Focotar 50/4.5 vs Edixagon 50/2 vs Travelon 50/1.8

[Alex H]

I recently got a bunch of old lenses in M42 and Exakta mount that came as an add on for something I was looking for. So before putting them up on eBay or tossing them into the recycle bin, I will test them for UV transmission, by comparing them with my main UV-lens, Leits Focotar-II 50mm F/4.5. The first test is below,and the test subjects are Edixagon 50mm F/2.0 (left) and A.Schacht Ulm Travelon 50mm F/1.8 (right).

 

 

Test was performed hand held to guesstimate approximate UV transmission, ISO, WB and F-stop were the same for all lenses.

 

1) Focotar – exposure 1/60th of a second

 

 

2) Edixagon – exposure 1/13th of a second

 

 

3) Travelon – exposure 1/25th of a second

 

[OlDoinyo]

Assuming the white balance is the same, neither of these impresses in comparison to your reference lens.

[Alex H]

White balance (WB) was the same, I mentioned it in the original post.

[msubees]

Alex,

 

How does the Focotar compare with Nikkor EL 80mm old style? Thanks.

[Alex H]

Sharper. Slightly better transmission, but that may not be statistically significant.

[Andy Perrin]

The Travelon seems somewhat the better of the two, if I'm interpreting this right? That is, you seem to get the same image as the Edixagon with half the exposure time and a similar spectrum.

[Alex H]

Travelon appear to transmit deeper into UV and has shorter exposure. I may try to shoot with Travelon at F/2 and see how it does, if I have time this weekend.

[Andy Perrin]

Travelon appear to transmit deeper into UV

How are you deducing that from the images? I don't trust my eyes much, since I'm somewhat red/green color-blind, but when I use Photoshop's eyedropper tool set to 11x11 pixel average, and I choose a spot on the highway, I get the following results for the ratio of blue to red:

 

Focotar: 180:171 = 1.05

Edixagon: 254:151 = 1.68

Travelon: 246:154 = 1.60

 

While I'm not sure how the colors are correlated to wavelength (to whatever extent they are...), it certainly seems that the Edixagon and the Travelon are about the same.

[Alex H]

The question is - what is the delimitation between "about the same" and "slightly different"? I wonder what will happen if you try to blur the entire picture for all three lenses and do the same test. Will there be any differences.

 

To my eyes, picture taken with Edixagon is "cooler", has more blue, than the picture taken with Travelon. I have observed same color differences in this test done long ago: http://www.ultravioletphotography.com/content/index.php/topic/1366-quest-for-a-uv-capable-wide-angle-lens/ (first and second page of the post). I also found that such small differences as we see in the shots between Travelon and Edixagon can have considerable impact when editing images further (adjusting white balance and "pushing up" color information) - for example it will be a lot harder to get any false-yellow in post-processing of the image taken with Edixagon without having weird color shifts and increased noise.

[Cadmium]

What would happen if you white balanced the middle one, or the bottom one, and applied that white balance to all three?

[OlDoinyo]

The Edixagon image is bluer because its bandpass is narrower. If you shift display intent to BGR, the color cast becomes orange (indicating primarily longer wavelengths.) It will be hard to coax much chromaticity out of images taken through this lens (after color cast removal.) The Travelon has slightly more bandpass (it transmits a little more of the shorter wavelengths.) Neither is in the league of the Focotar (which is an enlarger lens, no?)

 

To give an analogy, think of taking an ordinary color photo through various grades of yellow, orange, or red filters. The images taken through the deeper filters will have less bandpass and be more monochromatic.

[Andy Perrin]

I understand that, I was proposing that the Edixagon and the Travelon are cutting off around the same place. Alex seemed to be indicating that the Travelon was significantly better, and I just wasn't seeing that. Checking the blue:red ratios was my way of confirming that the bottom two images were not all that different from each other (1.68 vs 1.60). Alex argued, though, that small differences in that ratio could correspond to large differences in chromaticity after white balancing.

[Alex H]

Alex seemed to be indicating that the Travelon was significantly better...

 

And where did I say "significantly better"?

[Andy Perrin]

And where did I say "significantly better"?

That's how I interpreted your remark, "Travelon appear to transmit deeper into UV." If you thought they were about the same, I don't understand the reason for that comment, then. I think we all agree that this judging by color differences is imprecise, so the real way to resolve this is with a sparticle test or something like that which provides a more objective output.

[OlDoinyo]

Look at the vehicles in the center of the frame. In the reference exposure, two of the vehicles are a distinctly different color from the rest. Through the Travelon you can still see this, but it is less apparent. Through the Edixagon you can scarcely see it at all. I would be very surprised if the two lenses had identical transmission curves. But I agree that a sparticle test (or even a pinhole test) would be more objective.

[Alex H]

That's how I interpreted your remark, "Travelon appear to transmit deeper into UV." If you thought they were about the same, I don't understand the reason for that comment, then. I think we all agree that this judging by color differences is imprecise, so the real way to resolve this is with a sparticle test or something like that which provides a more objective output.

 

Your interpretation is incorrect. I said, that "Travelon transmits deeper into UV", nothing more, nothing less. I did not say how much deeper, because I can not estimate this. I did not say how much better, significantly or insignificantly, because it is a subjective interpretation. I never said that they are about the same, because it is also a subjective interpretation. I am sure that two lenses have different transmission curves (based on my own experience of testing many lenses pictorially and comparing my test results with transmission charts made by Klaus), that Travelon transmits deeper into UV and that is exactly what I said. Just deeper, without any quantitative evaluation.

 

The main reason I did not put any comparative statements in the original post is to make everyone come to their own conclusion by themselves. You are free to make your own conclusions based on what you see and your own experience on the subject, but please stop misinterpreting my statements and attributing something to me that never said.

 

I do not see any substantial differences between the test the way I do and the pinhole test.

[Andy Perrin]

Look at the vehicles in the center of the frame. In the reference exposure, two of the vehicles are a distinctly different color from the rest. Through the Travelon you can still see this, but it is less apparent. Through the Edixagon you can scarcely see it at all. I would be very surprised if the two lenses had identical transmission curves. But I agree that a sparticle test (or even a pinhole test) would be more objective.

Yeah, they are not precisely the same — even my eyedropper test had a different ratio. I think this is boiling down to a question of what people's threshold is for "significant difference." I can see what you are saying about the cars, OlDoinyo.

 

Alex, I don't know why you are getting so upset. It is a subjective test to begin with, so yes, I made a subjective interpretation. I also did not put words in your mouth, I just interpreted what you said in a way that went beyond what you intended apparently. It happens. Peace.

[Cadmium]

I think the idea here is that by applying the white balance of the best personally available UV transmitting lens for the white balance of the other two lenses shows a comparison between the UV color palette range differential between these lenses.

Given that idea, then the first lens has the most color range, the third lens has the next best color range, and the second lens has the least.

Which would usually mean that the least amount of color palette range (and the bluest) would have the least UV transmitting depth.

This may go without saying, but even though lens two and three have less color when lens one's white balance is applied to them, they may have a wider color palette if they have their own white balance.

My point in saying this it that if we applied this test to say a Tamrom 21mm, or a Nikkor 18mm, the color palette might not look very good.

So I am just saying this for whoever is out there in the 'balcony' (so to speak) thinking that these white balanced shots are representational of the actual white balance they might expect from each lens individually.

[OlDoinyo]

Perhaps these test swatches will clarify. Top to bottom, frames taken from the reference, Travelon, and Edixagon frames. I have tried to remove color cast as best I could from the latter two. I displayed as BGR because I can see things more easily that way:

 

 

The top frame shows the vehicle colors clearly, and also the crosswalk paint.

The center frame shows the vehicle colors somewhat less clearly, and the crosswalk paint is faint.

The bottom frame could not be decasted completely due to blowing-out of channels, and is very noisy. Vehicle colors show faintly; the crosswalk paint shows only as faint noise-ghosts. It becomes difficult to separate genuine chromaticity from various artifacts, gradients, and noise at this level. I would recommend the Edixagon for b&w UV only (at best.)

 

Color casts, cast gradients and noise aside, a lens with a narrower bandpass will have a narrower palette, simply because it transmits a smaller range of wavelengths and allows less wavelength information to be recorded. There is no way to give the Edixagon the width of palette of a lens with broader bandpass.

[Alex H]

I will try to address all the issues in chronological order.

 

... when I use Photoshop's eyedropper tool set to 11x11 pixel average, and I choose a spot on the highway, I get the following results for the ratio of blue to red:

 

Focotar: 180:171 = 1.05

Edixagon: 254:151 = 1.68

Travelon: 246:154 = 1.60

 

While I'm not sure how the colors are correlated to wavelength (to whatever extent they are...), it certainly seems that the Edixagon and the Travelon are about the same.

 

My version of PSE does not allow to do that, so I can not comment on your results. I already suggested to you to average blur images and measure color composition on those. So here it is (Focotar,Edixagon,Travelon):

 

 

Unfortunately, I am not sure how to interpret these number, just like the numbers you produced. So I am not going to comment, but feel free to use your eyedropper approach and check if you find any difference from your original data.

 

What would happen if you white balanced the middle one, or the bottom one, and applied that white balance to all three?

 

Steve, I am not going to do this, but if you are really interested, I can send you RAW files.

 

The Edixagon image is bluer because its bandpass is narrower. If you shift display intent to BGR, the color cast becomes orange (indicating primarily longer wavelengths.) It will be hard to coax much chromaticity out of images taken through this lens (after color cast removal.) The Travelon has slightly more bandpass (it transmits a little more of the shorter wavelengths.) Neither is in the league of the Focotar (which is an enlarger lens, no?)

 

You are absolutely right, Clark. Focotar (to be specific, this is a Focotar-II 50mm F/4.5 lens) is indeed an enlarger lens optimized for actinic light.

 

Alex, I don't know why you are getting so upset. It is a subjective test to begin with, so yes, I made a subjective interpretation. I also did not put words in your mouth, I just interpreted what you said in a way that went beyond what you intended apparently. It happens. Peace.

 

I am sorry to say this to you, Andy, but you are wrong. I am not upset, and I see nothing in the text that I wrote that can be interpreted as an indication of me being upset. All I was trying to do is to explain myself as clear as possible, to avoid any additional misinterpretation. I have absolutely no problems with you making subjective interpretations. That was in fact the purpose of me posting these pictures online and letting anyone decide for themselves. But this statement of yours is a clear misinterpretation of my words and attribution to me something that I have never said: Alex seemed to be indicating that the Travelon was significantly better, and I just wasn't seeing that.

[Alex H]

This may go without saying, but even though lens two and three have less color when lens one's white balance is applied to them, they may have a wider color palette if they have their own white balance.

 

Steve, in my experience the color palette will still be smaller for the lenses that pass narrower wavelength transmission band even if they are applied individual white balance. Here are the results from same lenses, each with its own individual WB:

 

Focotar

 

 

Edixagon

 

 

Travelon

 

 

The pictures from Edixagon and Travelon may look very similar, and may be "interchangeable" in practice. Unfortunately, this is not the best example as I was not able to include any subjects that reflect in the range below 370nm and are typically represented by false-yellow. Such objects would show much more pronounced differences in color palette in shots taken with Edixagon and Travelon. I have seen it in other lenses I have tested in the past that have similar overall UV transmission to Edixagon and Travelon. I am not going to try to prove this to anyone, it is not such a difficult test to make yourself.

 

So I am just saying this for whoever is out there in the 'balcony' (so to speak) thinking that these white balanced shots are representational of the actual white balance they might expect from each lens individually.

 

Steve, could you please explain what exactly you mean here.

 

Perhaps these test swatches will clarify. Top to bottom, frames taken from the reference, Travelon, and Edixagon frames. I have tried to remove color cast as best I could from the latter two. I displayed as BGR because I can see things more easily that way:

 

 

The top frame shows the vehicle colors clearly, and also the crosswalk paint.

The center frame shows the vehicle colors somewhat less clearly, and the crosswalk paint is faint.

The bottom frame could not be decasted completely due to blowing-out of channels, and is very noisy. Vehicle colors show faintly; the crosswalk paint shows only as faint noise-ghosts. It becomes difficult to separate genuine chromaticity from various artifacts, gradients, and noise at this level.

 

Thanks for this test, Clark. Next time you do such tests with my images, please ask me for larger resolution files, so you can make larger examples. These are rather small.

 

I must say, though, that I have difficulties interpreting your BGR composites, and much more comfortable working with my original files.

[Andy Perrin]

But this statement of yours is a clear misinterpretation of my words and attribution to me something that I have never said: Alex seemed to be indicating that the Travelon was significantly better, and I just wasn't seeing that.

Yes, I misunderstood you on that, but I didn't attribute something to you that you didn't say. The words "seem to be indicating" in that sentence means "this is my interpretation of what was said," not a verbatim repetition of what was said. I was talking about what I thought your meaning was, not your words.

[Alex H]

OK, I understand now. Thanks.

[Andrea B.]

I was just gifted with a Focotar 50/4.5 as a result of Bjørn's Spring Cleaning. I'm eager to get it all kitted up and in use after seeing Alex's results here. Mine is one of the variants prior to the II, but I don't know the construction.

 

A reminder to anyone new to UV photography that comparing false colours to draw conclusions about UV-capability can only be regarded as an extremely rough, imprecise method because of the large number of factors which go into producing UV false colours. I think that such comparisons would best be done by making use of the "actual" raw data as would be seen in apps such as Raw Digger which can display a raw composite without any effects from a particular converter's ideas of white balance, saturation, contrast, sharpening or profiling. Just very tiny shifts or differences in any factors can cause display of anomalous false colours. And it is well known that no two converters convert the same.

[Alex H]

And I would like to point out that even using RawDigger, one will never be able to do proper quantitative comparison on lens transmission. It is still only distinguishes between "more" UV or "less" UV, be it exposure or bandwidth. Nothing more.