In a perfect world I would love to own the Canon, but I can live with any small difference in IQ for $1000 or so. I love L, but I'll buy anything that can get close for a fraction of the price. And in this case the lens is also well built and has HSM, so AF is fast (and hopefully accurate).
Bokeh Smokeh! It's clear that the Canon 85L renders the EXIF much better. Time for round 6: EXIF rendering. Don't be surprised if the Canon 50mm 1.8 pulls way ahead of the big boys here. Another thing, the Sigma EXIF looks like cheap digital. The Canon 85L EXIF has more of film look to it.
Doing some formal testing, I noted something that I thought was new; on further research it seems others have known about for quite some time... first formal test I've found
Canon cameras "compensate" for large aperture losses by adding additional amplification to all lenses that the body recognizes as "genuine Canon lenses" with apertures larger than F/2.8... :-)
Further note:
This extra amplification is determined by the body you mount the lens on. I'll try to check the 5D2 later some time. It seems that this (the 60D) has one of the higher amplification ratios - a 5D2 might show a bit less than this (larger photosites, less incident angle loss). But the 40D has roughly 115% at F/1.2 - so I doubt it will be much lower than this.
Anyway - the main point is:
Don't EVER trust anything coming out of a Canon camera as "the whole truth and nothing but the truth" from a scientific PoV. A lot goes on behind the scenes that you aren't supposed to know about.
So there's a very good reason the Canon 85L shows a brighter picture than the third party F/1.4 lenses... - it's been artificially amplified in digital postprocessing in the camera before the picture is saved to cr2.
This knowledge should be applied to comparisons between the 50/1.4 lenses too, like the Zeiss, the Siggy and the alt. lenses you can mount on a Canon.
That's interesting. I would expect that your rule to never trust Canon also applies to all the other manufacturers. We simply don't know all that is happening 'under the hood.' For example, I'm sure Nikon with their natively clean high ISO files are baking in a fair amount of luminance and especially chroma NR.
Perhaps this is one case where someone needs to test both lenses on a film body.
But I can understand why Canon might do this, because images shot at maximum aperture tend to be a bit on the darker side than those stopped down a stop or more. It's something I noticed back in the film days with Canon and also see with digital, with some lenses being more obvious than others. One that immediately comes to mind is the EF 35mm f/2.
Yes, this seems to be true for other brands also. But I think I can say for sure that only Canon keep this correction reserved for "only native brand lenses" - I can only test Nikon (D90, D700, D3x) and Sony (A850, A550, NEX5) right now, but it would seem that they correct for the given aperture regardless of lens brand - as long as it recognizes the lens communication scheme and can read the aperture setting.
BTW, the "NR" on Nikon raws is quite limited - it consists of a very simple pulse noise rejection scheme, a "speckle damper". It is not entirely defeatable. But neither is the quite more elaborate floating window median that Canon applies to all ISO1600 (and up) files that the cameras save... BEFORE sending the data through the in-camera settings of Noise reduction. :-)
This is getting totally out of scope for the thread - I just wanted to mention that the brightness difference between the Sigma 85/1.4 and the 85L isn't what it seems to be. The Sigma doesnt get the correction that the 85L gets. That's why the "difference" is visible.
The REAL difference is - just like I said earlier - about 1/5 Ev.
I don't understand why this is a mystery to you guys. Canon even advertises it and it's on every body Canon makes. It's called peripheral illumination correction. It artificially corrects for optical vignetting. The reason Canon only corrects for it's lenses is because it plots all of it's lens' vignetting patterns and enters them into your camera. Vignetting is a product of lens design and focal length, not just the aperture setting. Therefore, unless Canon took all third party lenses and individually plotted their vignette patterns they won't be able to accurately correct for the EV difference.
theSuede:
You should get paid for this, so you could be able to publish this kind of information earlier and on a more regular basis.
The scaling really annoys me. In reality, Canon then reduce the highlight dynamic range when using Canon lenses, and add another step of quantization noise in the shadows.
Ironically, what they do to seemingly favour their own lenses, works the opposite way for people knowing how to deal with it. The Sigma now looks even more attractive than it would if we hadn't been through this discussion.
dancam:
Peripheral illumination correction is something entirely different, and applied in software when decoding the RAW file. The sensor light falloff correction is applied before writing the RAW file, and is irreversible.
alundeb wrote:
The scaling really annoys me. In reality, Canon then reduce the highlight dynamic range when using Canon lenses, and add another step of quantization noise in the shadows.
This is what we have seen when using fast lenses, especially on 5D2. Look at my 5D2 posts, most of them using 50L.
The thing is, if you didn't have this kind of compensation, then reciprocity would no longer hold. That is presumably why Canon implemented this scaling. Because the light falloff due to oblique incident rays is dependent on the particular sensor design, each sensor must have its falloff characteristics measured and a custom curve applied. Otherwise, two images of the same subject shot at the same exposure settings with two different cameras could be of different brightness.
Canon's sensors aren't the only ones that are limited by this physical property--it is just that they decide to compensate for it in the ADC in combination with the gain calculation for ISO. Yes, it does reduce DR but the better way to think about it is that you never had that DR to begin with--just like with ISO, shooting at a high sensitivity comes at a cost.
Now, I want to be absolutely clear that I don't necessarily like the approach that Canon uses here. I'm simply putting forth the facts that (1) explain their reasoning for doing it, and (2) show that such light falloff is not specific to Canon-made sensors.
The DR was there to begin with, if it was limited in the low end by amp/conversion noise as it is at low ISO.
I can understand Canon's reasons for hiding it in the camera rather than leaving it for the software, and it is no surprise either than loss of DR is the price they are willing to pay.
There is still a 1/3 stop shutter speed difference between the Canon and Sigma consistently through all apertures, even at f/2.8. The images from the other thread with even smaller apertures are gone, but the difference in brightness at manual exposure with the same shutter speed was also consistent. One would expect the difference to be only at wide apertures, no?
How can we know that the Sigma images here are not also corrected for sensor light fall-off? The exif information is complete with correct aperture information, so Canon could do it. If I got CR2 files with both Canon and Sigma lenses, I could check the histograms for combing.
