Math Problem: If f0.95 = f1.9 in full frame...

[siminona]

I hope some of you can help me. I've been pondering... If on the Four Thirds sensor f0.95 is roughly equivalent to f1.9 with regards to DOF on full frame sensors, what is f0.95 equivalent to on APS-C crop sensors? Would I multiply it by 1.6 to get f1.52?

If Four Thirds is supposedly at a 2-stop disadvantage when it comes to DOF (not light-gathering) compared to full frame, how many stops is APS-C at a disadvantage? And what do you think of this article: http://www.have-camera-will-travel.com/field_reports/full_frame_vs_crop_sensor_-.html

I hope this doesn't start a debate or anything. Admittedly, this stuff isn't all too important to me. I just like taking photographs, and I'm very happy with m4/3. I'm just trying to figure out the math.

Thanks!

 

[seachicken2000]

I think the numbers are roughly that APS-C is 2/3 of a stop better than m4/3 and FF is 1 1/3 stops better than APS-C.

From what I've gathered from these forums, it's not that there's any advantage of any of the formats in theory. It's just that the f-numbers are not directly comparable, and that in reality lenses equivalent to the really fast FF lenses can't be easily made for the small sensors.

--
A rose by any other name is still a chicken.

 

[Bogdan_M]

I think the numbers are roughly that APS-C is 2/3 of a stop better than m4/3 and FF is 1 1/3 stops better than APS-C.

Roughly that, given that APS-C is between 1.5 and 1.7 crop.

From what I've gathered from these forums, it's not that there's any advantage of any of the formats in theory. It's just that the f-numbers are not directly comparable, and that in reality lenses equivalent to the really fast FF lenses can't be easily made for the small sensors.

That is not true. The bigger formats have 2 advantages. They can give you less DOF when you want maxiumum separation for a portrait. Thay also give you the flexibility to trade DOF for low light performance. A FF ISO 3200 should be theoretically equivalent in noise to 800 ISO on 4/3.

Thta being said the differences with APS-C aren't significant, because the difference of size between sensors is not that big. The availability of certain lenses and technological developments tend to mask differences between standards. For instance what you can achieve with that 0.95 will also be achieved in Canon scope with a Canon 35mm f 1.4, but without any advantage is size, or price.

--
A rose by any other name is still a chicken.

--
http://www.flickr.com/photos/bogdanmoisuc/

 

[seachicken2000]

That is not true. The bigger formats have 2 advantages. They can give you less DOF when you want maxiumum separation for a portrait.

There was a thread on here earlier about this. As I understand it, m4/3 can give you the same DOF as FF, in theory .

For example a 50mm f/2 FF lens is equivalent to a 25mm f/1 m4/3 lens in terms of:

• angle of view (most will accept this)
• depth of field (easily checked http://www.dofmaster.com/dofjs.html )
• total amount of light gathered (From wikipedia: "If two cameras of different format sizes and focal lengths have the same angle of view, and the same aperture area, they gather the same amount of light from the scene." - the AoV is the same, and both have 25mm diameter apertures)

They also give you the flexibility to trade DOF for low light performance. A FF ISO 3200 should be theoretically equivalent in noise to 800 ISO on 4/3.

The total amount of light gathered, determines the noise. So I don't see any advantage.

In the above example, the sensors in the cameras will be ISO 3200 for FF and ISO 800 for the m4/3. But the m4/3 camera is concentrating the same total amount of light down onto an area 1/4 the size. The sensor is 2 stops less sensitive, but receives 2 stops more light.

So in theory, you provide me a FF camera and lens, and I can design for you a m4/3 camera and lens that has the same photographic characteristics: angle of view, depth of field, and noise. The markings on the lenses for focal length and aperture will differ as will the ISO markings, but they'll produce images that you can't tell apart.

The issue is that in the real world you can't build the fast lenses for m4/3, or even if you can they'd be huge and expensive. The 25mm f/1 lens does exist in the form of the Voightlander 25mm f/0.95 lens, but ... it's huge and expensive. If you wanted a lens with equivalent characteristics to a 50mm f/1 on FF, it's not available.

I wonder why these lenses can't be made for m4/3. I suspect it's because they need to be telecentric. If sensors could be designed to accept light from a large range of angles like film, I wonder if these fast lenses could be made for m4/3.

--
A rose by any other name is still a chicken.

 

[KenBalbari]

Well, the way you are stating it isn't entirely accurate. If you change only the f-stop, and the format, no the depth of field is not going to be the same. But if you change the focal length as well, in order to get the same framing, that's when you have to change the f-stop to get the same depth of field.

Really, what it comes down to is the same aperture size will produce the same depth of field. So 50/0.95 = 52.63 mm aperture, and 100/1.9 = 52.63 mm aperture. So those will both produce the same depth of field. If you shot them both on the same camera, and then cropped the larger (50mm) image down to the same field of view, they would also look the same then.

This really isn't as complicated as some like to make it. You can view it as, given a 2x crop factor, the larger format requires 2x the focal length, 2x the f-stop, and 2x the ISO (to keep the shutter speed the same) to get pretty much the same photo. Or you can view it as the smaller format taking the same image with the same lens, and cropping it. Anyone who thinks those two views are in conflict, doesn't really fully understand either.

The primary advantage of the larger format is generally going to be more resolution. In the example above, where you change focal length, f-stop, and ISO, the only difference left is that the larger image likely produces a higher resolution image (in lines per picture height for example). In the crop example, you might get roughly the same resolution (assuming comparable sensor technology), but that is only because you are wasting a large part of the sensor (why not use a longer focal length and get a higher resolution instead?)

The other wrinkle here is that the reason depth of field becomes an issue has to do with limitations in lens design. If it were as easy to find a 25 f/1 for 4/3 as it is to find an equivalent quality 50 f/2 for full frame (35mm), then there really wouldn't be any advantage for the larger sensor in achieving narrow depth of field at that effective focal length.

 

[Bogdan_M]

That is not true. The bigger formats have 2 advantages. They can give you less DOF when you want maxiumum separation for a portrait.

There was a thread on here earlier about this. As I understand it, m4/3 can give you the same DOF as FF, in theory .

For example a 50mm f/2 FF lens is equivalent to a 25mm f/1 m4/3 lens in terms of:

• angle of view (most will accept this)
• depth of field (easily checked http://www.dofmaster.com/dofjs.html )
• total amount of light gathered (From wikipedia: "If two cameras of different format sizes and focal lengths have the same angle of view, and the same aperture area, they gather the same amount of light from the scene." - the AoV is the same, and both have 25mm diameter apertures)

They also give you the flexibility to trade DOF for low light performance. A FF ISO 3200 should be theoretically equivalent in noise to 800 ISO on 4/3.

The total amount of light gathered, determines the noise. So I don't see any advantage.

In the above example, the sensors in the cameras will be ISO 3200 for FF and ISO 800 for the m4/3. But the m4/3 camera is concentrating the same total amount of light down onto an area 1/4 the size. The sensor is 2 stops less sensitive, but receives 2 stops more light.

So in theory, you provide me a FF camera and lens, and I can design for you a m4/3 camera and lens that has the same photographic characteristics: angle of view, depth of field, and noise. The markings on the lenses for focal length and aperture will differ as will the ISO markings, but they'll produce images that you can't tell apart.

The issue is that in the real world you can't build the fast lenses for m4/3, or even if you can they'd be huge and expensive. The 25mm f/1 lens does exist in the form of the Voightlander 25mm f/0.95 lens, but ... it's huge and expensive. If you wanted a lens with equivalent characteristics to a 50mm f/1 on FF, it's not available.

I wonder why these lenses can't be made for m4/3. I suspect it's because they need to be telecentric. If sensors could be designed to accept light from a large range of angles like film, I wonder if these fast lenses could be made for m4/3.

--
A rose by any other name is still a chicken.

Yeah, you've got it right. Basically, you can have perfectly equivalent pictures (noise, perspective and DOF-wise). Just that there are some situations that are not obtainable on the smaller formats.

Why is that? You can search the forum for Joseph's Wisnievski's FF wizfaq, he expalins the matter very well. Basically, FF (and not APS-C) falls in a physycal sweet spot, where fast lenses are possible without too much weight and cost. It's just physics and luck

--
http://www.flickr.com/photos/bogdanmoisuc/

 

[misolo]

I hope some of you can help me. I've been pondering... If on the Four Thirds sensor f0.95 is roughly equivalent to f1.9 with regards to DOF on full frame sensors, what is f0.95 equivalent to on APS-C crop sensors? Would I multiply it by 1.6 to get f1.52?

If Four Thirds is supposedly at a 2-stop disadvantage when it comes to DOF (not light-gathering) compared to full frame, how many stops is APS-C at a disadvantage? And what do you think of this article: http://www.have-camera-will-travel.com/field_reports/full_frame_vs_crop_sensor_-.html

I hope this doesn't start a debate or anything. Admittedly, this stuff isn't all too important to me. I just like taking photographs, and I'm very happy with m4/3. I'm just trying to figure out the math.

Thanks!

To go from 4/3rds to FF: multiply both the focal length and the f-number by 2 and ISO by 4. This will give you setups which will produce two photos with very similar characteristics (approximately same field of view, depth of field, noise, and dynamic range - although the equivalence in noise and dynamic range only applies for two sensor of the same technology and generation).

To go from Canon APS-C to FF: multiply focal length and f-number by 1.6 and ISO by 2.56.

To go from 4/3rds to APS-C: multiply focal length and f-number by 2/1.6=1.25 and ISO by 1.25^2=1.56.

Of course, the equivalence is not exact. The aspect ratios are different, more recent sensors have less noise, etc.

 

[duartix]

I guess you deserve a prize for a thread well started.

It's the first time I see one of these threads not degenerating in a FOV, DOF, ISO flame throw show! :O
Congratulations on the posters too for keeping it very clear!
--
Duarte Bruno

 

[Everdog]

First which APS? Nikon's or Canon's? They are different sizes.
Second, which 4/3rds? Most are the same, but the GH1 and GH2 can be different.

Third, we are taking about different aspect ratios so that messes everything up. A Canon APS sensor is less that 14% taller than 4/3rds sensor, but much more wide.

The reality is you will see a difference between FF and 4/3rds for DOF, and between FF and APS too. It is very hard to see any difference between APS and 4/3rds because there are so many other variables (different lenses, etc).

My advice: don't worry about it and go take some pictures.

 

[Ehrik]

I hope some of you can help me. I've been pondering... If on the Four Thirds sensor f0.95 is roughly equivalent to f1.9 with regards to DOF on full frame sensors, what is f0.95 equivalent to on APS-C crop sensors? Would I multiply it by 1.6 to get f1.52?

For Canon 1.62 crop, you'd multiply with 2.00/1.62 = 1.23.
For NX/NEX 1.54 crop, you'd multiply with 2.00/1.54 = 1.30.

If you crop the final image to something closer to 4:3, then the coefficients go down a bit, closer to 3:2 and they go up a bit, unless you use the multi-aspect capability of the GH cameras.

One important point to remember with equivalency is that ISO 100 on 4/3" is equivalent to ISO 400 on FF, so the 4/3" ISOs are "allowed" to be more noisy.

When talking about extreme lenses like f/0.95 there's the question what the f-number of the microlenses is. That may be the limiting factor and all the light from f/0.95 may not be available, not even in the centre of the frame.
--
Just my two öre,
Erik from Sweden

 

[TorsteinH]

I think the numbers are roughly that APS-C is 2/3 of a stop better than m4/3 and FF is 1 1/3 stops better than APS-C.

Shoting with large f/stops you normally want all the DOF you can get! So I would say 4/3 is aprox 2/3 better than APS-C and two stops better than 35mm (FF is not a term I would use! If you photograph with a camera with a 45x60mm sensor, are you using Double FF?)

From what I've gathered from these forums, it's not that there's any advantage of any of the formats in theory. It's just that the f-numbers are not directly comparable, and that in reality lenses equivalent to the really fast FF lenses can't be easily made for the small sensors.

--
A rose by any other name is still a chicken.

--
http://www.ohb.no/foto
************
Torstein

 

[Jonas B]

Shoting with large f/stops you normally want all the DOF you can get! So I would say 4/3 is aprox 2/3 better than APS-C and two stops better than 35mm (FF is not a term I would use! If you photograph with a camera with a 45x60mm sensor, are you using Double FF?)

Torstein,

Certainly the term FF doesn't make you unsure what we are talking about?

There is no DOF advantage to 4/3 - just stop the FF camera down and you get the same if needed. However, in some cases you can equal the playing field. When you need a certain shutter speed and a certain DOF and the FF camera has to be at 4 times the ISO value because of this - then we get the same result as the FF user.

In all other cases the FF camera either gives more DOF control and/or less noise and other possibilities.

The best system is the system suiting to the individual's needs and wallet. I have after having used the 5D, the G1, the 5DMkII and now the G1 again settled for micro 4/3. It's a good system for me, in some cases it matches the FF system, in all other cases it is technically less stellar.

Jonas

 

[CWilt]

Seems like everyone wants to look at it as the glass is half empty because its harder to get less in focus. It is much easier to get everything in focus with m4/3 than it is with FF. For a set amount of light and shutter speed FF will need a 2 stop smaller aperture to achieve the same DOF. That often requires a 2 stop higher film speed.

Maybe the glass is half full instead.
--
Charles
My family images are at http://www.stakeman.smugmug.com
Be sure of your subject.
Never, force the shot.

 

[KenBalbari]

To go from 4/3rds to FF: multiply both the focal length and the f-number by 2 and ISO by 4.

This is correct. Of course I managed to misstate it above (right after saying it wasn't really that complicated).

 

[Glenn Smith]

N/T
--
Glenn Smith

You miss 100% of the shots you don't take... ~ Wayne Gretzky

 

[Glenn Smith]

If you use a hand held light meter you will find it is the same for all size film or sensors.
--
Glenn Smith

You miss 100% of the shots you don't take... ~ Wayne Gretzky

 

[Bogdan_M]

I think the numbers are roughly that APS-C is 2/3 of a stop better than m4/3 and FF is 1 1/3 stops better than APS-C.

Shoting with large f/stops you normally want all the DOF you can get! So I would say 4/3 is aprox 2/3 better than APS-C and two stops better than 35mm (FF is not a term I would use! If you photograph with a camera with a 45x60mm sensor, are you using Double FF?)

It just depends on how you define "normally".

There are 2 reasons for using large apertures: one is to obtain thin DOF for better separation between subject and background, and the second is because you want all the light you can get, when available light is scarce.

I never heard of people using large apertures (orlarge f stops) in order to obtain a large DOF. I think your definition of normal is contrary to the "normal" normal .

--
http://www.flickr.com/photos/bogdanmoisuc/

 

[Jonas B]

If you use a hand held light meter you will find it is the same for all size film or sensors.

True, but do you really think that is a reply to the OPs original question?

Jonas

 

[seachicken2000]

There are a lot of myths and pre-conceptions about how sensor size affects the way images from a system will look. I have held all of these beliefs, but recently by reading here have realized that they are all false. They just seem true because the cameras that are currently available behave that way.

• small sensors have more DoF
• small sensors are more noisy
• small sensors are lower resolution

What does seem to be true is that the lenses currently available for m4/3 and APS-C are effectively slower than their FF counterparts. This slowness is responsible for the noise, and the greater minimum DoF.

If it's the case that FF is the sweet spot in terms of fast lens size, perhaps the smaller sensor cameras will die out once FF sensors become cheap to make, as enthusiasts do want that narrow DoF.

It's surely possible to make FF cameras a fair bit smaller by having a range of slow lenses. These would give a system that had the same characteristics as a m4/3 system, but the camera would have advantage over a smaller sensor camera in being able to take fast lenses too.

I suspect this wouldn't work though, as some lenses would inevitably be a lot larger. A better plan would be a small mirrorless camera that could take both FF lenses and crop lenses on the same mount (the A900 does this, and automatically crops the sensor as needed). You could introduce the crop sensor cameras first onto the market (but they'd have a lens mount far larger than necessary and be the subject of ridicule). Then later introduce a model with a FF sensor compatible with the crop lenses when FF sensor production costs bring them into the mainstream.

I hereby predict this is Sony's game plan for the e-mount over the next 5-10 years, I think they've shown their hand by making the NEX lens mount preposterously large, and I will revive this thread at a later date when it all comes true to say I told you so. I further predict that m4/3 cameras will go a similar route.

I think it works as a game plan because:

• The step to FF improves the camera encouraging upgrading
• Sneakily keeps customers loyal to the system
• Gently 'helps' customers buy a whole new set of FF lenses
• I already want one

--
A rose by any other name is still a chicken.

 

[misolo]

Seems like everyone wants to look at it as the glass is half empty because its harder to get less in focus. It is much easier to get everything in focus with m4/3 than it is with FF. For a set amount of light and shutter speed FF will need a 2 stop smaller aperture to achieve the same DOF. That often requires a 2 stop higher film speed.

Maybe the glass is half full instead.

I believe this is not quite accurate. You can get the same DOF in FF by closing the aperture. The f-number is focal length divided by the diameter of the diaphragm. When a FF has an f-number twice as large relative to a 4/3rds lens, the two lenses have exactly the same size opening in the diaphragm . The sensor of the FF just happens to be twice as far from the diaphragm, but is seeing the same image, coming from an opening the same size and from similar optics (hence same depth of field, and therefore loss of sharpness due to diffraction will kick in for the same depth of field). In FF the sensor is four times as large and light is spread out accordingly with a fourth of the density, but it's the same opening and same total light (which is why at 4 times higher ISO the FF sensor has about the same noise as 4/3rd sensor -- this equivalence on noise is approximately only, of course, as technology differs from one manufacturer to another and improves over time).