Home
SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

Colours?
View previous topic :: View next topic  


PostPosted: Thu Sep 13, 2018 3:44 pm    Post subject: Colours? Reply with quote

I'm not an expert. There's lots of stuff I know little about, including some of the basics. One of these is colour. People often write that this lens has good colour reproduction or that lens has famous [insert-brand-name-here] colours etc. But what does that actually mean? I mean I know some lenses give warmer or colder colours than others at the same conditions but so far I haven't actually seen any consistency that could make me say that this or that line of lenses has certain signature colours that people seem to talk about.

So I went and did an experiment. I shot a test scene with 15 different lenses, all in the same (or at least very similar) conditions, at the same settings and the same processing. Everything set manually to the same settings, same WB, no auto anything. All lenses were shot at 50mm/5.6 (except for one which I couldn't close on my selected adapter). All 15 shots were taken within a 12 minute timeframe. This is the result (click for full size):



Now I know even less what to think. Yes, there are some minor differences. Three of the lenses produce noticably warmer colours than others, and two or three others lean in the opposite direction (on one or two the lighting might have changed ever so slightly judging by the hardness of the shadow on the wall). What really bothers me is that I can't see a pattern. There's not enough consistency for me to be able to say that some lenses have specific signature colours that are better/different from others. You know, the colours people talk about in reviews and such.

What do you think? Is my test flawed? Light wasn't consistent enough? Am I looking at the wrong thing?


PostPosted: Thu Sep 13, 2018 4:31 pm    Post subject: Reply with quote

I gave up on "testing" lenses, since I found out that it most often just "tests" aka documents the (in-)ability of the tester - and not really the lenses. Those are often THAT good that only a capable optical lab can really show the slight differences lenses have. Twisted Evil


PostPosted: Thu Sep 13, 2018 4:55 pm    Post subject: Reply with quote

kds315* wrote:
I gave up on "testing" lenses, since I found out that it most often just "tests" aka documents the (in-)ability of the tester - and not really the lenses. Those are often THAT good that only a capable optical lab can really show the slight differences lenses have. Twisted Evil

I would tend to agree with that when we're talking about resolution and other such precise pixel peeping/grain sniffing technicalities, but surely colours are much more subjective and differences are quite visible to everybody, even at this tiny size as I'm presenting here. People use phrases like for example "the famous Minolta colours" but how do you quantify that?

I wasn't happy with my first test because I think the sunlight changed a tiny little bit from shot to shot, so I reshot the whole thing. This time the light was really consistent and this is what I got (shots in the same order as first time):



Results are similar. A few lenses produce warmer colours, a few cooler, a few have more green than others, a few more magenta. Two clearly have lower contrast but others are very close to the same. The problem is I don't know what "Minolta colours" are and how I can recognize them. Reading various reviews and people praising lenses of various brands for their signtature colours, I'm sure such a thing must actually exist.

So can anyone possibly tell the Minoltas in this test from the others only going by the "Minolta colours"? (the tested lot consists of a bunch of Minoltas of very similar vintage and various russians and east germans)


PostPosted: Thu Sep 13, 2018 5:48 pm    Post subject: Reply with quote

I agree with Klaus (and always do!)

Much of this is perhaps over-emphasis of minor differences. Some lenses, and families of lenses, do have common traits. Some are relatively warm, some are cool. Some are a bit different in their quality of warmth - Komura is a bit "golden" for instance. But these subtle differences are adjustable in in PP of course.

The truth is that unless you are submitting your work to a truly anal-obsessive photo editor (and with the decline of glossy magazines, do these even exist?) this stuff doesn't matter except as a question of taste and maybe of personal morale and confidence.


PostPosted: Thu Sep 13, 2018 6:42 pm    Post subject: Reply with quote

This is an interesting (informal) test, and for once not about the usual nonsense such as corner performance at wide open aperture. Nice to see these differences, that are definitely there!

I haven't used any Minolta-lens for quite a while, but these are mostly famous for having consistent colour character over various focal lengths.

On the opposite side of consistency, my Samyang 12 was very cold, and the 8 very warm in character.


PostPosted: Thu Sep 13, 2018 8:21 pm    Post subject: Reply with quote

Thank you so much for these examples!!! Excellent test imho.

Lens formula CA correction. Coatings. Imho these are the primary factors affecting color signature.

Please show which lens made which image.


PostPosted: Thu Sep 13, 2018 8:54 pm    Post subject: Reply with quote

I can't even begin to see any difference between the images.

Ask me, I'd say your camera does a pretty decent job of white balance. Cool


PostPosted: Thu Sep 13, 2018 9:06 pm    Post subject: Reply with quote

White balance was set to a constant not auto... Smile

Actually, pixel peeping colors is the method to compare these images; compare rgb values...


PostPosted: Fri Sep 14, 2018 6:19 am    Post subject: Reply with quote

cooltouch wrote:
I can't even begin to see any difference between the images.

Ask me, I'd say your camera does a pretty decent job of white balance. Cool


Differences should be easily visible, most obviously on the tone of the background wall but also just the general feel of each image is slightly different. As I said white balance was exactly the same for each image and light was too (in the second test at least). Any differences are down to the actual character of the lenses that were used.

For example the image in row 1, column A (let's call it A1) is slightly warmer than average, image B1 is by far the warmest with an almost orangey feel, C1 is cooler with a slight magenta cast, D1 has a slight greenish cast and E1 is the most neutral at the WB settings I chose. (I'll take this as my reference for the rest.)

Moving on to row 2, image A2 is a tiny little bit warmer than E1 but otherwise quite close. Then B2 to me looks like it has a tiny little bit more green, C2 is noticably warmer than most and D2 and E2 are almost indistinguishable from E1.

In row 3 images A3 and E3 are again very similar with perhaps a touch less magenta/more green while the middle three (B3, C3, D3) are visibly cooler than the rest. They're not exactly the same but none is really similar to any of the others. B3 and D3 have an almost bluish purplish cast and C3 and especially D3 also show noticably less contrast.

Now here comes the problematic part. The lenses used were the following:

A1 = Minolta MD Zoom 28-85mm/3.5-4.5
B1 = Minolta MD Zoom 35-105mm/3.5-4.5
C1 = Minolta MD Zoom 35-135mm/3.5-4.5
D1 = Minolta MD Zoom 50-135mm/3.5
E1 = Minolta MD Zoom 35-70mm/3.5

A2 = Minolta MD 50mm/3.5 Macro
B2 = Minolta MD 50mm/2.0
C2 = Minolta MD Rokkor 50mm/1.7
D2 = Minolta MD 50mm/1.4
E2 = Carl Zeiss Jena Tessar 50mm/2.8

A3 = Carl Zeiss Jena Pancolar 50mm/1.8
B3 = Meyer Optik Gorlitz Domiplan 50mm/2.8
C3 = Industar-50-2 50mm/3.5
D3 = Industar I-61 52mm/2.8
E3 = Jupiter 8 50mm/2.0

I see very little consistency among the Minolta zooms (top row). Each image is visibly different than the rest. In the second row B2 and D2 are consistent with each other and very similar to E1 but the other two Minolta primes are visibly different and are more similar to D1 and A1 respectively. What's that all about? There are 9 Minolta lenses in the test and I see about 5 or 6 different colour signatures. So what are these famous "Minolta colours" people rave about? And the same goes for pretty much any other brand.

To make things worse, E2 produces pretty much exactly the same colours as some of the most highly praised Minolta zooms and primes while A3 and E3 are also extremely close. (B3, C3 and D3 are clearly outliers and let's say they are "inferior")

The more I think about it, the more I feel that all that talk about Minolta colours, Canon colours, Pentax colours, Zeiss colours etc. is nothing but bull[word-we-can't-use-on-this-forum].


PostPosted: Fri Sep 14, 2018 7:05 am    Post subject: Reply with quote

miran wrote:
The more I think about it, the more I feel that all that talk about Minolta colours, Canon colours, Pentax colours, Zeiss colours etc. is nothing but bull[word-we-can't-use-on-this-forum].
Exactly, this. There are numerous people(photographers) who did test "the X's colors". The result was always the same - they are non-existant.

The only general noticeable difference is a yellow tint caused mostly due usage of thoriated glass.


PostPosted: Fri Sep 14, 2018 7:56 am    Post subject: Reply with quote

y wrote:
miran wrote:
The more I think about it, the more I feel that all that talk about Minolta colours, Canon colours, Pentax colours, Zeiss colours etc. is nothing but bull[word-we-can't-use-on-this-forum].
Exactly, this. There are numerous people(photographers) who did test "the X's colors". The result was always the same - they are non-existant.

The only general noticeable difference is a yellow tint caused mostly due usage of thoriated glass.


Got any links to those numerous people's tests of the X's colors?


PostPosted: Fri Sep 14, 2018 8:07 am    Post subject: Reply with quote

visualopsins wrote:
Got any links to those numerous people's tests of the X's colors?

I don't remember all the blogs/articles. But this one stuck in my head: https://mbphotox.wordpress.com/2016/06/14/colours2/


PostPosted: Fri Sep 14, 2018 9:10 am    Post subject: Reply with quote

y wrote:
miran wrote:
The more I think about it, the more I feel that all that talk about Minolta colours, Canon colours, Pentax colours, Zeiss colours etc. is nothing but bull[word-we-can't-use-on-this-forum].
Exactly, this. There are numerous people(photographers) who did test "the X's colors". The result was always the same - they are non-existant.

The only general noticeable difference is a yellow tint caused mostly due usage of thoriated glass.


Colour perception is highly subjective. Many people seem to turn the colour control on their televisions right up to near maximum, I can't watch it like that and prefer it turned down a bit from mid range.

I certainly don't think a yellow tint is the only bit at all. I've used a spectrometer at work to try to measure the transmission of some of my lenses across the UV/visible/NIR region. There are very distinct differences at the extremes but often noticeable changes in the visual region. Many of the changes might be lost if using the same sensor/film for testing all the lenses, but then the same 'canon colours' etc. is used to talk about digital sensors too.

This is compounded by variation in the eyesight of the photographer - I've read scientific papers that suggest young people typically see further into the UV than older people and saw a DRAMATIC change in my blue perception when I had minor cataracts removed - at one point having a very different sense of colour from one eye to the other..

Colour perception is a very complicated subject, people think there are three colours that get blended to make the entire visual range, but in reality its an (effectively) infinite continuum that the eye/film/sensor splits into 3 groups. In humans (at least those without colour blindness) there are actually 4 light sensitive cells in the eye, Long, medium & short wavelength cones, (which normally give us colour perception from the relative proportions of their overlapping sensitivities) & rods (used in low light). As light levels drop our colour perception changes as more of our vision comes from the rods, cameras do not experience this change.

In any photographic system a minor gap in sensitivity in the visual band of a few nm wavelength range would not be too noticeable in most circumstances, but if the light source is purely monochromatic at just that wavelength suddenly there'd be nothing for it to see. Optical systems are not normally quite so dramatic but can make very noticeable changes - just try adding a weakly coloured Didyminium filter (AKA 'red enhancer') to a scene lit by low pressure Sodium lights.

Just imagine if our camera sensors were trying to replicate the colour vision of a mantis shrimp which has 12 colour receptors! The Bayer matrix would have to be much more complicated Twisted Evil


PostPosted: Fri Sep 14, 2018 10:25 am    Post subject: Reply with quote

y wrote:
I don't remember all the blogs/articles. But this one stuck in my head: https://mbphotox.wordpress.com/2016/06/14/colours2/

Great article. Thanks. Smile


PostPosted: Fri Sep 14, 2018 10:32 am    Post subject: Reply with quote

I don't dispute the complexity of color science field. The "X's colors"... it's the same in HiFi world - "directional cables", "demagnetized CDs", etc. Ppl will say they can hear/see the difference. Is it subjective? Yes, it is.

Will a blind-test prove anything?


PostPosted: Fri Sep 14, 2018 10:54 am    Post subject: Reply with quote

I played around with colour channels a bit looking at the grey background and I find that the differences are exaggerated more with the purple and blue channel for the colder ones and the yellow channel for the warmer ones. Aqua also has a bit of influence. Not much happening in the other channels.

Here I tuned op purple blue aqua and yellow and maxed out vibrance.



The czj tessar seems the least affected. I guess it is down to coatings.


PostPosted: Fri Sep 14, 2018 12:05 pm    Post subject: Reply with quote

y wrote:
I don't dispute the complexity of color science field. The "X's colors"... it's the same in HiFi world - "directional cables", "demagnetized CDs", etc. Ppl will say they can hear/see the difference. Is it subjective? Yes, it is.

Will a blind-test prove anything?


I see a very clear difference between e.g. A2 and C3, especially on the background, but also other areas of the image.

Maybe the fact that different people see more or less differences (or none at all) comes down to monitor used, eye sensitivity, or other factors.

Posted the 2nd test image again for convenience:



PostPosted: Fri Sep 14, 2018 3:15 pm    Post subject: Reply with quote

miran wrote:
cooltouch wrote:
I can't even begin to see any difference between the images.

Ask me, I'd say your camera does a pretty decent job of white balance. Cool


Differences should be easily visible, most obviously on the tone of the background wall but also just the general feel of each image is slightly different. As I said white balance was exactly the same for each image and light was too (in the second test at least). Any differences are down to the actual character of the lenses that were used.

Oops, missed that part about your white balance. And yeah, now that I take a closer look, I can indeed see small differences. To me, the three biggest standouts are the middle three images in the bottom row -- noticeably cooler than all the rest. But these minor differences are not what my eye tends to be attracted to. I tend to notice most strongly the saturation and contrast, and viewed from this perspective, there's just not a whole lot of difference between any of the images.

And I agree completely with:
Quote:
The more I think about it, the more I feel that all that talk about Minolta colours, Canon colours, Pentax colours, Zeiss colours etc. is nothing but bull[word-we-can't-use-on-this-forum].

Sometimes you just gotta call it like it is. If it's bullshit, then that's what it is.


Last edited by cooltouch on Fri Sep 14, 2018 3:18 pm; edited 2 times in total


PostPosted: Fri Sep 14, 2018 3:16 pm    Post subject: Reply with quote

I call into question validity of the test reported in blog. White balance is key. Setting wb in pp using resulting image, rather than setting a constant in the camera for use by all test images for comparison makes no sense. Miran's test is careful with white balance and pp, clearly shows there is indeed a difference in color balance between lenses.


PostPosted: Fri Sep 14, 2018 3:34 pm    Post subject: Reply with quote

visualopsins wrote:
I call into question validity of the test reported in blog.
It seems as if the purpose of the test on the blog-post was to show that with processing, you can get close to the same output with different lenses. But that is a very different test than Miran's.

I think that if some here do not see the differences in Miran's images, it could very well be related to our individual monitors (and settings), eyesight, etc.


PostPosted: Sat Sep 15, 2018 1:43 am    Post subject: Reply with quote

kds315* wrote:
I gave up on "testing" lenses, since I found out that it most often just "tests" aka documents the (in-)ability of the tester - and not really the lenses. Those are often THAT good that only a capable optical lab can really show the slight differences lenses have. Twisted Evil


Agreed--%100. jt


PostPosted: Sat Sep 15, 2018 8:56 am    Post subject: Reply with quote

miran wrote:


The more I think about it, the more I feel that all that talk about Minolta colours, Canon colours, Pentax colours, Zeiss colours etc. is nothing but bull[word-we-can't-use-on-this-forum].



Firstly, thanks for taking the time to post this up. I've found it fascinating.


I'm one of the ones who can see quite a difference between these lenses. I'm actually surprised at how much difference, but as others have noted, it could be my screen that is contributing to it.



What springs to mind, and is probably as far as I can judge, is the saying that Japanese lenses tend to be warmer and German lenses tend to be cooler. This seems to more or less stand up if you judge them as two groups rather than individually.