guyfawkes: Reading many comments about "equivalence" both here and posted elsewhere, there does seem to me to be one question I'd like to ask: what practical applications are there that one would need to know this?
For example, if one owns a FF sensor dslr, does one really need to know what the so-called equivalence is for DoF for another format? Or turn this around. Would anyone moving from an APS-C system to FF need to know? How on earth would this impact on images they proposed taking?
In my days with film, using anything from sub-min up to 5x4, equivalence never even entered the equation. Photographers simply worked within and knew how to use the tools they used. So why has it assumed a proportion totally out of kilter with its importance now we use digital cameras?
It very much smacks of pseudo science to me.
I have to ask, have you ever shot photos with FF prime lens and 1/1.6f aperture?
I understand that choice of such a wide aperture is not suitable for every photographic situation, but sometimes it is "just right". And that is something you cannot do with m43 since its "1/1.6f" is actually 1/3.2f.
So yes, "equivalence" does have "very practical application in digital imaging".
Mike FL: As far as I can see from all the posts that WHO wrote this article created LOT of confusing.
I do not think there is any confusion, just denial.
Amin Sabet: Relationships between format size, depth of field, angle of view, focal length, etc, are not new or controversial.
What is referred to here and within the forums as "total light equivalence" is intellectually interesting but of more limited practical value. When you consider the vast array of sensor efficiencies we encounter, from films of various makes and speeds to digital sensors of different generations and designs, assumptions about image quality based on total light gathering will often fail to hold up.
What is underemphasized in this article is that for shooting in relatively limited light where shorter exposure times and non-shallow DOF are desired (applies to most street photography for example), there is no inherent image quality advantage to larger formats. On the contrary, smaller sensors tend to be slightly more efficient.
There is no accepted definition of "photographic equivalence", regardless of what this article or loud voices in the forum echo chamber suggest.
Hi Amin, thanks for the links above. There were solid (economical) rationale given by Eric Fossum for smaller, higher volume sensors leading in pixel/sensor technology. It is al about sensor design NRE amortization over higher volumes and exponentially lower manufacturing yields of larger FF sensors.
So I agree with your point that in the specific handheld (short shutter time), low-light, small aperture / large DoF use case smaller sensors have (some) advantage over FF due to their newer/better sensor technology.
Jylppy: Wow! 1000+ comments. What a great article!
I think there is a minor error in the paragraph about diffraction. Diffraction is pixel size, not sensor-size, related. Sensors with smaller pixels (ie. more pixels per area) suffer from diffraction "earlier" with smaller F-number. Usually smaller sensors (i.e. larger crop-factor) use smaller pixels to reach "required" MPx-levels (who would buy 5MPx m4/3rd when FF cameras have 20MPx+? These would have roughly same-sized pixels).
Good read about diffraction at excellent Cambridge in Color site:
And @bobn2, if you find a reference to a source describing diffraction as you see it right, pls share a link. I'll take a look. I am not an Optics expert, but I do have M.Sc. In Engineering Physics.
What I try to say that diffraction is purely an optics-space phenomenon (non-discrete math), where as digital sensors are discrete in the sense that each pixel is capturing photons arriving certain area. So no, diffraction in optics space does not depend on pixel size, and is by no means caused by sensor pixels, but once the photons arrive on sensor and image its being captured the diffraction impact on image quality at pixel-level depends on pixel size since smaller pixel pitch can and will detect lower levels of diffraction. I think we talk about the same thing.
But once I get back from holiday I can check this with colleagues who do design camera systems and hold PhD in Optics...
@bobn2. Could you please cite a reference to back up your claim? I was talking about pixel level diffraction and I find it theoretically impossible for pixel size not to impact on pixel-level diffraction.
Wikipedia entry says the same: http://en.m.wikipedia.org/wiki/Diffraction-limited_system
Figuring out this in the beginning of my hobbyist-photogapher career was a big "Heureka"-moment for me and the reason I sold off my 4/3 Olympus and bought Canon 5D. Yes, plenty of bulk to carry around, but I finally got the bokeh I was after. Everybody makes their own choices for the reasons important to them, but understanding "Equivalence" is fundamental to not to be fooled by the "F-numbers".
To clarify. Diffraction is an optical phenomenon, but reaching "diffraction limit" (i.e. The F-number at which it is diffraction that reduces the lens resolution) depends on pixels size. With larger pixels, larger "airy discs" will go unnoticed _at pixel-level_.
@Sabet. I think you miss a point here. Certain ISO-number on a given sensor is just a _calibrated_ standard sensitivity. It tells nothing about signal gain-levels in the actual sensor. Since mFT-sensors receive 4x less light vs. FF at the same aperture/shutter speed, it means the mFT-sensor needs to be driven at much higher gain-levels to reac equivalent broghtness and this lifts up signal-to-noise ratio. Without in-depth rationale and industry knowledge, I am not going to buy an argument that mFT-cameras have 4x "better" pixels (i.e. have 4x higher signal to noise ratios) than FF-sensors. Sony has the best pixels/semiconductor technology in business and they make FF-sensors too. Economical reasons might cause some difference since physically large FF-sensors are bloody expensive to manufacture.
Wow! 1000+ comments. What a great article!
Akpinxit: I would really like to see how 409600 ISO , full size still image looks like
I do not think that is the way to assess _still_ image quality under low light conditions. In video one can do noise reduction based on previous frames so the video shutter times do not provide fair comparison to equivalent still image shutter times. One needs to shoot still images and compare those.
There is now question the new Sony will be _awesome_ in low light. Sony leads in sensor technology, the camera has full frame sensor size and very large pixel (since low number of pixels). Sony Alpha 7s is the new Low Lght Beast(tm).
And I shoot Canon, so not a Sony fanboy here. I would be interested to buy a Sony Alpha MLC as a second camera, but I do not have the extra 2-3000€ to spend :-(
And yes, with ~10k€ in Canon camera and lenses, I am curious to see where does this camera system war end to. I think Mirrorless will win over DSLR.
Jylppy: Ah, Fanboys in the air! Since it is Nikon it _must_ be better than the independently reviewed award-winning Sigma 35mm/1.4, right? 1/1.4 aperture - who cares! Robust metal construction - plastic is lighter! ;-)
This new Nikon might well be a great lens, but we can know it only after thorough independent testing. Before that it is pretty pointless to make strong statements what this lens might or might not be. We'll see.
Ah, Fanboys in the air! Since it is Nikon it _must_ be better than the independently reviewed award-winning Sigma 35mm/1.4, right? 1/1.4 aperture - who cares! Robust metal construction - plastic is lighter! ;-)
Michael Ma: Wow, Sony a 7R leapfrogs any Canon or Nikon. Canon really needs to step up. Their days of purposely making cameras with lesser technology than their flagship is over. They need to not hold back now to stay in the business.
Sony 7R leapfrogging? Just compare it to Nikon Df at ISO 6400 and above. Yes, 7R has more pixels, but still. Df's low light perf is fantastic.
DPR folks, Pls add Canon 5DII so I (and many 5DII owners) can reference to a camera which performance they know.
Positive although very cautious move from Canon. This seems stop-gap measure at the best. I am still hopeful for a better dual-pixel AF EOS M with EVF emerging early next year.
The camera market seems to be dying from the bottom. Smartphones killed pocket cameras and entry DSLR sales / mirrorless sales are dropping. Growth is in pro / semi-pro segments, therefore the rumors of Canon looking to enter into medium-format.
Nevertheless, as 5DII owner I am looking to buy a second camera. The use case is to travel with two cameras/lenses so no need to change lenses on the road. The other camera should be smaller, i.e MILC. 100D does not cut it. So I am believer in mid-high-end MILC segment, but how well does my personal need correlate with larger market? Anyway there are plenty of interesting opportunities in MILC with new imaging technologies...
"Biological_Viewfinder, while you may be right in general, some of the lenses actually beat their Canon/Nikon equivalents (at least Sigma 35mm/1.4f). It seems Sigma and Tamron have greatly narrowed the gap to Canon/Nikon top models with their new products. I own three L-zooms, but I have been very satisfied to my Sigma purchases.
Mateo Miller: Sounds interesting but it weighs half a pound more than the Canon 24-105L 670 grams vs 885 grams (23oz vs 31oz)
Auts! Whether sealing would be a high priority for this kind of general purpose travel lens. Oh Sigma, what have you done?
I have accidentally tested my Canon 24-105/4L's whether sealing and it works. It got poured about 20 liters of water directly on camera (5DII) + lens on a river boat without any problems at all.
No, I am not gonna test my luck second time.
Karl Summers: Please DPR, make this a testing priority. A test against the Canon f/4 24-105 and the f/2.8 24-70-II, as well as the Nikon equivalents. I don't expect the build quality to compare, but if the optical quality is like their 35mm ART lens then I would love to get this piece of glass.
Well, I have Sigma 35mm/1.4f Art series and its build quality is superb. Solid construction, beautiful industrial design. I love it. It easily compares well to Canon L series lenses in build quality, but its industrial design is more stylish. Go Sigma, go!
siberstorm27: These comparisons will inevitably lead to fanboy flamewars here in the comments section, most notably from Nokia and Sony fanboys with little sense and big egos. The cameras, especially on the Nokia, have overblown claims of superiority based on fancy tech jargon. The end result is a mixed bag, and all these phones produce phone quality images regardless. You can focus on the "phone" part without fear of losing much image quality from one phone to another. The Galaxy S4 Zoom, however, stands out as having a unique 10x zoom that you can't approximate on other phones. It's also way thicker.
And this comment was written by ... Samsung fanboy? ;-D
I have used the version 3 and I am not satisfied at all.
The sync is all but reliable and tends to run problems so often so I just gave up. I re-installed the plugin and app many times without success. First it seemed to work OK, but then it start to run in to problems with sync - could not figure out which way to sync or got stuck completely to certain images.
I hope they can manage to fix the issues so my money would not be completely wasted.