"Minolta Colors" - urban legend or reality?

[stevemark]

Some weeks ago "tf" did a carefully prepared test, comparing many Minolta lenses with two Canon lenses from the 1950s and a recent Sony lens. No differences were visible in his test results:

https://lensqaworks.com/2019/09/09/minoltas-unique-color-rendition-a-faith-or-a-physics/

tf concludes that there's no such thing as the "unique Minolta color rendition" - in other words: it's an urban legend. When discussing his results, everyone did agree with tf, congratulating him for debunking some well established fake news. You may remember that I did not agree with tf, promising to add some information from my own experience.

Of course the test results shown by tf are valid - no question about that. They document the color transmission of the tested lenses under certain conditions. It is very important not to forget that!

When i saw the test results, i immediately knew what had gone wrong.

In real life photography and especially in landscape photography you rarely have large areas with A) exactly the same color and B) without any texture. Yet, tf has tested the color rendition of lenses using relatively large areas of unstructured colors:




In real life photography, we usually have lots of textures - grassland, trees, clothes, brick walls, roofs ... even the surface of a lake usually has a rich texture and often lots of bright reflections.

Look at this image:

Of course everyone agrees that the grassland is green (left square), and that the tree is dark green (right square). Every lens, according to tf's test, should render these areas identically. But I have some doubts.

Let's have a closer look.


It's immediately obvious now that both the grassland as well as the dark part of the tree have lots of fine textures with different colours. Everyone with some experience in photography will agree that a good, sharp and high-contrast lens will render these different from a "bad", blurry and low-contrast lens. And in turn, the overall colour (="average colour") will look different!

Here are a few crops from the above scenery, taken with six well known vintage MF normal lenses. The conditions are ideal: The light is very soft, there's no back light, all images were taken at f5.6, there are mainly different shades of green (no other colors, most photography lenses are specifically optimized for green light!) and the crops are more or less from the center of the image. These conditions are the easiest for any lens - no lens should have difficulties under such conditions !!

Yet, the results are striking:


There's a Minolta MD, a Zenith, a Zeiss CY, a Canon nFD, a Mamiya, and a Topcon lens up there (NOT in this order!! ... try to find out yourself!!). I hope you can see the differences, and i hope that you now understand that "tf"s test was really misleading, even though, technically, it was very well made.

Now imagine how big the differences would be in difficult conditions: back light, wide open, corners ... let alone a mixture of really different colors!

Yes, there are huge differences in color rendition.

Stephan

[benadamx]

the samples in the big test diagram are too small to really get any good sense of the color rendering (something the eye/brain is notoriously poor at dealing with), other than that even in that ideal controlled environment, subtle differences can be still be detected.. your whole-scene tests plainly reveal the flaw; perhaps if we could see the set of full photos of the colored pencils set, we could make similar comparisons

[Alun Thomas]

In the landscape samples posted, the differences are fairly obvious. In general when faced with such a comparison, I would describe the difference as more one of contrast, some of the images display much more than others. Of the ones showing poor contrast, they appear to be suffering from veiling flare likely caused by less effective antireflective coating(s). The example given does lend credence to the idea that better contrast can also be desribed in terms of increased colour saturation.

I've tried lenses from a wide range of mainly Japanese manufacturers (i.e. Nikon, Canon, Pentax, Olympus, Yashica, Mamiya) and, especially in the 40-60mm range, have yet to experience a bad lens. Having said that, I have tried almost all the Minolta (SR mount) 40-60mm lenses, and they did all stand out (to my mind) at their initial test, in terms of the contrast and subsequent colours seen, forming a quite pleasant to view image. The lenses range from the early 55/2 and 55/1.8 up to the MDIII 50/2, and almost all variations in between, except for the 53mm and the 58/1.2 (pricey). To my mind, it seems Minolta mastered that aspect of lens design quite early on, whether it was coating type or some other design parameter.

[aidaho]

[hemeterfilms]

As long as individual human eyes are responsible for making the final judgement , the results will be unreliable to say the least.

[Sciolist]

Has anyone seen the plate of mixed pasta that is the melange of receptors and neural pathways used to transmit colours to our consciousness? It's a wonder we see anything at all.

We don't even perceive identical hues across a range from each eye, never mind comparing to what other people are perceiving. And I am referring to within the range of 'norm', not colour blindness.

When trying to agree a consensus on a special attribute to a colour range, the fly in the ointment will firstly be us, not the gear or method, surely.

[aidaho]

[D1N0]

From the greenery comparison I see mainly differences in contrast. Apart from the bottom right one which is obviously cooler than the other ones.

[tf]

Thanks for those tests.

As a 'starter' of this "color challenge," I wouldn't like to push my own opinion, because we still didn't make an agreement about what everyone should mean under the 'lens color affecting'.
My position is based on the maximum possible exclusion of aberrations. Because aberrations are unstable, so, for 50% of cases it will be better coloring, but in other 50% - worst colors. In other words, I avoid to say something like:
"Aberrations of this lens are better for landscapes colors... but... sky should be clean, grass should be dry, the Sun should stay under 45 degrees over the skyline and little clouds should lie over the trees close to the horizon. But this lens is worst if the sun goes to lunchtime and it absolutely useless in winter".

From this point of view, for me it looks directly - 'all lenses are the same'. I've never chosen a lens because of the colors but select my preferable lens because of sharpness, fewer aberrations, size, and convenient control.
Just please, don't mix it with simple "the less aberrations - the clean colors" - we all realize it.

That's why I fixed the conditions, cut the center of the frame - and yes, in this case - all lenses display similar colors.
(Those who can't recognize it on 100% size on 1/3 of the frame and need bigger pictures - may use Photoshop to up-size, it should underline a difference even more, but don't expect a miracle, I've checked )

A few weeks ago such tests have been performed on FaceBook. I recommend looking at it for a better understanding of how even the small changes in light (a little bit more or less clouds out of the scene) may change picture colors.

[stevemark]

[visualopsins]

I'm a fan of Takumar colors, myself. lol

I can see the Minolta colors.

I think coatings and CA formula affect color balance.

[tf]

[cbass]

I am not sure if all Minolta lens generations had Minolta colors or if they gave up. However, I have noticed a difference in how Minolta lenses render reds. Mine are all Rokkor lenses though.

[dan_]

As I remember, back in their time Minolta lenses were known not principally for a special color rendition, but for having a consistently equal color rendition from lens to lens, regardless of the type of lens. You could change lenses and have exactly the same colors in your photo. In the film time you had no Auto White Balance and film was not forgiving - that's why having consistent colors between lenses was very important, leading to the fame of "Minolta colors". Probably a test on film could better reveal if the "Minolta colors" label is a fact or an urban legend.

Based on my experience with the Minolta lenses I too think that they have a distinct rendering, especially for greens and reds. I like their colors but I can't say I like them more than the colors from other lenses.

[visualopsins]

Excellent and I think correct point dan_! That's same with m42 Super-Multi-Coated Takumars - colors are consistent for all focal lengths. I think due to same coatings and same CA corrections.

[Antoine]

The Minolta colour"consistency" taking into account
- Films colours unlike digital cannot be modified and cameras had no impact (unlike today where Nikon, Canon and Sony do not have the same neutral colours/rendering)
- Still pictures taken with different films yielded different colours (Kodak, Fuji,...)

The Minolta "consistent" rendering may have been impacted by the following:
- Made their own glass, control all its processes
- Made their own lenses until the early 80s
- Their early lens apochromatic multi layer coatings (1958) + evolution with their own modern processes (which does not contribute in tf tests)
- A deliberate consistency/continuity policy... also imposed by making everything inhouse (until the AF revolution... which killed their business...)

No other manufacturer had a similar policy of consistency and progressive evolution (but Nikon went further when they moved to AF...)

-

[visualopsins]

Lens can't affect film color balance?! What are all those filters used for?

I remember video guys came here looking specifically for sets of lenses with consistent color balance over different focal lengths...

[RnR]

Minolta and Contax Zeiss glass is widely recommended for alt-glass cinematography, because of their consistent colour signature.

There are two threads over at the reduser forums discussing this;

https://www.reduser.net/forum/showthread.php?92246-Minolta-Rokkor-Survival-Guide
https://www.reduser.net/forum/showthread.php?92044-Contax-Zeiss-Survival-Guide

[michelb]

[RnR]

[Ernst Dinkla]

One day I will find time to measure lenses with my spectrometer. Probably gives more insight than creating camera/lens profiles with a Color Checker Passport + software. Finding a stable continuous spectrum light source is the main problem. Which aperture suits best is another thing, wide open the spectral transmission of many lenses is different as at say 8.0. Wonder if the internal reflections are playing a role in that.

[Teemō]

[RnR]

[Ernst Dinkla]

[Gerald]

Ideally, a lens should be absolutely transparent to all colors, but in practice, the spectral transmission is not flat, so color balance varies from lens to lens. The reasons for this non-ideal behavior of a real lens are various: types of optical glasses used, different coating formulations, etc. The coatings, in particular, can be manipulated by the lens manufacturer to make the tonal balance neutral, or slightly tending to a warm or cold tone.

The graphs below show the spectral transmission curves of two common normal lenses:



(from: https://www.researchgate.net/figure/Normal-objective-spectral-lens-transmissions-for-the-Nikon-50mm-f-14-AF-D-red-solid_fig7_40484060)

It important to note that correction of chromatic aberrations has NO relationship to the color balance of the lens! To be more exact, no doubt that CA can introduce false colors into micro-details, such as fringes and colored halos, but CA does not change the colors of large areas of the image. So, it is a mistake, for example, to imagine that an APO lens necessarily produces purer or brighter colors than an ordinary lens.
-----------------------------------------------------------------------------------


Back to the interesting Stephan's test, in my opinion the differences in color rendering of shadow areas of the image are not the result of differences between spectral transmissions of the lenses, but of differences between light scattering by the optical surfaces or the optical glasses themselves.