Is there something special about the depth-of-field offered by medium format? Chris and Jordan look closely at the effect of format size on depth-of-field to find out.
Recent Videos
Shallow depth-of-field is often cited as being a key part of the 'medium format look' but is that actually true? Our Canadian friends shot some side-by-side examples to check.
Inevitably, since this video compares sensor sizes, it touches on equivalence and why it can make sense to think about equivalent f-numbers when comparing formats.
Please do not reproduce any of these images on a website or any newsletter/magazine without prior permission (see our copyright page). We make the originals available for private users to download to their own machines for personal examination or printing (in conjunction with this review). Please refrain from using them for any commercial purposes.
As theory goes, definitely there isn't any meaningful difference, in practise though, all presented pictures here has entirely different look, hardly to be describe as similar enough.
Cause there is thing called perspectivity compression. And MF has a bit more resolution to work with.
Not that obvious on these pics, but I saw what can be done with some decent mpix Phase One and Contax Zeiss lens and only similar combo with Hasselblad lens can touch that (well, Zeiss for Hasselblad really). FF is far from what can be done with such equip.
Its fairly simple, bigger area of sensor when mpix are not too densly packed is fairly forgiving to lens quality and MF lens are already pretty good.
Bigger size also means its easier to achieve subjectively higher resolution (cause that resolution simply can be spread accross bigger area).
It's worth noting that a large part of this is that FF lenses have really pushed the envelope lately. There are good 50mm and 85mm f/1.2 lenses that are sharp and contrasty wide open, unlike back in the film days.
I'd bet that the Medium Format 75 and 80mm f/2.8s back in the day were just a lot nicer looking wide open than most 50/1.7s (pro vs amateur optical quality), and while there have been 50/1.4 and 1.2 for ages, the sharpness wasn't there wide open, and even if it was, the much larger frame of 120 film allowed significantly sharper prints at any given size, since there was much less magnification needed. The filmstock was the same, so a bigger frame just adds more and more resolution, at a time when you needed it.
Modern FF lenses are just really fast and really sharp, a lot sharper than their film equivalents. That's probably the biggest thing that's closed the gap.
Digital medium format is also not "medium format" by film standards. There is 645 which is the smallest one (6x4.5 cm), 6x6, 6x7, 6x8 and 6x9. You could just as well call digital medium format "fuller frame" :-)
@AllNamesAreTaken - Digital MF is essentially 127 film shot 4x3, and I'd call that Medium.
@Snoddas - To be sure. But I guess what I mean is that the gap has closed. There was a big gap between formats in the film eras (perhaps somewhat unfairly, since there's probably a lot of pro 120 to consumer 135 comparison), and the process of printing photos where high MF resolution was a big factor. The gap today is a lot smaller today. Digital MF options have great lenses, immense color depth, and a lot going for them. But FF has grown up, too. Resolution gap is smaller, and not having to stop down means the lenses are brighter in practice rather than theory.
@AvailableMight - Many 617 cameras (I'm thinking Linhof and Horseman) almost are Large Format. They use LF lenses and focus using screens, but roll film is easier than sheet. To that end, non-crop 35mm panoramic cameras like Fuji TX-1 or XPan use essentially medium format lenses with 135 film. No critique, just musing.
I do not know why many people think about uber thin depth of field. I have my Mamiya 7 I occasionally use (which uses 6 X 7 format). Most lenses' widest aperture is 4 or 4.5.
It might be that a lot of people 'now growing up' are used to smartphone-like DOF- everything in focus. What's the best way to tell someone to get a discrete camera? Why, bokeh and all of that thin depth of field (and sure, low light).
I mean that's also getting torn down since modern phones (talking latest iPhone or Samsung) have started to brute force this using computational photography to get a depth map and emulate that small depth of field now.
As long as that fake bokeh is acceptable, people are fine with it. I've seen some poor ones, but my friends and family don't care- as long as the background is blurred, it's a good picture lol.
For me, the MF gives a technically “better” picture then an FF camera - for example the Sony a7Ivr (which I own, btw). After looking to different photos from digital MF cameras, seeing “same shot” comparisons with FF, I see an improvement in overall quality of the MF. I would describe the difference as “more natural” or “closer to reality”.
Now, what causes that difference for me? Well, probably a combination of neutral and more “separated” colors on the photo, more color nuances in skin color, hair color, So, for me, it is NOT about a difference in the photo’s absolute resolution! The MF photo is MORE NEUTRAL, MORE NATURAL and LESS AGGRESSIVE - that's what I see. Nothing wrong with my Sony or Pentax K1 (which usually gives me, btw, a little bit more of that “MF look” than my Sony - I do not know why, but that’s what I see…)
I understand that this is a totally subjective post, but why not to share my impressions with you guys?
I understand what you mean. Try to explain a different person why you love your partner. He probably won't understand, but it's there. Details. Overall feeling. The whole something. Dunno... But it's there.
Oh no - another “equivalence” sermon from DPReview. Please learn the relationship between focal length, aperture and distance to subject/distance to background. You will then realise that dof is about the lens and not the sensor...🤪
I'm not that convinced about your argumentation. If I grab my Arsat 30mm and i mount it on my Kiev 88, it is a fisheye lens. If I would do it on a 35mm camera, it is a mild wide angle. On an APS-C is roughly normal. If I take a picture with the same lens, with those 3 cameras, keeping the distance to the subject and I see them on the same screen, each picture is different, and the DOF too. And that's why the equivalence exists, to make our lives easier.
Of course, if you crop it to the area of the other sensor, you get the same picture. But guess what? If you want to know how that in advance, you can use the equivalence rules, and you will know. :-) Video-vs-photo, unfortunately I don't have an adapter to mount the P6 lenses to any digital camera.
Please learn about testing for single variables. People can be confused when a test is made to test for a single variable, because it is too simple. First you have to reduce everything to a common denominator, or the test is meaningless. 1 meter + 3 feet = How many feet? Conclusion: Don't buy a MF camera to get an advantage in DOF. Buy it for greater resolution and tonal range.
This article was about one - and only one - aspect of different sensor sizes.
Chris was objective in his approach to isolate the single variable that was the major point of the article - depth of field.
He didn't say that MF does or doesn't offer significant benefits over other formats. Those variables were irrelevant to the point he was trying to make.
He put together a test which indicates that that one particular variable isn't significantly different across those three formats when equalized in a manner that seems reasonable for such a comparison. He didn't say that that manner would be a legitimate method to compare the overall benefits of MF vs FF vs APS-C.
Seems like a good number of the comments are reading a lot more into this article than what the author actually said.
I love you Chris and Jordan. Why do you make such a Northrup-like video? Photography is more then depth of field.
Equivalency is also a kind of nonsense. How do you compare apples and oranges in a sweetness-equivalency, or a number of seeds equivalency? Equally sweet, same number of seeds… same fruits?
I have only shot medium format with film. It was definetily different from small-film. I understand, you don’t see the benefit of a medium format, but trying to prove it with a handful pictures….
@I like Lenses Math or physics are just tools! And you can create tons of nonsense with them as every other tool! Do not try to make from them something flawless and indisputable as usual pagan idols ......
Every theory you can build in maths and physics using the tools correctly is provably the opposite of nonsense. Example: non-euclidean geometry. Long thought to be useless flexing by nerds, today totally useful. And "pagan" is an insult invented by the followers of one particular religion against other religions, which makes it about as relevant to a well-informed discussion as the word "ante-deluvian" is to a modern geologist.
I like Lenses - Yes, it is a kind of nonsense. Don’t even you get the point of my comment.
How focal length-sensor size-aperture on the same subject can give quite similar DOF is clear. There are plenty of equivalency tables around.
But here is the conclusion, medium format has no unique depth of field! Based on equivalency they have “tested” with 3 cameras, 2 lenses each and a handful of pictures analized… Is it not a kind of nonsense to abuse physics and mathematics to produce a quite northrupish video?
It could have been interesting to use more cameras, more lenses, more subjects and a lot more pictures. So at the end maybe it could be concluded: Sensor size with the equivalent lenses and apertures does not seem to have significant effect on the quality of the DOF. Such a test would be honest, this looks more like YouTube food.
As you like lenses, you know that even on the same sensor, with the same focal length some lenses produce beautiful bokeh, etc… others not.
The point is, MF has slower lenses than FF, so you won't achieve superior bokeh because the depth of field will be at best the same. While you pay more for the exclusive MF gear. The only thing you get more is 100 MP resolution for the more expensive models.
"So at the end maybe it could be concluded: Sensor size with the equivalent lenses and apertures does not seem to have significant effect on the quality of the DOF. "
You don't seem to disagree with the conclusions, only the amount of samples used to get there. I'm not sure why this is- it's a strange nitpick, like you're uncomfortable with it being revealed so simply. You don't need a lot of test photos to bust the depth of field myth. Points about tonality of larger negative size etc aren't what the video is talking about either, if you want to bring that up.
@hetedik - I suspect you haven't bothered to inform yourself about that which you are arguing about. It is in fact quite simple to calculate that if you ignore wavelength effects and assume an ideal thin lens then the image produced produced on two different sensors by two different lenses with equivalent focal length and aperture are literally *identical*. There is simply no difference between the image produced on a medium format sensor or a FF sensor (or a phone camera sensor).
In other words any perceived benefit of medium format, or any "medium format look", outside of pixel-peeping resolution is either a hold-over from when film medium format was actually significantly larger than FF and therefore equivalent apertures were larger, or simply a confirmation bias illusion.
"assume an ideal thin lens" - this is the crux, lenses are not created equal and no lens is ideal, they just embody different compromises. Given that mathematically equivalent DOF can be produced, this is a good time to remember to buy lens systems, and then attach the cameras that fit them, not the other way around.
All lenses are not imperfect equally. Just because something is true for all systems qualitatively, does not make it the same quantitatively. An informed photographer buys into systems based on which compromises were taken in designing its lenses, so it is very relevant to the discussion of which system to buy into, all other things (such as DOF) being the same.
@I like Lenses " "Equivalency is also a kind of nonsense. " Sure, if you think maths or physics are also nonsense."
The thing is when have you ever used equivalences when out shooting? I shoot both MF and APS-C (yeah I'm a fuji guy) I can't remember a time when out shooting and I thought oh let see my 23mm GFX is equivalent to an 18mm on FF, no I probably thought I want a wider lens lets put on the 23mm. So yeah nonsense.
Equivalency is great for arguing about in forums and making sales of camera equipment to people who don't understand format sizes. But please tell me what is the practical application of equivalency?
In the end it comes down to lens quality/properties, assuming pure image resolution is of no concern. The faster a lens is, the more complicated it seems to corrects for aberrations. MF (even 44x33) allows for moderate wide open apertures still being fairly shallow in DOF, with no excessive lens aberrations to worry about. Sensor size alone is not the answer to the question this video tries to answer.
Surprised DOF topic keeps popping up every year depending on the site. I don't recall there being a wide debate.
Crop 645 MF allows you to get a wider FOV with a focal length. Thats the universal difference(other benefits as well but leave those out and just focus on the hard numbers). You can use a crop 645 MF 50mm lens and have a wider FOV over a APS-C/FF at 50mm. Using a FF 40mm or APS-C 35mm isn't the same thing. That just about matches the Crop 645 MF 50mm FOV but the focal length is different.
Medium format does have a secret sauce - sharpness. Though rarely included in equivalence debates, it's a key factor and easy to see if you print big (or pixel-peep).
Smaller formats demand higher resolution from the lens for an equivalent level of detail (because they need to be enlarged more). It varies by crop factor, so 30 lines-per-mm on full-frame is equivalent to 45-lpmm on APS-C, but only 23-lpmm on medium-format.
Sharpness is a combination of resolution (the fineness of detail) and contrast (how clearly those details are shown) and it's contrast that actually contributes most to our visual perception of sharpness. A fact of physics is that as resolution demands go up, so contrast goes down. Resolution and contrast are the two axes of an MTF lens test (Modulation Transfer Function - those squiggly graphs that lens manufacturers produce).
Smaller format push lens performance harder and they inevitably run out of puff sooner. When taken to the limit, larger formats always win.
Certainly true. For any given sensor resolution, the larger sensor has a significant pixel density advantage. Or in other words, a larger sensor will be able to provide a lot more resolution with an equal - or lower! - pixel density.
Now, this isn't so much relevant to DOF, but some other IQ factors that are frequently discussed are more a function of pixel density than they are of sensor size. (WIth the obvious caveat that beyond a certain resolution today's technology really only works well with larger sensor size.)
While true on paper, the actual sensor size difference is only 1.26x larger. So the differences are minor. It's like comparing m4/3 and Canon APS-C. Differences are academic at best.
I would argue the lens quality is higher on average, but you pay for it.
The other thing that's in play is learning curves in design and production which are a function of production levels. These effects likely used to privilege more mass produced formats such as APS-C, but increasingly are now more on the side of FF (as the amateur market falls away). Outliers like MF may not get the attention from producers they need.
In other words, whatever is the most popular gets the most attention from manufacturers - so those designs are likely to be higher quality for a given level of cost...
"Sharpness is a combination of resolution (the fineness of detail) and contrast (how clearly those details are shown) and it's contrast that actually contributes most to our visual perception of sharpness."
Well I thank you for bringing this up because I believe this part of photography seems to be far less understood compared to depth-of-field. IMO. I value a certain amount of details an or resolution way over what I think some folks perceive as being sharp.
It isn’t just sharpness of the lens, but diffraction comes in sooner. That said you’re defiantly in the realm of pixel peeping to notice such differences and even at decent print sizes, it’s not something most people are going to notice unless they’re allowed to zoom into 100% on the original file or are printing at wall-mural scale.
Resolution was no doubt the core of old film Medium Format mystique.
A standard 135 frame as 36x24mm exposed, 864 mm^2. 120 film shot 645 had 2324mm^2 exposed with a size of 56x41.5mm. That's a huge difference. So even if the 75mm f/2.8 lens is going to be pretty similar to a 45/1.7 (common on a lot of integrated lens rangefinders), you've got a tonne more resolution on 645. Sharper in-focus areas, better transitions to OOF areas, nicer prints, etc.
The thing that 135 film had going for it was shutter speed. Want a deep DOF, get everything in focus with a standard FOV lens? F8 gets you there in FF, but you'd need to between f11 and f16 on 120 film, and even higher on sheet film, because the standard focal length keeps growing. The old Ansel Adams "Group f/64" shooting 8x10 with something like a 400mm lens... is just f/8 and be there on 135 with a 50mm, so you can get away with a faster shutter speed. But you've lost detail, and can't reproduce as clean an image as large.
Isn't APS-C the only format that's actually pushing things harder? M4/3 too to an extent but the pixel density of a 61MP FF sensor is about the same as that of a 16MP M4/3 sensor (20MP isn't much of a jump) or a 100MP GFX MF sensor no? Heck most APS-C bodies are in the same ballpark too except the most recent ones going >30MP, Fuji's are right on the same line at 26MP... Is my math wrong?
I guess 42MP FF bodies and particularly 24MP ones are under that curve but looking at the highest res bodies from each format the demand on lenses doesn't look that different, I do see a lot more exotic UWA designs for larger formats tho and as an UWA lover that's hard to ignore.
I think you made a typo when you suggested resolution and contrast are the two *axes*. Typically resolution is the 30 and contrast is the 10 lines per mm curve. So they are the two line pairs not the two axes. (mtf curves come in different layouts but typically with four lines sagittal and tangential at two line pair densities)
Contrast is much misunderstood as it has two meanings in photography. The mtf one and the common usage of high dynamic range.
@Impluses the point is that at a certain level of detail the larger format sensor will be asking less of the lens so the image will be rendered with a higher standard of contrast. Pixel density is part of the whole picture so to speak but is a bit of a distraction here. The MTF graphs that lens manufacturers produce are purely optical and ignore the sensor completely, ditto Roger Cicala's.
@Lessiter No typo from me but what perhaps I should have said is that 'traditionally' the two graph axes are resolution and contrast. While it is always true that resolution and contrast are the two parameters being measured with MTF, there are different ways of presenting them. Canon for example puts % contrast up the vertical, and the horizontal axis is distance from the image centre; resolution is expressed with different dotted lines.
Canon has also changed the explanation for their MTF graphs in recent years to make them easier to understand, but what they now say is misleading IMHO because both the 10-lpmm and 30-lpmm graphs do indeed show % contrast. What I think they're meaning to say is that if you want nice crisp and punchy images for normal viewing then look at 10-lpmm, or finely detailed landscapes for larger exhibition prints then check 30-lpmm. Personally, I want both and would argue that 30-lpmm is a better indicator of overall performance.
@Richard Hopkins now show me examples of award-winning photographs from the digital era that are examples of "taken to the limit", and would not have won the award and looked the same with the next smaller sensor.
And that is the myth of the large sensor: yes, they "pull away" when you get to the extreme edge case, but hardly anybody spends any time at the edge case.
Classic example: people using 44x33 cameras on tripods and shooting at base ISO. Well, gee, if you had used the same techniques with 36x24 you are nowhere near to an edge case, and would look great as a final product, viewed whole.
What's weird to me is that lenses are basically projectors.
Sensors are basically screens - the further away the screen (sensor) the larger the image.
It's really as simple as that.
25mm micro four thirds 50mm full frame
Those millimeters are the distance from the focal point inside the lens to the sensor. (ignoring retrofocal designs for wide angle lenses in SLRs etc.).
So why should moving the sensor (screen) further away from the lens (projector) make the image sharper?
@saltydogstudios - It doesn't change the sharpness of the image projected onto the screen. But maybe you can take more measurements to get a sharper image.
The sensor isn't just a neutral screen--it has a fixed amount of photosites, of megapixels, collecting information.
Thought experiment: a slide projector projecting an image onto graph paper: 8x10 with quarter-inch squares. Fill in each square with the main color of the image projected onto the screen, 320 squares. Suppose you tape four pieces of paper together, move the projector further back. The image is now projected onto a space that's 16x20, and you'll fill in 1280 squares. Second image has more measurements and more detail.
Now, FF doesn't necessarily have more detail, it just usually does. A 25mm lens on 20mp m43 image should have more detail than a 50mm lens on 12mp FF image, and there are well-selling camera with those specs.
But if you're using pixels the same size, you can put more pixels onto a larger sensor.
Not to forget that superior sharpness also increases (perceived) subject separation, as the difference between in-focus and oof areas is more apparent. Conversely, I've had non-photographers remark "this photo is really sharp" when seeing an image recorded wide open - they saw a strong sharpness contrast, which helped the sharp area stand out more.
I hope this is useful, it is often overlooked but explains why everything in the imaging chain that affects magnification - from capture to viewing - affects the final DoF.
The whole concept of DoF is simply based on our eyesight, ie the average Joe can detect detail in a print down to 0.2mm wide when viewed at a distance of 10in (say a 10x8in print).
That's it, and everything in the DoF formula is worked back from that basic assumption of 0.2mm. It's been universally applied for decades and still works because while technology marches on our eyesight stays the same.
In any assessment of DoF, it is essential to maintain the correct viewing distance or the whole thing falls apart. Happily, so long as the viewing distance is roughly equal to the diagonal measurement of the image everything holds good, eg a billboard from the other side of the street looks sharp. But if you are pixel peeping and viewing a small section of an image on-screen without moving your chair back, all is lost.
You can’t get the “medium format look” with Fuji’s slightly larger 44x33 sensor because Fuji’s native GF lenses are just too slow. Full frame cameras can actually get closer to the “look” thanks to their ultra fast lens options.
So if you bought a GFX camera for high resolution work you might have gotten your money’s worth. But if you bought it for producing a “medium format look,” you just flushed a bunch of money down the drain.
There are very few f/1.4, f/1.2, or f/0.95 medium format prime lenses out there (or f/2.8 zooms for that matter).
That said 100+MP resolution can be very impressive if you need it either very very large prints (or covering the wall with high resolution wall paper), allowing viewers to zoom into 100% on websites with zoom viewers, or cropping to fine details in post. Working for museums and doing art reproduction is about the most common use for these kinds of uses. But 100MP on smaller formats runs into issues of softeness due to lens limitations and diffraction. So for that we tend to stick to formats a little larger than the 44x33 and go to the 645 format sensors that cost as much as a car.
This is not something unique to Fuji; even in the film days you could barely get f/1.4 equivalency on Hasselblad 500 cameras, with most lenses topping out at f/2.8 at best. (the focal plane shutter 2000 series with its one-stop faster lenses was much less common due to reliability issues from what I've read)
If you really wanted ultra shallow DoF, you were probably best off with the Pentax 67 system, which had a huge negative and a modest range of f/2.8 lenses (and one f/2.4).
I believe the 'medium format look' to be more about using a longer focal length to get the same field of view than it is about aperture size. The longer focal length offers a stronger compression of the scene which is what the look is about IMO. Using the 105mm on a Pentax 67 gives the same field of view as a 55 mm on a full-frame but a totally different looking photograph. You get a very different subject isolation and scene compression. It kind of feels more relaxed and less stressed than smaller formats. It's hard to explain.
Now I agree that the current 44x33 sensors don't get you as far, but I definitely see the difference vs 35mm.
@JelleNL Look at the video because they did use longer lenses on Medium format to match the framing… it didn’t happen as you describe. Compression and isolation is due to distance from the subject, so if you use a longer focal length to stay at the same distance with the same framing, it doesn’t change compression or isolation.
I agree that compression is the same when you equivalence it all out. I wonder though if anyone knows if the fov/focal length relationship is important for lens design? Does it help MF that the focal length is longer for the same fov. At wide and standard that is. Or is focal length irrelevant and only the fov determines the complexity and compromises you have to make?
good video, it clarifies some misconceptions widespread especially around crop sensor users. FF is where its at when you want to use DoF as a creative tool without breaking the bank. No such thing as 35mm f0.95 in native GFX mount. Gotta admit, in some cases i think MF lenses will render bokeh a bit nicer than smaller sensors.
When m4/3 cannot match FF equivalent aperture with their expensive lenses, people say it's a weak and overpriced format for amateurs. When GFX is in the same boat it's because their owners are too advanced and sophisticated for such things? Ironic.
your creativity is stimulated most by FF lenses, theres no doubt about it. Affordable lenses from 9mm (rectilinear) to 800mm (fresnel). If you can live with limited options in either UWA or Supertele or DoF you might be good with other formats than FF. Better be sure beforehand that in your project you dont need all the (extreme) options you couldve had with a FF lens selection.
If you want options, neither mft nor medium format is for you. They are quite locked in and limited, also the limited R&D cameras get mostly go to fullframe.
There is not 3 things that affect DOF only 2 ..if the framing is the same ONLY the aperture and distance from the camera will affect DOF ..focal length will not
There are actually five factors affecting DoF - focal length, distance, aperture, format and Circle of Confusion. If format and CoC are given, then you can reduce that to two - magnification and aperture - but that's not very helpful in practice.
To be able to measure and compare DoF in practical terms, you need to know focal length, distance and aperture (assuming that format and CoC are fixed).
It really comes down to magnification. If you make your subject larger in the image, the DOF will decrease. If you make your subject smaller in the frame, DOF will increase. You can do this via focal length changes or changing distance to subject. For a given magnification of subject, DOF will be the same at a given aperture and sensor size.
Focal length and distance to subject are really just inputs to calculate magnification.
Magnification is also what impact DOF on the output size, so the CoC, format and print size part of the equation. If you magnify your image for display, DOF will decrease. If you make it smaller, DOF will increase. You can even see this on a digital image by comparing thumbnail view with full screen.
"Magnification is also what impact DOF on the output size, so the CoC, format and print size part of the equation. If you magnify your image for display, DOF will decrease. If you make it smaller, DOF will increase. You can even see this on a digital image by comparing thumbnail view with full screen."
Not if you respect the standard viewing distance and keep that the same, ie viewing from the same distance as the diagonal measurement of the print. This is a fundamental of the DoF formula.
Just to add another point that often gets overlooked - if you crop an image, you change the DoF (because you have effectively changed the sensor size).
Richard yes if you get pedantic enough according to some, your reading glasses and the paper that the photo is printed on changes the DOF but in a practical sense no and yes you need the 3 parameters to calculate DOF the FL, Aperture, and distance but what i say still stands that if the framing is the same, only the Aperture and the distance has bearing on the DOF and not the FL
"Not if you respect the standard viewing distance and keep that the same, ie viewing from the same distance as the diagonal measurement of the print. This is a fundamental of the DoF formula."
If we respect that, then they are always the same and can be considered irrelevant to the formula.
Cropping an image only decreases DOF if you then display it at the same size as you would have the uncropped image. In order to do that, we are back at increased magnification. But if you cut a print in half the DOF has not changed because we left it half the size of the original print. So there was not a Change in magnification.
Take an image with some nice “bokeh balls” in the background. Crop it. Compare both crop and original at same size side by side. Bokeh balls are now bigger in crop, right?
tkbsic what does that prove??? if your subject is in focus and you crop Compare both crop and original at the same size your subject is bigger but still in focus.
@tkbslc Sure, but we are comparing DoF not some balls size! As if I do not have lights on the background, I will have no Bokeh balls so no DoF? :) Or if I just magnify 2x original picture, here we go double DoF? He he he, easy-peasy!
It’s just an easy to indicator to see, it’s not the sole proof. If you compare the actual plane of acceptable focus, it will also change. Same as how DOF changes based on viewing size or viewing distance.
well nobody answering my OP lets move it up a notch ..now you have a 70mm F2.8 and you have bokeh balls 10 feet behind your subject ..you keep the framing the same but you are now at 200mm F2,8 ..your balls are huge but the DOF is the same ..prove me incorrect..... like i say FL has no bearing on DOF if the framing is the same ..the vid is incorrect
As davev8 if you keep subject of interest same size does not matter how and what will crop nothing really changes. Just framing or FoV. Who cares about bokeh balls or focus change when your subject is half in the frame or out of the frame...... Or we will make our subject microscopic size to get better DR and noise performance, and to get all frame in focus..... So in general all this is total useless and not practical nonsense :D. Pure example how science could be used for totally useless comparisons and theories.
It is very useful to understand the relationship between DOF, magnification and bg blur. For example, a headshot at 200mm f5.6 will have the same bg blur as one at 100mm f2.8, but the f5.6 will have double the DOF. Now you get your bg blur, but you keep nose in focus, too. Or maybe you use this knowledge to understand that the DOF is the same at 1x macro regardless of the focal length used.
That's a very good point from tkbslc - DoF and background blur (bokeh) are not the same thing and the f/number is only part of the story. In fact, in pactical terms longer focal length is often the better option (and cheaper) as you get all the blurred bokeh, enhanced by the enlarged and simplified background, plus more DoF. This would make a good follow-up video, especially as I think it's bokeh that most people are actually wanting when it comes to shallow DoF.
To extend your examples, in terms of background blur 200mm f/2.8 is roughly the same as 135mm f/2 (my choice for portraits), same as 100mm f/1.4, 85mm f/1.2, and 50mm f/1.0 - provided you are able to adjust the shooting distance and move back to maintain the same framing of the main subject in say a portrait. This also requires the background to be at infinity to hold good, but it is also substantially true at closer distances. This is a very good blur and DoF simulator https://dofsimulator.net/en/
When it comes to DOF and bokeh look, there is another myth, the difference between mechanical and electronic shutter that is supposed to create a slightly different bokeh look.
Fully electronic and fully mechanical focal plane shutters should not differ in that regard. EFCS however, which moves one curtain electronicaly and other mechanically has this effect at higher shutter speeds.
The leaf shutter DOES impact DOF because the shutter starts opening at the center meaning the photo is essentially an integral over a series of photos of different depth of field with the small aperture getting the most time.
Medium format cameras used to have leaf shutters (electronic shutters for slow or non moving subjects eliminate the issue) and leaf shutters will tend to slightly increase depth of field and change the transition from in to out of focus subtly.
My little Pentax Q lenses also have leaf shutters. It is a little surprising with the all the benefits those have that they are not used more often. The Minolta Alpha/Dynax/Maxxum 7 film camera had an STF mode that would take several exposures while varying the aperture to get a similar optical effect with the drawback of all those exposures.
It does if there’s movement aka moving leaves. Before stacked sensors, e-shutters were typically slow enough to have artefacts so the bokeh would look weird.
There's no myth, but it's getting conflated and confused, EFCS *will* have a negative effect on bokeh with faster lenses at very fast SS and DPR has covered this in the past, it leads to a fairly obvious truncation of bokeh balls but it's still subtly visible in other areas.
Most manufacturers simply disable EFCS and go full mechanical (or full e-shutter) past a certain SS threshold, I think Sony might be the only one that doesn't or doesn't offer the option to do so, hence you have to micro manage the EFCS a bit if shooting bright primes wide open in daylight without an ND (so you're topping out SS at 1/4000 & 1/8000).
E-shutter itself won't really have an impact on bokeh per se unless it leads to banding or rolling shutter distortion, and if there's any motion it will on most bodies that read at 1/30 or slower.
DOF is often confused with background blur, the latter of which is a function of the physical aperture size (focal length / aperture) relative to subject size. Here are some articles on the subject, including a link to an online background blur calculator where I plugged in the telephoto values from Chris's video...
Yep! DOF takes into consideration the WHOLE field. It involves foreground, mid ground, and background. It’s that relationship that’s being compared in an accurate test like this. When people only discuss background blur that misses the point. Given a large enough f-stop and a position to stand in, you can blur out the background even with a wide angle lens. But that ignores what the rest of the field will look like and the relative distances to the near and mid fields. The whole argument is mostly academic and ridiculous. Does it REALLY matter? Just go out and shoot, people! Get out of the basement and BE a shooting photographer not an armchair philosopher. 🤣
@tirmite, There is no direct relationship between DOF and the quantity of background blur. Perhaps you're conflating how DOF is described in terms of the foreground and background. That's separate and distinct from the concept of background blur. Perhaps have a read of the articles I posted.
"the latter of which is a function of the physical aperture size (focal length / aperture) relative to subject size"
That's how you work out infinity blur. Calculating the blur for things closer to the subject is not so easy. For example, imagine photographing someone standing in a field of flowers. The blur for flowers right behind the subject will be more impacted by the DOF and how fast focus falls off. The flowers at a distance will be more impacted by the infinity blur (aperture size). The math is a little complex in the real world.
But yes, this is the reason why longer slower lenses can have both deeper DOF and more melted backgrounds at the same time.
Nope. Not if it’s what you want or need. If you’re rich you don’t have to justify it. If you’re a working pro like me you can justify the cost because there is an obvious quality advantage for your clients IF they need it and/or appreciate it. You pick the camera in your arsenal that’s appropriate for the task at hand. Often 35mm FF is plenty of resolution and there are unique lenses available (wide angle tilt/shifts for architectural) that make that format desirable. But there are times (or clients) that demand a step up. If I’m selling BIG prints, they are just going to look better because the image doesn’t have to be enlarged as much to get the same size dimensions. PERIOD. More pixels and spread out less over a large print are just going to have higher resolution. PERIOD. And THAT is often worth the extra investment in the more costly equipment.
" there is an obvious quality advantage for your clients IF they need it and/or appreciate it"
99,98% of your so called clients will never see the difference, nor perceive it. The quality difference is there, but unnoticeable if you don't know what to look for.
"If it can 'justify' the cost" is highly dependable upon the work you do.
You say Full Frame is good enough for architectural work. Now let that be one of the fields in which you rather have a larger sensor than a smaller... It shows you have no clue why you are actually opting for (small) Medium Format.
Not to say that there are hardly any true digital medium format cameras around. The GFX is a toy and a marketing ploy. A smart move of Fujifilm to let you believe you have the advantage, until you work with true medium format, wether it be a Phase One, a Hasselblad or just analogue Medium Format and use film instead.
If you truly think pixels matter so be it. But a true professional knows its not about resolution nor DoF.
I should have specified any 50MP semi medium format camera, such as the one in this comparison. If someone wants to blow the price of a Phase One, more power to them. As it sits now, I have 60MP at my disposal, slightly outresolving any 50MP mini "medium format" camera. This year we'll probably top out at 100MP for the a7rV.
But then again I seldom shoot fewer than 5 to 10 frames stitched for my scapes, producing a 1-2 GB tiff with much better geometry and sharper corners than any single shot medium format camera PERIOD. As far as portraits are concerned, I don't want to see anybody's mug at 30X40.
Supposedly medium format gives better ISO performance because of the larger sensor but the lens choices are the problem and mainly suited for a small number of genres. With the A7R4 you could do wildlife, good luck with the GFX on that regard.
I’d rephrase your statement… A lot of money is wasted on Medium format by people who may get it just because they think it’s better. However I regularly use medium format to have the highest resolution possible because I’m photographing artwork that is very old and expensive, and I want to have an image that can be reused and repurposed as many ways as possible in the future without having to re-handled the artwork again. So there is value in medium format for me. But I’m not everyone.
@Francis85 Obviously, you vastly underestimate the intelligence/perception of your clients. A foolhardy posture for a professional if one knows what's good for one's business, reputation and artistry.
Always funny how the people who have never used the gear and lenses are so quick to provide an uneducated opinion about a topic and the quality and benefits thereof.
I have put in two more 50MP cameras. The Canon 5DS R and the Pentax 645Z. Interestingly, it is just the Sony where the People have yellow faces. (Despite WB beeing corrected in ACR:
Now compare the Phase One IQ4 (150MP but 645 format, larger than the Pentax or Fuji 44x33 sensors) to the A7R IV in pixel shift mode (which should be about 200MP). It’s an extreme example but shows you will eventually hit diffraction at high enough resolutions on smaller sensors.
You can throw in the Olympus E-M1 III in pixel shift. The jpeg holds up pretty weill against the 50MP 24x36mm Sensors. Not quite the same, but pretty good.
The Problem with smaller sensors is indeed the problem of the entrance pupil. The lens should perform very weill at f/2 or so. The test there is done at f/5.6.
There is only one advantage and that is 100 MP top resolution. I'd like to see the prints where it makes meaningful difference over 50-60 MP FF sensor. And for capturing art, the two recent projects I saw used stitching with Canon R5 and Sony A7R cameras. Apparently it was good enough.
I mean sure, if you have specific clients which will pay you more than double just because you shoot with MF gear costing much more than FF then it makes economical sense. But to me that looks like very small niche. For the rest there are advantages like focusing speed, shooting and processing speed, better video quality and options, more lens options, and all that at much cheaper price than for MF gear.
1) E-M1 Mk III is a 20MP camera (5184x3888). Pixel shift would double both directions... making it an 80MP camera (10368x7776). And you say it holds up decently "not quite the same, but pretty good" to a 50MP 135 format camera that tells me that it's not quite delivering 50MP worth of information. Maybe it's resolving closer to 35 or 40MP? You're getting 160% the amount of pixels but still not getting 100% of the information.
2) The pixel pitch of a 80MP M43 sensor is 1.678µm. A 50MP 135 sensor is 4.14µm and at 240MP (a7r IV pixel shift): 1.89µm. A 100MP 44x33 sensor is 3.76µm. A 400MP 44x33 sensor is 23296 1.88µm. For 645 sensors the pitch are 100MP: 4.6µm, 150MP: 3.8µm, 400MP: 2.3µm.
3) In my experience even at f/4 there isn't noticeable gain in detail with a 44x33 at 400MP or 135 at 260MP. On a 645, a 400MP capture can render more detail but you need f/6.3 or wider
4) f/2 is problematic if you're looking for max detail because of lens sharpness, field curvature, and DOF.
Why are so many folks always talking in terms of printing images when I am pretty certain, most images are never, Printed. Why are some folks worried so about pixel piping. I do it to check how well I achieved critical focus. Also, if you really want to get into something real deep, I think Hyperfocal focusing would be an good choice next.
Concerning this topic, I bet almost nobody has changed their mind, one way or the other.
You are correct. In this era most people are viewing images on a screen which has a much lower dpi resolution than printing with dots of ink. But those of us who used to make prints or still do or who shoot for clients who will make prints, brochures, run magazine articles or ads STILL DO have to be concerned about how images will look at 300dpi printing resolution. That JPEG that looked good on your smartphone or iPad screen (72-144dpi) will suddenly look like crap when printed at “photo quality” in a double-page magazine spread or a back-lit poster size ad in an airport terminal. You just can’t take a 100 piece puzzle that’s meant to be 10X10 inches and then spread it out (expand it) on a 20”X20” tabletop and expect it to look as good.
I get that for folks that actually do print their work, mostly. It's just for most folks, it's really not something they have to worry about much. IMO. An even those that do rarely print, most usually understand the parameters.
Funny.... magazines, and brochures is offset print CMYK. Hardly any of these media is printed at 300 dpi. Take a loupe and you will see each individual dot with ease.
Brochures, magazines and actually all printed media including those large sized billboards don't actually need more than 6 megapixels to look good already.
In the last 20 years we have gained on sensor resolution. Yet, the perceived extra value in print has been mostly unnoticeable.
It shows sensor size and resolution don't matter much. I personally would even go as far as don't matter at all. In that sense cameras like the GFX are a waste of money these days. Now if you do true fine-art it may be another thing. But for commercial photography any camera will do as fine.
You can check for yourself. All you need to do is take an old magazine (let's say one of 15 years ago) and lay it next to one you bought last month.... Good luck noticing the difference.
I never used to print in my youth but I'm doing more of it now as I near 40, I'll also look at photos on my 65" OLED and/or crop with more liberty on a higher MP body (tho I had no qualms with a 1.5x crop even on my 16MP M4/3 bodies, as long as I was using a good lens)...
Just because *most* people don't print doesn't mean that should be taken as a given when you're handling subjects like these amongst enthusiasts who are much more likely to do so.
Of the people that print, most print at 300dpi or less on canvas or similar decorative items, not fine art paper at 2880 dpi and hundreds of bucks a pop, so there's that limitation in the output medium, as well. I have printed canvases from 12mp 2/3 sensors and a shower curtain from a 16mp APS-C sensor and they all look *great*. Landscapes at F8 at 24mp printed at 24 by 36 inch, I can glue my nose to the canvas and pixel peep all the detail to my heart's content, amazing. That said, I took some portraits with the GF 110 and the lens quality jumps right out at you, great skin and iris detail and such buttery sharpness falloff. Love it. The fact that you attach it to a 4433 body is a minor detail.
You buy a medium format camera for ONE reason: the sensor is bigger so you don’t need to enlarge an image as much as the smaller format cameras. Therefore the resolution is higher comparatively. OR it allows you to enlarge more without losing quality. Also… is you really, really want super shallow DOF, get a fast FF lens that produces a large enough image circle to cover the MF sensor, put an adapter on it, and mount it to the MF body. No need to wish for some exotic high speed MF lens that will probably never be made because there’s not enough demand for the manufacturer to invest those kinds of resources.
The key advantage of large sensors comes down to diffraction. We can make sensors ridiculously high resolution so enlargement isn’t the issue it’s diffraction onto the sensor. Bigger sensors can use smaller apertures before diffraction costs resolution allowing GREATER depth of field. Good for Ansel Adams, not so much for bokeh fanatics.
M43 loses resolution to diffraction at f4, FF at f8, MF f11.
Except that the common / cheap medium format sensors aren't that much bigger than FF.
When people started talking about the differences between MF and FF they were comparing film, and comparing 60mm x 45mm (or bigger) with 36mm x 24mm. There is a far bigger difference between those than these tiny 44mm x 33mm "fake" MF sensors.
It would be interesting to revisit this conversation when someone brings out a 60mm x 60mm sensor. Might have to wait a while, though...
We had 3.5 x 4.5” (8.9x11.4 cm) “sensors” 20 years ago. They were scan backs. It could take 12 minutes to take an exposure but it was basically a large format.
A full sized 6x6cm single-shot sensor would be prohibitively expensive without resorting to tricks like photographing a ground glass with a smaller sensor or going back to scanning sensor.
AlephNull "...aren't that much bigger..." At nearly 70% more area, I'd consider current digital-MF a whole lot bigger than the 24x36 'miniature' format.
I wonder how many of these "It's not that much bigger" types spent most of the last decade finding fault with M43 compared to APS-C because that sensor size difference was so, so huge and important.
Tonio Loewald: But depth of field also depends on the format, so you don't gain anything. In your example, both depth of field and diffraction blur is the same for M43 at F4, FF at F8, MF at F11.
If you want to do it by area, then a REAL MF sensor would be 56mm x 56mm. So let's look at those areas:
36 x 24 = 864 mm2 44 x 33 = 1452 mm2 56 x 56 = 3136 mm2
These fake "MF" sensors are less than half the area of the real thing, despite having almost 70% more area than FF. To be real MF sensors they would need 260% more area than FF. Now THAT would be "much bigger".
For that matter, an APS-C sensor is around 25 x 17 = 425 mm2, which is a bit under half the area of FF, just as these fakes are just under half the area of a real MF. They are more like the APS version of a real MF sensor.
If you stop call them MF sensors - call them JF sensors, perhaps (because J is halfway between F and M, so JF is halfway between FF and MF), then I'll happily agree that they have the potential to outperform FF sensors, all tech being equal (a stacked JF sensor could be quite an impressive piece of work).
Just stop pretending they are MF sensors. They aren't.
In the digital realm, they are between full frame and much larger sensors. And if you think GFX is portly compared to FF, Canon makes a sensor big enough to fill an 8x10 view camera.. but you probably wouldn't want to lug it around unless you had assistants.
However we're no longer in the film world. So a moderate amount more of surface area, such as 70 percent greater (than full frame) in this case can and does make a lot of difference.
One thing Chris's at-home test didn't take into account, is how much wider the FF and APS-C cams had to shoot in order to match the more moderate apertures of the MF camera. It's easier to design a lens that's optically excellent wide open if that max aperture is f2 than it is f1.2. Full frame lenses that manage this become as big as their slightly slower MF counterparts, and often times even more expensive, lens for lens.
There is so much more to photography than just looking at DoF. If DoF is all that concerns you then you have chosen the wrong hobby or should immediately start learning what photography is truly about. You'll soon see that sensor size really doesn't matter.
Well this topic is about DoF, not everything else one might might be concerned with. So to suggest otherwise might mean you need to learn more about photography as opposed to assuming what you think everyone else knows about photography.
Hi Francis85, this is a gear site where people like to discuss different technical aspects and capabilities of gear. Maybe you should visit a car forum and say the size of your engine doesn't matter when you need to get groceries.
Sure - If choosing the largest possible aperture (preferably lenses with an largest f-stop of 0.95) , and a camera with the largest sensor is what you care for most then you are not a photographer, but a gearhead. (By the way good luck getting sharp images with that combination...)
The concept of equivalence is very clear to me so don't worry. If you have concerns about the size of the sensor in your camera or you think you only need lenses with an extreme large aperture than definitely there is some misunderstanding for you over the concept of what photography is truly about.
@Francis85 .... I certainly don't care most about it, nor do I think a lot of other people here. Besides, here's a shocker for you: it's okay to be a photographer and still be interested in gear and its technicalities. It's called *gasp* curiosity...! It might be a good idea for you to just go out and enjoy what you like doing instead of worrying about what concerns other people.
@francis: I bet that if you show a series of (full body) protrait images to a bunch of random people, some with strong separation (f1.2) vs some with moderate separation (f2.8) of the subject (model), most people will prefer the one with the more blown out background. There is an aesthetic to it, people do see and appreciate it. If its not your thing, i think youre in the minority. if you use f1.2 in a group shot however, youre an amateur, or very very talented (if you get everyone in focus that is)
For those saying your 50mm f1.2 "just fine" on GFX -
Your 50mm f1.2 will not look like a f0.95 FF lens, or will not have the equivalent depth of field, on GFX because depth of field will be deeper towards the edge of the frame - caused by excessive vignetting using a (fast normal) lens designed for a smaller sensor.
A certain amount of optical vignetting is tolerated in fast normal lenses, that the excessive vignetting from adapted fast normal GFX lenses is not equivalent to. And you can't ignore this because you won't have as shallow depth of field.
The things in the background on the left side of the wardrobe are farther away from the focal plane thus they look more blurry. Imagine where the focal plane would be and follow it farther outside of the image, you would notice it will touch the wall the right side of the wardrobe. Take some Photos at a perfect 90° angle towards it or it wont prove anything. Also field curvature will come into play, some lenses really do what you describe however another copy of it might do the opposite (but getting foreground objects more in focus due to the focal plane being bent closer towards the camera and background objects even more out of focus than the in the center).
It's the physical size of the hole (the ƒ-stop) which controls the depth of field. An ƒ-stop is a fraction of the focal length of the lens which gives the same 'light passing' equivalence regardless of focal length. So smaller formats have smaller focal length lenses and thus smaller physical apertures, therefore the greater depth of field.
Except you’re missing the whole fact that that only smaller sensors will have the image enlarged more making the circles of confusion represent larger. Of course it takes a little more math to understand that.
kb2zuz, you are of course correct. But depth of field is determined by the physical diameter of the aperture regardless of the amount of the enlargement. The circles of confusion may be enlarged but so too will the depth of field which remains constant dependant only on the aperture. It's a bit difficult to explain in just a short comment and would benefit from some diagrams.
Well, that’s good, I am pretty sure that I heard Chris saying the focal length of the lens effects DOF. I know that people have told me, in the forums, that I am wrong if I think that. I was informed, lens focal length has nothing to do with DOF.
But even so, my experience has been that it does, a lot.
If you are wrong, then DoF calculators are wrong, too. Every DoF calculator I have seen requires the input of the focal length and the f number and the subject distance.
Maybe the people you have been reading in the forums are using some other factor which works out to provide the same thing? For example, the f number is the focal length divided by the diameter of the aperture. If they use some other measure for the aperture, maybe that accounts for the focal length?
For example, let's imagine that instead of using f/2 (meaning the aperture is half the focal length) they measure the aperture against the height of the sensor or the size of the individual photo site / pixel - that removes the focal length from the equation.
DMartin92, rules like that omit critical information on how to make the comparison. You and Chris are right if you don't change the distance to the subject. But if you move further away with the telephoto so you keep the field of view constant, the depth of field depends only on the f stop, but not on the focal length. (And both statements assume the same sensor size.) The background blur, however, increases with focal length.
It's pretty simple to test out if you have a lens that does wide angle and one that does telephoto. Even point and shoots can do that. Shoot the same distance to subject and same f/stop and the DoF is the same at 14mm and 200mm.
The framing will be completely different but if you crop the wide image to the view of the telephoto image they have the same DoF. F/stop matters, distance to subject matters, focal length doesn't.
Kerensky97, no. Same camera, same distance, cropped to same field of view: in that case depth of field depends on one angle only, which is determined by the aperture (diameter) and distance to subject. A longer focal length means a larger aperture, hence a larger angle for the same f stop.
I won't argue it here. You won't get this stuff from a depth of field table, except by error-prone calculation. If in doubt, discuss it the Photo Science and Tech. Forum.
What I'd like to see is the same image fov (cropping the larger sensors' image to the smallest one) taken with the same lens (take a Pentax MF lens adapted), from the same distance at the same aperture from MF -> FF -> APSC -> M43
What strikes me as obvious here is the general availability of faster lenses for FF. There are plenty of 1.4 and 1.2 lenses for full frame, and even f1.0 and f.0.95 are possible. The equivalent of these (calculated in the way the video does) are simply not available in MF or in APS-C with anything like the same common availability, if at all. When they are available, they are huge and cost a fortune.
Those who shoot Sony or Leica M (both available with excellent 60MP sensors) not only have the excellent Sony and Leica very fast lenses, but also the superb, fast Voigtländer lenses at very reasonable prices. Where are the 35, 40, 50 equivalent f1.2 lenses for MF or APS-C lenses even remotely as small as the Voigtländer lenses at under €1000?
Full frame, in this respect, seems to be an obvious choice. For many other reasons as well.
Yep. Many even reasonably priced FF lenses need to be stopped down to achieve DOF equivalence with MF lenses wide open.
FF is the sweet spot for total light collection and minimum DOF if that’s what you’re into. There are probably a few exceptions in different systems at particular focal lengths but for the most part it’s pretty clear.
My deepest respect for this video and your courage. It Is well known that mentioning the 'e'-word inevitably brings you into deep trouble. Gear fanatics love their myths. The small sensor hardliners will insist that their camera and lenses can do everything, just in a smaller package ('f-stop is f-stop') and the lager sensor worshipper will be offended simple by the comparison. There are not many decent gear heads out there who appreciate the naked facts about photography. Like me - as long as you don't question the magic Leica look of my 50 mm Summilux 😜.
But discussing whether an ever decreasing DOF is generating photographic value would be such a distraction from gear tribalism. Don't want that, obviously.
Because before you use that gear, you have to buy it. And when choosing, it makes sense to consider whether, given your shooting style and needs, one format or another has advantages for you.
@ Adam007: People buy cameras from sensor size and features, and pick lenses from angle of view, aperture brightness and resolution. We chose brand from preferences and bias.
Equivalence is of very little help for guiding you to what to buy, since this way of thinking and comparing typically introduce more confusion than clearing up things. This mostly something from the SLR age, brought into digital when most DSLR sensors filled the aps-c format, the half format compared to the 135 film format
Sensor size and equivalencies are probably the most disputed topic across DPR forums. I mean, it comes up all the time - sometimes by indirection - and there are plenty of people basically arguing about which format is best. There's no good answer to that question, but there are better answers to the question, "Which is best for *you*?" - and these discussions are useful for anyone willing to learn a couple things before dropping X grand on a new system. Even beginners somehow know to wonder about whether they should get APSC or FF (to pluck the two most popular candidates for new ILC users).
Magnar When my Z7 dies and I have to use my Nikon Zfc as a backup I can still use all my full frame lenses but I want to calculate in my head which one will do the job I want.
I wanted to take the Z7 out with the 40mm f/2.8 lens for some street photography. What lens can I swap out for a similar experience now that I'm stuck using my Zfc?
@ Kerensky97: Why not just look into the viewfinder? I am sure that when you move from the Z7 to the Zcf, you don't calculate all aspects of "equivalence". You just use what you have, and shoot.
@ QuietOC: Why don't people calculate "equivalence" when cropping files? This has the same effect as using a smaller sensor ... people post tight cropped pictures of birds, but don't calculate "effective" focal length, aperture, etc. ...
"Who on Earth do think in terms of "equivalence" when doing real world work? Why should I compare the gear I use with something that I never will buy?"
I've not seen anyone advocate for thinking in terms of equivalence when out shooting. But if someone is thinking of buying into a system or considering moving out of one, it's a useful tool to understand the envelope of capabilities offered by each system (or lens within that system).
Equivalence shows that a significant range of photos can be shot with any format and comparable image quality. People find a way to insist their camera model or brand or sensor size is better, regardless of terminology (or provable fact, a lot of the time).
I think a lot of photographers do have an intuitive understanding of the concept, but some seem to get confused when they can't see the limitations of truisms such as f/2=f/2=f/2. I suspect there'd be a lot less push-back against it if it were widely understood (because the outcome is the same whether you understand / engage with it or not).
And sure, you could make yourself an Excel spreadsheet to calculate the equivalent crops of your images, if you wanted. It'd take about five minutes. I'm not sure what you'd use it for, but it would give results that are consistent with what you'd see in the images.
I'm not sure that multiplying one number by another, when you want to understand the impact of format is a lot of maths.
@Magnar W By time I'm looking through the viewfinder I'm out taking pictures. I don't want to tell everybody on the street, "Wait here in that pose for a few minutes, I need to run back to my hotel room and pick up my other lens. This is way wider than I thought because I don't understand equivalence."
Or maybe I'm going to rent or borrow a lens and want to know what will give me roughly the same thing in crop sensor because I don't have the money to just rent all the lenses and sit there swapping them out until one looks right.
It seems like you're working really push this "just get out and shoot" mentality but real photographers think, visualize, anticipate, and plan. I don't want to be in the Decisive Moment only to find out I was incorrect in the framing I have. As many photographers advocate, I see the scene and composition before I even bring my camera up. I don't want to wait till I look through the viewfinder only to find my lens isn't as wide as I thought.
Nope, wrong again. First of all digital "medium format" isn't even medium format. Only the Phase One cameras get close to the true medium format. Secondly, photographers have already done these tests on youtube and it's 100% clear that format drastically changes the background blur.
Equivalence doesn't work in the real world and here is one reason why... While it's true that you can get closer with a smaller sensor and achieve the same BG blur, you can only get so close if your subject is a fixed size. For instance, human beings are generally 5ft - 6.5ft tall. So, you may be thinking, oh, my 4/3rds sensor will achieve the same background blur, I'll just get closer and voila, equivalence!!! Sorry, but no, that would only be true if you could shoot tiny people who were like 2-3ft tall. Then the equivalence fantasy would make sense. Now, if I take a picture of my pet hamster with a cellphone, I can get decent background blur. See, that is how equivalence works.
KingKwannon, if you can find a wide enough lens you'd get the exact same effect with smaller sensors. But there are generally limits to feasible aperture. That said, any blur created with MF is possible with aggressive FF lenses.
Yep, that's pretty much why 35mm won. The lenses Canon was making were often very sharp, but the 50mm and 85mm f/1.2's were never great wide open. The effect the big aperture had was to make the image quality losses irrelevant. Not only for the perception of sharpness, but also for the film, where the larger grain of 35mm would have no effect on resolution across most of the picture. In fact, a whole subculture developed of people who liked the extra grain of 35mm when using fast lenses.
There are many ways to skin a cat as the saying goes. Take the new Canon C70, it allows the use of a speed-booster. But that isn't how optics are typically measured. And the key thing there is that the lens is still larger. You can make the lens focus down on a tiny sensor, but the lens is still big if you want the equivalently the same performance as the bigger sensor.
KingKwannon: you completely misunderstand equivalence, it's not about getting to the subject closer, it's about what lens you need on a different format to get the same image from the same spot. So to achieve the same framing and BG blur on 4/3" sensor in the exact same situation, you need a lens with half the focal length and half the f-number, compared to FF.
Really? Then why do you say things like "While it's true that you can get closer with a smaller sensor and achieve the same BG blur, you can only get so close if your subject is a fixed size."
Even with the 80mm F1.7 lens, I always shoot at F2 to prove to everyone that there is no difference with a FF system and 1.4 / 1.6 lens. Proving it to everyone means everything, not being able to shoot at F1.7 for better OOF BG or bokeh is not important at all. I only use medium format to compare full frame equivalency.
Jokes aside, I'm tired of seeing this type of tests. I do not buy a 100mp camera to crop it down to 26mp and compare with a APS-C camera. I do not buy a MF 1.7 lens and stop it down to compare it with a FF lens. If a test needs to be done, it should be with every system at their max performance / feature / potential, not dumbing down because the other systems have smaller sensors / not a good enough lens or something else. You won't race a Toyota and Ferrari at the same speed and compare how it feels.
This was my main misunderstanding about the video. If you take away the advantages of a certain system to make it level with another, of course the results would be similar.
It's how I felt about their testing of noise in high MP versus low MP sensors.
You do realize that you need to stop down fast FF lenses to be as slow as MF, right?
The point of the test was to see if there was any *inherent difference* in the sensor sizes. There wasn't. Now you can look at what glass is available and understand why MF isn't as shallow if that's your bag.
"I do not buy a 100mp camera to crop it down to 26mp and compare with a APS-C camera" I mean, isn't that one of the advantages, that you can crop heavily and still have a decent image?
Literally, that it can be cropped so heavily and still have 26MP left.
@Eggplant: Of course it is an advantage. But this kind of tests always do that, as if everyone buys MF to just dumb down and use it like that. That is what I meant.
Agree - this is a impossible test. It would only make sense to compare the same lens/focal length on all cameras e.g. E-mount 24 mm F1.4 on FF and APS-C. Since there is no comparable system with all three sensor formats you will never see a difference in DOF. Chris tried to eliminate all differences by using math and obviousely succeeded. Compare it on E-mount only and at the same distance to the object you will definitelly see a clear difference - in a different frame and if E-mount would offer a MF sensor it would show the same.
I think Chris was pretty clear that they are testing DOF only in this test. SOME people claim that shallow DOF is an advantage of MF cameras. This video shows that a 35mm f/1.4 lens is capable of achieving the same shallow DOF as an f/2 MF lens, which is absolutely true. They were even used short edge equivalence and cropped the 35mm and APS-C images, which maximizes the effect at a .73x crop, rather than the .82x crop on the long edge. Some people considering MF would find this video useful in showing available MF lenses don't really buy DOF beyond a fast 35mm lens.
As for your example MF lens: 80mm f/1.7 is equivalent to 55mm f/1.24 (short edge) or 66mm f/1.41 (long edge) on 35mm. 50mm f/1.2 lenses are available for many 35mm systems, and 85mm f/1.4 lenses are widely available.
Another thing not covered here is shutter speed and ISO equivalence, which does favor MF, since MF effectively has one stop of lower ISO available than the FF cameras, assuming that both have the same base ISO. However, again base ISO performance is already very good in modern cameras.
The overall point is this: Modern 35mm cameras and lenses have gotten so good that while medium format does have advantages, those advantages are subtle, and only show up in specific shooting conditions. This makes modern MF cameras more of a niche tool. If 1.5x improved linear resolution and 1 stop lower base ISO matters to your workflow, then get MF. If you need more than that, you may need to be looking at larger MF sensors like Phase One.
Funny that the most liked comment misses the conclusion of the video. Which is that looking at depth of field alone there is no advantage to MF compared to FF or APS-C. Or to FF compared to APS-C, for that matter.
@xtam667 I don't think that the video shows that there's no advantage to FF/MF compared to APS-C. They had to stop down the MF lens to reach equivalence to the APS-C lenses. This is in contrast to FF, where the MF lens was used wide open.
Using short edge equivalence as they are using, the MF crop factor is 0.73 and the APS-C crop factor is 1.5. 110mm f/2 MF = 80mm f/1.46 FF = 53.5mm f/0.97 APS-C.
While there might be a few specialty f/0.95 APS-C lenses, they're certainly less common, especially with autofocus. However, 85mm f/1.4 FF lenses are available with AF in most mounts. Based on those results, there aren't many options to get the kind of shallow DOF with APS-C that you can get easily with FF/MF.
Hmmm...the equivalence discussed is incomplete. That's because the blur due to depth of field (and, if you wish to be more accurate, including diffraction and other blurs as well) must be multiplied up from its value on the sensor to its size in the final presented image (be it on a monitor or a print). Because smaller sensors require their image to be magnified more to reach a given viewing size, their blurs are also magnified more and this effect undoes much of the so-called advantage in depth of field. Computing all this is not trivial, but the net result is that any on-sensor depth of field advantage of a smaller sensor is greatly reduced if you require the final viewed image blur to be the same. And, that additional multiplication also raises other blurs: camera motion, subject motion, lens aberrations (e.g. coma, chromatic, astigmatic), diffraction, etc. Net result: not much if any advantage left.
The subject of "equivalence" will never be settled, because there's no agreement on how to resolve differences in aspect ratio. Issues of DOF and diffraction are mathematically easy because terms in the numerator and denominator cancel out. That done, for the same FOV and relative aperture (f/stop), DOF is inversely proportional to the cropping factor. Diffraction is directly proportional to the relative aperture. In this article, multiplying the relative aperture by the cropping factor levels the playing field.
Once you begin to "think" in a particular format, you use it to its best advantage without diddling with equivalency.
Ed, I'm not sure that's quite right. DoF at the sensor level depends on magnification, and that is a function of focal length and distance to subject only, at least in the classic thin-film-lens approximations: m = (F/(u-F)) and DoF = 2Nc (m+1)/(m^2) where u is dist. to subj., F is focal length, N is focal ratio, c is circle of confusion diameter, and m is the magnification from reality onto the sensor. It does not depend on aspect ratio. If you plug the m formula into the DoF formula, you get DoF = 2Nc(u/F)((u/F)-1) which depends on N, c, and the u/F ratio (rather than u and F separately). See Kingslake, "Optics in Photography." However, if the final viewing images are all to be the same then aspect ratio can affect the sensor-to-viewing magnification, and therefore will need consideration in that respect.
The blurring effect of diffraction relative to the size of the full image is exactly the same when you use equivalent apertures.
The only difference is in lens quality; a larger sensor can use a narrower relative aperture and so you'll have less lens aberration marring the image.
CarVac, the diffraction blur is 0.854N(m+1) = 0.854Nu/(u-F) microns if you use the radius of the Airy disc as the "effective size" of the blur, or 1.708Nu/(u-F) if you use the diameter. Since the effect is much stronger toward the center of the Airy disc, I tend to use 0.854. Now, if you have the same blur on the sensor (which is what you seem to be arguing), then when magnifying up to viewing size, the 16X24 APS-C sensor must be magnified by 1.5X more than the 24X36 sensor, and that will also magnify the on-sensor diffraction blur of the APS-C by 1.5X as well (0.82X replaces 1.5X when comparing the 33X44 sensor with the 24X36 sensor). So the APS-C diffraction blur is larger in the final viewing image; nobody looks at sensor-sized photos, be it on the camera's back-panel screen, in a print, or on a monitor. There was some of that in the "old days." Remember contact prints? (though they were usually viewed with a magnifying loupe, increasing the sensor-viewing magnification).
The APS-C sensor causes diffraction blur gets magnified 1.5x, but it starts out with a smaller Airy disc because the equivalent f-number is 1.5x lower.
buynoski: You are wrong, equivalence is about the effect of blur (due to shallow DOF or diffraction) on the final image, not the sensor. As CarVac says, that's why you multiply by the crop factor, as the blur disc/Airy disc is smaller on the smaller sensor due to the shorter focal length, even if the physical aperture is the same.
Popular MF lens designs, like the Zeiss Planar, begin to fall apart above f/2. Mirrorless MF simplifies design requirements, but more glass translates to a lot more mass and expense. Growing up, the price of an Hasselblad was a ticket to the moon, and still is if you look at studio versions.
I think another consideration is at which point an even smaller DOF becomes photographically irrelevant. I find this question particularly prudent when a lot of modern day equivalence discussion is predicated on the default assumption that an ever shallower DOF is desirable.
Super shallow DOF is super useful at times in the real world. Sometimes you just end up in crappy locations, or locations that have tons of people in the background and the ability to really throw the background completely out of focus can save the shoot and allow you to deliver the client usable results.
@The Mad Kiwi: I understand the benefits of shallow DOF in general. But the question is whether the difference between, say, f1.2 and f0.95 on FF is photographically relevant to strive for. There might be a point of legitimate diminishing returns. Either from a cost/dimensions/weight perspective, or where the DOF is effectively too shallow for most applications. Of course stopping down is possible, but in the above scenario you would be carrying a substantially larger, heavier and more expensive lens, without taking much advantage of the feature differentiating the lens from its slower alternative.
A lot of these considerations are very subjective tbh. Like, personally I'd probably never wish for something faster than f1.8 on GFX and the one f1.4 lens I've bought for FF I got as much for it's overall excellence vs the slower options as for the extra speed itself, but that's just me.
FL is a big factor too, if anything I tend to wish for slightly faster wides where getting shallow DoF even with a wide aperture can still be challenging, whereas with teles I actually with we had more slower/smaller options (specially >100mm on FF), yet plenty of people love their 135/1.8 & 105/1.4s so there's clearly a market for that.
Of course the DOF is same with any size of sensor when the aperture is adjusted to make it the same, so what is new?. The question is "which size sensor gives the least DOF"* and that depends solely on the lenses available for the format, not the sensor size per se. Here the winner is 135 (a.k.a. FF), as that is where the fastest lenses are, many f/1.2, even f/0.95 or some such. There are no f/1.4 lenses that I know of for mid format cameras yet, also no f/0.7 lenses for APS-C. It is geometry AND supply and demand..
* what combination gives most DOF is not nearly as interesting, and besides diffraction starts to eat into the resolution.
"There are no f/1.4 lenses that I know of for mid format cameras yet" GFX mount lenses: TTArtisan 90mm f/1.25 Lens for FUJIFILM GFX Mitakon Zhongyi Speedmaster 65mm f/1.4 Lens for Fujifilm GFX IRIX 45mm f/1.4 for Fujifilm GFX
Not sure what this was trying to prove. If I put a 35mm lens on my APS camera and set it to f/32, the 50mm lens on my FF camera won't stop down down to f/45 so larger can't get d.o.f. as deep. And if I set the 50 on FF to f/1.2 I can't open the 35 up to f/0.9 so smaller can't get as shallow. If I avoid shooting wide open to get best lens performance, then with 50 at f/4 for shallow but not paper thin d.o.f. the 35 isn't giving it's best performance. And if I want deep d.o.f the 35 is less impacted by diffraction at f/11 than the 50 is at f/16.
If one sticks to mid range apertures it is possible to get an equivalent.
But one of the attractions of small frame (disc and 110 on film, simple phone cameras) is more is roughly in focus.
@James O'Neill: "And if I want deep d.o.f the 35 is less impacted by diffraction at f/11 than the 50 is at f/16."
Diffraction depends on the aperture diameter. Since f stops are defined as focal length divided by aperture diameter diffraction is the same for equivalent values on the two systems (35/11 roughly the same as 50/16).
"And if I want deep d.o.f the 35 is less impacted by diffraction at f/11 than the 50 is at f/16. " This is not true. First, if you use the same 50mm F/16 lens on a 24MP APS-C and a 24MP FF sensors, the effect of diffraction on the image will be more noticable on the APS-C sensor because the same small area in a picture is smaller on the APS-C sensor. Second, when you use an 35mm F/11 lens on the APS-C camera, the iris hole diameter is the same than with the 50mm F/16 lens, so same result than above.
One of the epiphanies of mirrorless cameras is that you can actually use their lenses wide open. That's something Leica shooters have known for decades, and SLR/DSLR shooters never had. Now that the same lenses can be tested on the same camera, the deficiencies of SLR lenses can no longer be ignored.
Do not confuse the amount of diffraction created by the aperture diameter, with the amount of degradation on the image, which also depends on the pixel size.
The video is clearly addressing one common myth: that even at equivalent apertures, medium format has some magical improvement to shallow depth of field.
They used the widest apertures available on the Medium format system (which people claim has shallower DoF) and then used the “equivalent” apertures on the smaller systems.
We know and intrinsically understand that because the smaller formats were stopped down, we could get even shallower DoF on those smaller format.s
A minor defect of this study is that the equivalence multipliers were calculated on the basis of sensor diagonal, and the comparisons were made at 3:4 aspect ratio. The equivalence multipliers should have been calculated on the basis of image height, since image height is what's held constant in the test.
Though if you look at the framing they did match up the perspective and framing. They just cropped the smaller sensors so they were effectively even smaller to compensate as needed. So the smaller sensors were put at a “disadvantage” compared to the medium format that used it’s full sensor.
Not only is the diagonal an issue, but they also don’t know the exact focal length of any given lens. It’s rare that a 50mm is exactly 50.00mm it may be 52 or 48mm for all we know.
@JimKasson the great majority of real-life framing choices are made to conform the subject to the side-to-side width in horizontal aspect, and the top-to-bottom height in portrait aspect photography.
In which case, the appropriate crop factor between 44x33 and 36x24 is 44:36, or 0.818, not 0.727. So, 0.58 stops difference, not 1 stop. In the general real-life majority case.
@TN Args while I think you're right for the majority of use cases (except 8x10 prints) the stated methodology was to use short edge equivalence, and to crop the sides off of the wider FF and APS-C sensors. Given that methodology (the most favorable to MF,) @JimKasson gave the correct equivalence factors.
I think they further scaled their aperture equivalence to account for the slightly longer 85mm FF and 56mm APS-C lenses they were using. That gets 85mm f/1.54 FF "equivalence." MF still should have been stopped down to f/2.36, not f/2.2 when comparing with 56mm f/1.2 APS-C, which is probably why the APS-C didn't quite look equivalent.
I'd think maybe not in 5 years. I mean I don't think you would ever be able to fully simulate a true DOF with software. It has gotten good, but nothing beats bokeh like the optical characteristics of a lens itself, and that's why I think cameras will always remain the king for enthusiasts and pros obviously. The lower-end of the market may swap to phones (many have) and ditch their entry-level cameras and lenses, but I don't think phones are going to replace the enthusiast end of the market any time soon.
Agree that might not be a substitute for very specific applications, but for general, everyday use there is no question they killed at least the low-end dedicated camera segment. And if "enthusiast" means $3000 camera bodies like the Sony A7iv, the enthusiast segment will kill itself.
« Perfect » for that sort of stuff, it can be very, very elusive. Nature, the sort of stuff that comes from good glass, will have subtle qualities that will be hard to replicate.
" And if "enthusiast" means $3000 camera bodies like the Sony A7iv, the enthusiast segment will kill itself. " -Foskito
I've got the same fear but I don't think we're anywhere near there yet, and the A7 IV is $2,500 USD... Anyway, I do think the sweet spot for enthusiasts has moved upmarket for a few reasons, phones are encroaching a lot more on smaller formats and manufacturer focus has turned to more premium options but it's also easier to get well rounded bodies at price points that used to be a dead man's land of crippled models.
So whereas most enthusiasts might've reached for bodies around or under $1K in the past, I think a good chunk of them would find it easier to justify a $1-1.8K body these days that offers a far better value and feature set than bodies at those price points would've 5+ years ago, luckily lenses haven't gotten too much pricier and there's more 3rd party options than ever.
It *could* become a slippery slope with echoes of the HiFi market a few decades ago, but I think a few other things played into their demise and hopefully camera manufacturers have a better plan.
Of all the computational tricks smartphones have i still find the bokeh added in post to be the least convincing, like even stacking/HDR usually has less obvious errors than the fake bokeh, but that's only obvious if you're looking for it and zooming in or viewing large. I do think the errors can have an unconscious effect on how images are looked at even when people aren't aware of them, but anyway...
For the average buyer phones already took over and the vast majority of people out there will throw a grand at a new phone without ever considering a dedicated camera. I think that's precisely where camera makers can still make some gains tho.
One more thing that might or might not hold back phones from complete dominance is the FLs involved...
The normal/tele module on most phones is usually not as ggood or as fast as the main wide one, I think the general public might actually be getting used to the look of wides shot closely (and some phones just crop from it in portrait mode) but that's a whole other story... Again, something people may not be consciously aware of but which will continue to differentiate phones from ILCs even with perfect computational bokeh.
Foskito: LoL! Not Mr Bean so much (and not a wannabee either), but Blackadder... yes. You should watch it some time, particularly the final series (Goes Forth I think it's called).
It's both. The size of the circle of confusion is determined by focus error, aperture (a) and square of focal length (f). It has no idea how big the image is. You can calculate the size for a point at infinity and a lens focused at distance D the circle size (c) is f^2 / Da
The noticeable circle size is roughly image-width / 1000 so we use 0.03mm for 35mm , 0.02mm for APS-C. To get the focus distance to make a point at infinity into a circle that size just needs the equation to be swapped round: D = f^2/ca. The ratio of that (hyperfocal) distance, to where you've focused sets the d.o.f zone - e.g. focused at D/10 the zone runs from D/9 to D/11 If you double the image size keeping the same angle of view, you double F and C but because the formula is f^2 / ca for the same aperture you double the hyperfocal distance (or halve the the d.o.f).
Well I thought this show, " Chef!", was pretty funny it we are now talking about comedy aka overseas.
"Chef! is a British situation comedy starring Lenny Henry that aired as twenty episodes over three series from 1993 to 1996 on the BBC."
Concerning presenters, I prefer the Gordon Lang/ Christopher Frost style, more direct. That's my preference concerning a way to get info about camera gear.
When I usually want comedy, I watch real comedians by trade. So I never felt Chris/Jordan were adding to what they do by adding an comedy act, but some do like it, everyone's different. Other than that, they do an wonderful job.
BackToNature1: Yeah, I recall Chef! Good fun. I don't believe in somewhat serious matter ("Hey folks, here's a new camera!") presented in a faux jokey, funny way. But that's just me I guess, across the Great Divide...
I would say the same like you, it's a lens thing, but the two comedians just said it's NOT a format thing. Actually they said you get lessk depth of field in FF because of the faster FF lenses available but they said that very quiet
Say I want to demonstrate that there is nothing magical about Porche's acceleration and that of a Prius, and I drive them both in an "equivalent" way. See - the Porche is as slow as the Prius because I chose to drive it slow. The real answer to your question should include available fast lenses for the formats considered, and what they offer.
You are not testing DOF, you are testing BG blur.
Is that really a myth?
You pretty much confirm the "myth" with the Fuji 56/1.2 but then dismiss it because it does not prove what you want to prove. The Fuji shot has considerably weaker BG blur, by about a stop.
You need to look across the whole frame. Lenses wide open have much weaker BG blur away from the center. A larger format lens stopped down, to match say 56/1.2 on crop would have very similar blur in the center but much stronger one near the border. Cropping out a part of the frame proves nothing.
Now, I know the video has a unifying message in its closing words. But give it a bit of time and inevitably this comment section will be full with users arguing over why their preferred sensor size is "the best".
A 44x33mm Sensor is 70% bigger than a 36x24mm Sensor, all right? Then the Pixelpitch for the same MP count onto the bigger sensor is by laws of physics therefore bigger, than onto the smaller, crammed 36x24mm small picture format Sensor, all right?
The rendering of textures, shapes & formes, but especially the tonality, is differrent with (even small) Medium Format, than just Small Picture Format (Kleinbildfirmat) 36x24mm Sensor - if you guys don't believe it, see this:
And what's more - the GFX-100S does have 16-bit RAW, whileas the 36x24mm Sensors only all through the bank provide 14-bit RAW recording only, not the same kind of 3D Pop, tonality & rendition.
And now all freak-a-delics could bash, i stand my ground and don't give a damn.
Same kind of mis-information like PetaPixel posted (and some readers hier noticed & corrected) about the Ricoh/Pentax change, but only for Japan. And many folks posting the end of Pentax, go figure.
...and i'd like to add, that even faster MF Lenses do exist, at least one F1.7 for GFX, and from 3rd Parties even faster lens speed solutions.
For small webcontent presentation - or the usual bunch of un-social networks, therefore the picture size is resized for small viewing sizes, it does look the same, but not as a whole.
Most peeps are only looking for their pictures on screen, and never print it out - DIN-A3, A2, or even bigger print sizes format.
FOS: You just posted into between my 2 posts, i wasn't finished. :D
It's a reason that for archival purpose, very high-detailed shots, and even some kind of fashion photography, vogue, medium format is being used, instead of small picture format (ditch this pseudo "fullframe" moniker, some idiot created into the 2000's - every Sensor size is fullframe, if not being cropped)
marc petzold Thanks! In deed, let's kill off the term 'full frame' for what was once know as a 'miniature' format. The difference between 24x36 and even digital MF is obvious when printed. Many of the authorities on this subject are drawing their conclusions from miserable, online, compressed, RGB .jpg images. They need to spend more time making and looking at prints.
The DPReview video is only addressing DoF… not bit depth, diffraction, resolution, sharpness, or any of those other advantages of Medium format… they make it very clear and state it multiple times. The conclusions in the video you post only talks about the (“minor” according to the photographer) improvements in tonality, dynamic range, and recovery… nothing that the DPReview video was talking about.
Also is the Fuji GFX-100s Actually truest 16-bit/channel?
Yes, it is. The 50 MP Sensor Hasselblad XD1c 50 (II) claim to use 16-bit RAW, but the Sensor is providing 14-bit RAW only, this has being debunked long ago. GFX-100S is 16-bit RAW.
If you check the Leica SL2 vs Hassy named above from my previous post above - you clearly see, that even Leicas best L-Mount Lenses with the 36x24mm Sensor Combo gives less IQ vs. the Hassy with the now "very old" 50 MP Medium Format 44x33mm Sensor - in terms of rendition, tonality, -details.
This into contrast, that many ppl here said before "such an old Sensor" and like DPR often into the past, 50 MP MF is questionable vs 45-47 MP small picture format - nope, it's not. And even ISO 64 from the D800(e), D810 & D850 isn't rescuing it.
This was also being shown via YT videos years ago 1:1 via samples on YouTube. And Leica has some of the worlds best Lenses IQ-wise for the L-Mount, it doesn't cut it against the Hassy. Fact. With the "old" 50 MP Sensor. Whileas the 47 MP SL2 Sensor is quite new, in contrast.
44x33mm Sensor is 70% bigger than a 36x24mm Sensor, all right? Then the Pixelpitch for the same MP count onto the bigger sensor is by laws of physics therefore bigger, than onto the smaller, crammed 36x24mm small picture format Sensor, all right?
Sure. Would you like to provide examples/evidence to show the magnitude of this difference? (Note that the GFX 100 sensor out-performs the one in the GFX 50, despite having smaller pixels. Exactly the same sized pixels as the a7R IV and X-T4, as it happens)
The rendering of textures, shapes & forms, but especially the tonality, is different with (even small) Medium Format, than 36x24mm Sensor - if you guys don't believe it, see this:
That video gives the GFX 70% (2/3EV) more light and shows that this offers an tonality, but then shoots the Nikon 2/3EV above base ISO, preventing it gaining a similar benefit. If you give one camera more light, it'll perform better, that's to be expected.
And what's more - the GFX-100S does have 16-bit RAW, whileas the 36x24mm Sensors only all through the bank provide 14-bit RAW recording only, not the same kind of 3D Pop, tonality & rendition.
Can you, with specific reference to the interaction between linear data, gamma encoding and photon shot noise magnitude, detail where the additional tonality of 16-bit vs 14-bit can be seen? And also what '3D Pop' and 'rendition' mean, in terms of measurable properties?
The main benefit of a move to 16-bit is an increased capacity for DR (deep shadow detail) when mated to a suitably high DR sensor. There's barely any difference between the GFX 14 and 16-bit mode.
There are benefits to the GFX system. The 100MP sensor in particular performs exactly as you'd expect a 70% larger version of the a7R IV sensor to, and has the lenses to give that full benefit. It has a 2/3EV SNR / tonal quality benefit over FF cameras with base ISO of 100 and a resolution gain over the ones with ISO 64.
A larger sensor area gives you scope for greater light capture (which can give better tonal quality/SNR) and allow greater resolution (less demand on lenses). Just as FF offers those same benefits over APS-C and Four Thirds does over 1"-type and so on...
But stating that 70% extra sensor area gives you different rendering of 'sahpe and forms,' or invoking undefined qualitative assessments of '3D Pop' and 'rendition' looks like magical thinking to justify an existing belief. As does ascribing these properties to the presence of 16-bit ADCs, unless you can detail a way in which they would (and it's a LOT more complex than 'you capture with more precision' because 14-bit is already oversampling a lot of the tonal range because the noise is so great).
We like a lot of the GFX cameras but not because of any inherent magic.
Richard, when you need to write four (!) answers to my post - have you seen the Video from the SL2 vs the Hassy HD 50c II above?
I know you guys are all Small picture format lovers here - but for it's special purpose, it simply doesn't cut it, despite the comments. The Hassy looks really better here - and the Leica SL does have eye watering expensive lenses onto the front.
I am not talking about DoF 1:1 comparsions at equivalent lens aperture - i am talking about the possible IQ only, and in terms of rendition, tonality hereby.
The smaller Sensor looses. If you'd set the same (not equivalent - for DoF) aperture, shutterspeed - a 44x33mm Sensor due to it's 70% bigger, physical surface - captures simply more photons, than into the exact given same time, same settings - only 36x24mm Sensor into contrast.
Well Richard, you've seen enough Medium Format Pictures in your life all of the web, haven't you? Just see the various reviews, of the past ~10 years (into digital terms) and...
..check that -rendition, -tonality vs. ordinary Small picture format. It's not the same. If i could finance it, i'd go MF for my hobby, but only for the most interesting subjects.
16-bit RAW files give just a hell lot headroom for heavy editing, more than 14-bit, in that matter. Into a world, whereas gearfreaks and techheads are always lusting for neverending more - more MP, better AF, better lowlight performance, tack-sharp pictures at open aperture into all corners, and all other stuff - wouldn't it being good, to have that (almost) extra 2-bits as reserve, for editing?
If i'd hypertheoretically work for a living shooting various things, i'd only shoot large format Film 8x10, and digital medium format, ditching all the smaller things.
It took four answers because I includes your quotes, which took away form the character limits.
Just looked at that Leica vs Hasselblad video and he gives the Hasselblad 1 1/3EV more light per image (it'd have 2/3EV more light at the same exposure settings, but the aperture has been opened-up 2/3 on the medium format, giving much shallowed depth-of-field). It's so odd that he sets out what would be equivalent and then comes up with some rather questionable reasoning for doing exactly the opposite.
Again, give one camera much more light than the other, and it'll look better.
I've not just seen a fair number of medium format images, I've shot and processed them too. I've seen the 2/3EV improvement in SNR/tonal quality, I've seen the resolution benefits (including those of the lenses, though the GFX 50's intentionally separated microlenses can exaggerate this by giving improved apparent sharpness as a trade-off for a greater risk of moire).
But I've edited 14 and 16-bit files from the GFX 100 and I can't agree with your characterisation that there's a lot of additional headroom.
My understanding of the full impact of bit-depth is incomplete but I know enough to know that greater bit depth doesn't mean better tonality (or even more editing flexibility) in an of itself. I know camera makers' marketing departments like to say that it means your files have 'more colours' or 'more tones' but that's not really how it works.
I'm not a 'small picture format lover.' The nature of my job is to try to understand the strengths and weakness that different formats bring, and how physics plays out across all formats. All formats are a compromise (broadly speaking between size, price and image quality). There is no best format, each person and each use-case is best suited by different compromises/balances/formats.
But magical thinking that one format is special, and blanket dismissal that another 'just can't cut it,' doesn't help.
I think you've mistaken the word 'film' for the word 'true.' Silicon isn't film so (other than backward compatibility with existing lens systems) there's no reason to make sensors in the same sizes as film. Modern sensors are more efficient than film: a smaller digital format can perform comparably to a larger film one.
There are medium format models that use a ~54 x 40mm sensors (7% smaller than 645 film) from Phase One, Leaf and Hasselblad. Sony Semiconductor offers a 53.4 x 40mm, 151MP chip based on the same-sized pixels used in the GFX 100, the a7R IV and the X-T4.
Which leaves this digital 645 and the smaller 44x33 as what 'medium format' means in digital terms. But since most individuals are only likely to afford 44x33 it makes sense to compare this to the other option they might consider: 36x24. It wouldn't make sense to go down to a format they won't consider just because '44x33 is to APS-C as 645 was to 135 in the film era.'
For those obsessed with ultra-shallow DoF, full-frame with fast f1, 1.25, etc primes is the way to go. But, medium format look” is a combination of factors besides depth of field; color depth, tonal gradations, noise etc have a role in the final image too.
The SNR in midtones and highlights is where medium format sensors pull away, especially the larger variety. This makes a noticeable difference in the creaminess of skies.
The smaller sort of medium format is not much better than the D850 with its 64 base ISO though.
Back in the film days, there was a huge difference in detail, sharpness etc. between 35mm film and 6x6cm medium format....due to the properties of the film itself of course. Also probably in those days, the average quality of a MF lens was probably better than the average quality of a 35mm format lens (with equivalent difference in cost!!).
some of the 35mm format lenses i've tested on 42MP MILC (Canon FD, Minolta MD) perform very very well in terms of detail, even by todays standards. You have to stop them down and avoid flares tho.
On lens quality, it's actually the opposite. In pure resolution terms, high quality lenses for 35mm tended to be better than lenses for medium format film cameras, overall.
That's because they had to be; any medium format film image will be enlarged much less to make a given print, so there wasn't as much pressure on lens quality (or film flatness and several other technical matters).
The rub is "high quality" most 35mm lenses of old would probably not be in that category. Also, film resolution is so lousy compared to today's high res sensors that the lenses of old really didn't have to be that good....film resolution was the limiting factor.
Would've been interesting to show the shallowest DoF possible on each system, for instance comparing the look of a FF 85mm f1.2 to the Fuji 110 f2 and show how much of a difference f1.2 vs f1.6 equivalent looks. It's my impression that for equivalent apertures, FF has the most low-aperture options, but correct me if I'm wrong.
You are right, f1.0, f1.25, even f1.4 full frame lenses will give you shallow DoF than the current mini “medium format” cameras. 6x7, 6x9 film is another story.
You are right for GENERAL lenses used by most people but there ARE some MF (and large format) lenses that are as fast and faster. For instance some aerial photography lenses like 150mm f0.75. I guess the market for such lenses would be tiny and they would be huge and VERY expensive
In the analog age, medium format Hasselblad for instance was 6x6 cm which is about 4x the surface size of 24X36 mm. In the digital age medium format has become smaller.
Show me any 6x6, 6x7, 6x9 lens than is faster than 35mm equivalent. Or even 4x5, 8x10 sheets. As the format grows larger the lenses go much slower too, in the end it cancels out. Yeah, there are special projector, x-ray lenses or aerial ones but when you look at the images they don't have great quality wide open, the contrast is very poor.
I did a weight check between equivalent Nikon 1, m4/3, APS-C and FF lenses some years ago, and though different manufacturers threat the importance of low weight different with different design philosophies, I found that they followed each other quite well. Nikon 1 was actually worst, I guess aperture & AF motors, electronics etc in the lenses have a starting weight that doesn't matter of the imaging circle the lens is built for and the parts that scales with the image circle size is on a flatter weight curve.
But what does come with a larger image circle is a larger sensor. More possible pixels at the same or lower pixel pitch and in reality more possible bits per pixel it seems. Higher DR, more colors, better gradients. Same noise level at equivalent apertures but more pixels are processing better in modern noise reduction software like DXO Photolabs, so in reality one can go higher.
So that was the theory. But in practice, you need lenses that actually exist in the real world and in FF they go down to f/1.2 for extremely shallow depth of field. Are there any f/0.8 (*1.5) APS-C lenses? Any f/1.5 (*0.8) medium format lenses? It seems to me that the most extreme look is actually found in full frame, despite being neither biggest nor smallest.
There is a 40mm F0.85 for APS-C mirrorless cameras. You can also use smaller formats with focal reducers and most SLR lenses (like say the Canon EF 50mm F1.0L USM.) Focal reducers and teleconverters are fairly effective format converters. They even make the sensor appear to be a different size when you look through them.
Well, actually Nikon has a FF Noct 58mm f/0.95 for their Z system, so f/0.95 seems to be the new benchmark for FF in terms of fastest lens for FF that I can think of.
But then again, that lens was more for show by Nikon, not necessarily a practical lens for most people (due to its cost mostly).
I think the number of reasonably accessible lenses is probably more relevant than who's got the most extreme. F1.4 FF lenses are pretty common and within grasp of most enthusiasts, not to mention they're often relatively compact.
I think for most people 1.2 is probably the fastest they would probably want to go. I mean I personally (myself) don't see much of a need to buy anything faster than 1.8 as I wouldn't likely notice much of a difference, although this could matter depending on the focal length of the lens. But in general, for my needs at least 1.8 offers shallow enough DOF that I cannot really justify 1.2 and most 1.4 lenses. So I guess the photographer has to weigh out how important that little extra shallow DOF is to them and if they are willing to pay top dollar for it.
(Obviously the Noct I mentioned above is probably more for specific tasks and not your average portrait lens for example.)
Lomography's LomoChrome '92 is designed to mimic the look of classic drugstore film that used to fill family photo albums. As we discovered, to shoot with it is to embrace the unexpected, from strange color shifts to odd textures and oversized grain.
The LowePro PhotoSport Outdoor is a camera pack for photographers who also need a well-designed daypack for hiking and other outdoor use. If that sounds like you, the PhotoSport Outdoor may be a great choice, but as with any hybrid product, there are a few tradeoffs.
The Sony a7C II refreshes the compact full-frame with a 33MP sensor, the addition of a front control dial, a dedicated 'AI' processor, 10-bit 4K/60p video and more. It's a definite improvement, but it helps if you value its compact form.
Why is the Peak Design Everyday Backpack so widely used? A snazzy design? Exceptional utility? A combination of both? After testing one, it's clear why this bag deserves every accolade it's received.
The new Wacom One 12 pen display, now in its second generation, offers photographers an affordable option to the mouse or trackpad, making processing images easy and efficient by editing directly on the screen.
If you want a compact camera that produces great quality photos without the hassle of changing lenses, there are plenty of choices available for every budget. Read on to find out which portable enthusiast compacts are our favorites.
What's the best camera for travel? Good travel cameras should be small, versatile, and offer good image quality. In this buying guide we've rounded-up several great cameras for travel and recommended the best.
What’s the best camera for around $2000? This price point gives you access to some of the most all-round capable cameras available. Excellent image quality, powerful autofocus and great looking video are the least you can expect. We've picked the models that really stand out.
Above $2500 cameras tend to become increasingly specialized, making it difficult to select a 'best' option. We case our eye over the options costing more than $2500 but less than $4000, to find the best all-rounder.
Lomography's LomoChrome '92 is designed to mimic the look of classic drugstore film that used to fill family photo albums. As we discovered, to shoot with it is to embrace the unexpected, from strange color shifts to odd textures and oversized grain.
Sony's gridline update adds up to four customizable grids to which users can add color codes and apply transparency masks. It also raises questions about the future of cameras and what it means for feature updates.
At last, people who don’t want to pay a premium for Apple’s Pro models can capture high-resolution 24MP and 48MP photos using the iPhone 15 and iPhone 15 Plus. Is the lack of a dedicated telephoto lens or the ability to capture Raw images worth the savings for photographers?
Kodak's Super 8 Camera is a hybrid of old and new: it shoots movies using Super 8 motion picture film but incorporates digital elements like a flip-out LCD screen and audio capture. Eight years after we first saw the camera at CES 2016, Kodak is finally bringing it to market.
In this supplement to his recently completed 10-part series on landscape photography, photographer Erez Marom explores how the compositional skills developed for capturing landscapes can be extended to other areas of photography.
If you want a compact camera that produces great quality photos without the hassle of changing lenses, there are plenty of choices available for every budget. Read on to find out which portable enthusiast compacts are our favorites.
Sony, the Associated Press and 'Photo Mechanic' maker Camera Bits have run a month-long field-test to evaluate capture authentication and a subsequent workflow.
A color-accurate monitor is an essential piece of the digital creator's toolkit. In this guide, we'll go over everything you need to know about how color calibration actually works so you can understand the process and improve your workflow.
What's the best camera for travel? Good travel cameras should be small, versatile, and offer good image quality. In this buying guide we've rounded-up several great cameras for travel and recommended the best.
It's that time of year again: When people get up way too early to rush out to big box stores and climb over each other to buy $99 TVs. We've saved you the trip, highlighting the best photo-related deals that can be ordered from the comfort of your own home.
The LowePro PhotoSport Outdoor is a camera pack for photographers who also need a well-designed daypack for hiking and other outdoor use. If that sounds like you, the PhotoSport Outdoor may be a great choice, but as with any hybrid product, there are a few tradeoffs.
Sigma's latest 70-200mm F2.8 offering promises to blend solid build, reasonably light weight and impressive image quality into a relatively affordable package. See how it stacks up in our initial impressions.
The Sony a9 III is heralded as a revolutionary camera, but is all the hype warranted? DPReview's Richard Butler and Dale Baskin break down what's actually new and worth paying attention to.
What’s the best camera for around $2000? This price point gives you access to some of the most all-round capable cameras available. Excellent image quality, powerful autofocus and great looking video are the least you can expect. We've picked the models that really stand out.
DJI's Air 3 and Mini 4 Pro are two of the most popular drones on the market, but there are important differences between the two. In this article, we'll help figure out which of these two popular drones is right for you.
The Sony a7C II refreshes the compact full-frame with a 33MP sensor, the addition of a front control dial, a dedicated 'AI' processor, 10-bit 4K/60p video and more. It's a definite improvement, but it helps if you value its compact form.
Above $2500 cameras tend to become increasingly specialized, making it difficult to select a 'best' option. We case our eye over the options costing more than $2500 but less than $4000, to find the best all-rounder.
The iPhone 15 Pro allows users to capture 48MP photos in HEIF or JPEG format in addition to Raw files, while new lens coatings claim to cut down lens flare. How do the cameras in Apple's latest flagship look in everyday circumstances? Check out our gallery to find out.
Global shutters, that can read all their pixels at exactly the same moment have been the valued by videographers for some time, but this approach has benefits for photographers, too.
We had an opportunity to shoot a pre-production a9 III camera with global shutter following Sony's announcement this week. This gallery includes images captured with the new 300mm F2.8 GM OSS telephoto lens and some high-speed flash photos.
The Sony a9 III is a ground-breaking full-frame mirrorless camera that brings global shutter to deliver unforeseen high-speed capture, flash sync and capabilities not seen before. We delve a little further into the a9III to find out what makes it tick.
The "Big Four" Fashion Weeks – New York, London, Milan and Paris - have wrapped for 2023 but it's never too early to start planning for next season. If shooting Fashion Week is on your bucket list, read on. We'll tell you what opportunities are available for photographers and provide some tips to get you started.
Sony has announced the a9 III: the first full-frame camera to use a global shutter sensor. This gives it the ability to shoot at up to 120 fps with flash sync up to 1/80,000 sec and zero rolling shutter.
What’s the best camera for around $1500? These midrange cameras should have capable autofocus systems, lots of direct controls and the latest sensors offering great image quality. We recommend our favorite options.
Comments