I hear people talk about the beautiful colours produced by this lens or that lens, especially when talking about vintage lenses, something I have gained a passing interest in recently. I wonder if it is a moot point when shooting RAW and having the ability to manipulate colours in post processing, especially given that you can even calibrate with a colour checker chart. What are your experiences ?
Since what we interpret as color are various combinations of wavelengths of light, what a lens would do to influence that would be to attenuate parts of the spectrum. I believe modern lensmakers try to minimize such attenuation, aiming for achromatic lenses, but that is probably an imperfect endeavor. Still, I’ve made rather good camera profiles measuring spectrum through the lens (within 2 dE of lab monochromator measurements sans lens), so I don’t think that imperfection is major. I would call all of the above “informed conjecture”, so take it with a grain of salt…
All that said, I’ll bet older lenses might introduce major attenuation given the state-of-the-art in optics back then. “Warmer” and “cooler” would come from attenuation of the UV or IR ends of the spectrum, respectively, but that begins to look like white balance and would be cancelled out with the application of WB multipliers aiming to make a neutral patch R=G=B. Accordingly, I’d think you’d want to use one of the camera WB presets to retain such a lens influence.
I personally think all the influences on what light reaches the sensor are fair game, starting with the choice of illuminant(s) themselves, so I think lens influence does matter as long as you endeavor to understand the nature of the influence.
A lot of the above is thinking-aloud, so YMMV…
2 Likes
In my opinion - the hype about one lens vs another (especially when it comes to color) is a bit exaggerated. Not to say that it does not matter at all but other items (like sharpness and contrast) come to mind first.
I used to shoot Canon 70D and my 2 lenses were 18-135 STM and 50 mm 1.4. One time I made it a point to shoot one and the same scene with the 2 lenses on the same settings in camera. Even the zoom was the same (to match the prime). And I processed the same pictures the same way. The 50mm 1.4 came a winner. The colors were better for sure but the reality is that the contrast and sharpness were a big contributor in establishing my preference.
Now I use mainly Tamron but sometimes I attach ND (Tiffen). I have been happy with Tiffen in the past. It was surprising for me the first time when I noticed the significant change in the color. It was not just darker - it was somehow different. And it does come to issue when you are to shoot a person and some pictures are with ND and the others are not.
In my opinion if 2 lenses are optically identical when it comes to sharpness and contrast - probably the best way to asses would be to shoot one lens that you are familiar with vs the new and then when you compare the 2 judge if the difference matters to you.
At the end - a lot of it is personal preference.
1 Like
It may be possible to adjust colours in post processing but its not practical. The skill required to match the look of different optics in post is immense. The look changes in respons to a myriad of conditions and settings.
So in practice its worth looking for lenses with particular behaviour. Also in regards to colour and tone. Obvious things like colour temp are one thing but lenses can render mellow, pastel or with striking relationships between tones and colours. Perhaps the blue goes in one direction and the green in another giving landscape photos a distinct look regardless of “white balance” .
As with most lens properties it’s hard to find the conditions that reveal them. Over time you get the feel for it and over a range of photos the look will be apparent.
1 Like
not really. e.g. my tamron lenses on the nikon usually gave me a bit cooler colors compared to the Nikon native lens of my coworker.
Glass choice, bending of the light can influence which wavelenghts might interfer constructive or destructive.
1 Like
There is a reason why cine lens sets cost so much money. They ensure so many nuances in their designs that mostly don’t matter too much for photography.
- Same T-Stop actually passing through the same amount of light between lenses
- Color rendition matching between lenses
- Focusing not changing the FOV of the lens.
- On top for Cine zooms: not moving the focus point while zooming
- different focal lengths actually having the same physical size so that you don’t have to redo the rigging for the focusing motor used by the focus puller
and probably many more. It is kinda crazy and some times i am jealous. 
2 Likes
But if I use a colour checker to calibrate my photos taken with different lenses, and bodies for that matter, to make a custom profile for a given situation, does that not render it a moot point?
I am also curious, as an academic excercise more than anything particularly useful, is there a particular test that I can do to compare the colour rendition of different lenses? I mean, what would be a good subject to photograph that I could somehow meaningfully scrutinise?
It might be a fun play raw if someone could shoot a scene using a classic and a modern lens so we could see what the color rendition looks like between the two
If a lens varies the light spectrum smoothly, it is equivalent to white balance. If it varies abruptly, it can not be replicated by white balance. I would wager that most lenses vary smoothly.
What you are seeing as “better colors” may well be the influence of UV and IR light. They mostly affect the blue and red channel, respectively, and can lend an image a slightly glowy or hazy look. A lens with a stronger IR/UV cutoff may admit slightly less light, but be less prone to that hazy look, and therefore look “better”. Or “worse”, depending on your preferences.
Auto white balance is typically tuned to stick to “daylight” or “shade” when possible. Since lens color tints are usually fairly subtle, they are often not corrected by in-camera white balance. But can easily be corrected in post.
Here’s a test chart I made with a bunch of light sources and lenses:
The reference grey for each light source is simply the white balance without any lens attached. Two rightmost columns show the light source with two of RT’s white balance presets as a hue reference.
2 Likes
See
https://www.lensrentals.com/blog/2018/04/looking-at-cine-lens-color-shifts-using-spectrometry/
TL;DT: yes, there are differences, which can easily be fixed in post processing using a diagonal RGB correction matrix. Basically, white balance. Spectra are reasonably smooth.
I am not particularly looking to fix a colour shift, in fact I posed this question out of interest more than anything else. I wonder if the notion of a lens producing great colours, or punchy colours, or wonderfully saturated colours or whatever the observation may be, is a throwback to shooting film and not relevant to any large degree when shooting digital because we can post process.
I have an idea of shooting a scene with different lenses, maybe stretch it out over a couple of weeks to try different conidtions. I can use a grey card in the scene to white balance and set exposure. I can then perform the same, very limited post production on each photo and compare.
Besides those “standard” traits lots of cinematographers use classic or rehoused lenses for the specific look and aberrations they provide. See the Shogun TV series, the most recent Batman film, some dune ii scenes etc etc.
Mostly they are technically worse lenses that have expressive traits. Current cinematography colour grade their stuff to death so that aspect may not me critical but it’s starting to change and you can hear more and more people being sick or it.
1 Like
Try it! My feeling is that it’s quite difficult to find scenes that reveal characteristics of a set of lenses and shoot like for like. Moving images are a bit easier to compare as you can quickly and dynamically reveal how it behaves.
1 Like
If you were to do that for every combination of a lens and a body and every possible “given situation”, it would result in rather a lot of profiles … too many for this old man!
I shoot a color checker card, and color-pick the patches of interest (my preference CIELAB values). I enter those values into my spreadsheet which tells me the delta-E between one shot and another.
I don’t know if this is actually the case but my feeling is that what we see as colour is affected by veiling flare and other aberrations that reveal themselves most prominently in more complex situations than a test target shot.
I like the idea of shooting a test target, I do have one, and checking colours with a colour picker. I am not looking to over think this, I was just very curious (probably being nostalgic) about playing with older lenses because of those lovely colours people talk about.
I will take some photos !!
1 Like
haha, I’m not sure what you’re referring to with that image of colour such as we never see them. The flares are also of ghosting type not veiling flare. Perhaps I’m missing a joke here but for anyone not familiar with the terms. Veiling flare is basically flare that you don’t see the shape of. it affects the image by reducing contrast but can also skew colours.
1 Like
I would say the latter. Lenses do not affect color that much, manufacturers make sure they pass the visible spectrum more or less evenly (\pm 10\%), and any effect is trivial to fix in postprocessing. If you white balance your images, that will automatically fix any color cast that comes from the lens.
But even in the film era, lenses did not differ that much by color cast, or at least not by intention (eg some lenses using thorium oxide coating got yellowish, but that was considered a defect).