See Sigma 105mm f/1.4 DG HSM Art: Substantial Color Shift Going from f/1.4 to f/2 and Beyond for detailed evaluation of how color balance and tint from f/1.4 to f/2.
UPDATE: oops, Chris R was referring to the selfie aperture series. As it turns out, I used the wrong tint for that series (+8M); it should have been +2M according to my recent evaluation of the Sigma 105/1.4 color rendition. For accurate color be sure to shoot a SpyderCHECKR card for each lens and camera—don’t assume anything about the color rendition.
Subscriber Chris R writes:
I’m really enjoying your testing as usual and very interested in the new Canon 85mm 1.4L that you have just posted, I have seen other online tests where they have put it up against a few of the current 85mm lenses such as the Sigma Art, Tamron VC 1.8 and the good old Mr Otus, and yet they often pass over the colour rendering of lenses which is as important as their classic chart tests! Which are as much use really as a chocolate fire guard in the real world.
After seeing their results and after using a Canon 50mm f1.2L for many years, I noticed that the colour from the Canon lenses tends to have a slight pink/magenta balance, you can see this in comparison especially against the Zeiss lenses mainly and I can see the slight pinkish colour in your self portrait skin tones from those comparison images in your review here.
But I wonder if you got chance to run a few comparison shots of the Otus 85 alongside the 85mm 1.4L just purely to compare the colour, we all know about the sharpness but so many tests/reviews forget to cover the colour and rendering fully as you do. Also as you are probably aware, I think Canon were aiming the new 85mm at mainly event, portrait and wedding photographers, but your review is the most thorough iv’e seen so far.
I’m really interested and enjoying the images you’ve produced with the new Sigma 105mm 1.4 Art lens, I shoot mainly lifestyle and commercial work and I was really interested in seeing what this lens can do, especially the out of focus rendering, but i’m curious, as with the Sigma 135mm Art lens they tend to be slightly contrasty in comparison to say the Milvus 135mm and 85mm Otus, so again, I was wondering, could you possibly run off a few shots with your Otus against the 105mm Sigma mainly to show the difference in colour rendering and tonal quality, if your’e all done testing then fair enough.
DIGLLOYD: I agree absolutely that lenses have different color rendering, and that has been a weak spot in my testing, point taken and something I will address in the future. As to the Canon 85/1.4L IS, I do not own/have the Otus 85 on hand in Canon mount, so it was not something I could easily check (I do have the Nikon mount Otus 85, but I’m not going to stress the lens mount with an adapter).
As another example of color rendering, older Nikkors tend to be significantly more blue than newer ones with the more sophisticated coatings. The challenge of applying lens coatings to highly curved surfaces can result in off-center color shifts particularly with wide angle lenses, a challenge addressed with some of the lenses in the Milvus lineup versus the prior versions.
There are also ray angle issues (yes, even with a DSLR): the Sigma 105mm f/1.4 DG HSM is distinctly more magenta in color at f/1.4 than at f/2, a behavior I’ve seen in quite a few other fast lenses, and unavoidable on a digital sensor. It has to do with ray angle and off-center vs central rays striking the digital sensor. So the question of lens color actually applies even to a single lens, at different apertures. It is for this reason that one cannot just assume a correct color balance for all apertures during raw conversion! I had to work hard for the portraits for to get it right for that and also the mixed lighting—and at some point it becomes subjective.
Then there is the question of correction for secondary color, which itself can impart a significant greenish cast to background out of focus areas, and magenta cast to OOF foreground areas (which is why a gray card must always be shot in focus and never even a little out of focus). A lens with outstanding correction for secondary color will appear to have substantially better color rendering. Because it does, in practice since there is almost always out of focus stuff for normal and telephoto lenses. But this is really about correction for secondary color and thus distinctively separate and apart from color rendition for in-focus subject matter—yet the two cannot really be separated. This is one major reason why I hugely prefer lenses with outstanding correction for secondary color. The Zeiss Otus is one such line, as are most of the Sigma DG HSM Art lenses. Sigma does an absolutely amazing job in regards to correcting for secondary color—far superior to Leica, for example, even when Leica calls a lens “APO”.
The new “Adobe Color” camera profile requires +0M for the same conversion for which “Adobe Standard” requires +8M. When evaluating lens color, such factors can easily overwhelm any difference in lens color. This is why I prefer to shoot my own Macbeth color checker card.
One must not assume skin tones from past experience/familiarity. For example, due to a medical condition, my daughter (below) actually does have skin color which often goes distinctively reddish/magenta. I’ve heard this comment before (“slight pinkish colour in your self portrait skin tones”) for this same daughter, and it is a (quite understandable) error in assessing the skin tones. It’s the skin, not the lens or camera or raw conversion.
In the portrait below from Sigma 105mm f/1.4 DG HSM Art for Portraits, the lighting is mixed: warm frontal lighting reflected off a grassy slope but also with some extra green in it from trees in that area, plus surrounding green foliage, plus blue overhead skylight (see the bluish hair at top!). I did my best to find a pleasing balance. Assessing lens color rendition under such conditions is inappropriately error prone.
Color and contrast matter to images. There is something about Zeiss that I like a lot, as per below.