are these UV pictures?

[photoni]

5 hours ago, Kai said:

Good pictures Ming!

From my point of view everything is fine as far as the color is concerned.
If the IR attenuation is too weak, objects containing chlorophyll could lead to pale green IR overexposure. But I can't see that in your pictures.

I'll give one thing to consider: What is the right brightness?
If you put a PTFE or a matt aluminum plate in the picture and adjust the brightness of the VIS image and the UV image (i.e. same brightness of the PTFE/aluminium plate), your UV pictures with vegetation will appear much darker. This has to be the case, since the green parts of the plants use the light for photosynthesis, i.e. absorb it. Many photos then appear extremely "low key". Since then of course you can no longer distinguish many details and therefore some forum members brighten their pictures. The images then look similar to the familiar world of VIS images.
I then only find the comparison with VIS images pointless...

 

@Kai  yours is an intelligent philosophical speculation.

But it should take into account the "questionable" sensitivity of the sensor with RGB filters

that respond strangely to wavelengths for which they are not made.

Maybe I'm an Ortho BW photographic film is truer

[Ming]

7 hours ago, Kai said:

Good pictures Ming!

From my point of view everything is fine as far as the color is concerned.
If the IR attenuation is too weak, objects containing chlorophyll could lead to pale green IR overexposure. But I can't see that in your pictures.

I'll give one thing to consider: What is the right brightness?
If you put a PTFE or a matt aluminum plate in the picture and adjust the brightness of the VIS image and the UV image (i.e. same brightness of the PTFE/aluminium plate), your UV pictures with vegetation will appear much darker. This has to be the case, since the green parts of the plants use the light for photosynthesis, i.e. absorb it. Many photos then appear extremely "low key". Since then of course you can no longer distinguish many details and therefore some forum members brighten their pictures. The images then look similar to the familiar world of VIS images.
I then only find the comparison with VIS images pointless...

Kai, I don't have an opinion on the brightness of UV versus VIS images yet.  I am still learning the strangeness of UV.  Actually, I meant to ask you a question.  I read your post showing images by Canon EF 28mm f/2.8, so I bought one on eBay.  This afternoon I tested it with ZWB2 + BG39, visible bandpass, and a bunch of LifePixel IR pass filters.  I did choose to overexpose the UV images by 1 stop because they seemed a little dark in the viewfinder when compared to VIS and IR images.  Okay, here is my question for you: What 20mm lens do you use for UV?  Also, I read your post about Laowa 15mm f/4 macro.  What do you use for 14mm and 16mm, if any?  Thank you.

 

[Kai]

13 hours ago, Ming said:

Kai, I don't have an opinion on the brightness of UV versus VIS images yet.  I am still learning the strangeness of UV.  Actually, I meant to ask you a question.  I read your post showing images by Canon EF 28mm f/2.8, so I bought one on eBay.  This afternoon I tested it with ZWB2 + BG39, visible bandpass, and a bunch of LifePixel IR pass filters.  I did choose to overexpose the UV images by 1 stop because they seemed a little dark in the viewfinder when compared to VIS and IR images.  Okay, here is my question for you: What 20mm lens do you use for UV?  Also, I read your post about Laowa 15mm f/4 macro.  What do you use for 14mm and 16mm, if any?  Thank you.

 

Many questions... :)
So first of all, my personal recommendation for the exposure: Do what you like! If you're primarily an artist, adjust the brightness visually. Use a brightness reference if you are more scientifically interested. I can't make up my mind and I do this and that. But I know what and why I'm doing it.

Congratulations on the 28! It has become my go-to UV lens when traveling. I can comfortably work with it by hand when the sun is shining. I also like that I can quickly take simple comparison shots in the VIS with my smartphone (approximately the same shooting angle).
Your 40 lets a little more shorter-wavelength UV through, but you can't see much of that in common pictures. For plant shots (details, botany) I would recommend a longer focal length with better UV transmission up to 330 or 340 nm.

In the wide-angle range, the 15mm Laowa (or the Opteka replica) and the 28mm are enough for me. The pictures are so sharp and rich in detail that I choose sections if necessary. I do panoramas for larger angles of view, but with very large angles of view I like the cylinder perspective better than the central perspective anyway.

I hope this helps you!
Good light and lots of pictures!!! :)

 

PS:
I also like to record with the 90 TS-E (old version). Limited UV range but very sharp and of course the depth of field extension through the Scheimpflug effect :)

 

P.P.S.:
My EF 1:4 16-35 IS USM and unfortunately also my EF 1:2.0 35 mm IS USM are only transparent up to approx. 380 nm and are therefore not really suitable for UV.

[Andy Perrin]

Well, what this scientist was trying to tell you is that both the visible and UV pics are nonlinearly curved, and as Photoni said, the Bayer dyes add additional nonlinearity. So a brightness standard isn’t going to help much unless you have a mono sensor and process the RAWs with gamma=1. If you do that, you will get a few bright areas and a lot of muddy darkness in your UV photos. Conceivably you could do the matching up that way and then curve both pictures with the same gamma afterwards so they would be nonlinear but equivalent brightnesses would still correspond. 

[ulf]

One more take on this, I think, is that all digital images come from RAW-data captured by the image sensor. They are always very flat, both VIS and UV-images.

When processed to get good VIS-images, in camera, or by some dedicated software,  nonlinearity and other kinds of colour-, sharpness-, and contrast-, black-point  enhancement are applied.

For normal VIS-images the real world is a good reference and most processing software are well tuned for this, generating good images.

 

For UV-images and modified cameras that is not true anymore and you have to do all work yourself to get good images.

That is best done by processing RAW-files.

 

In the UV-world reflectance from most objects are quite low compared to the sky or illuminated PTFE.

 

The best way to handle this is to take a reference shot for WB of PTFE and then remove the PTFE-target and expose after whatever you think is interesting in your image.

If you expose after the sky everything else will be dark and dull. If you allow the sky to be over exposed, with a longer exposure time you will get a better exposed motif, with less noise.

 

Then the resulting image files can be tweaked to taste.

 

 

[Andrea B.]

I'll give one thing to consider: What is the right brightness?

 

There really is not a correct answer to that.

 

First we must stipulate that the only accurate, repeatable and precise method for measuring the UV-absorption/reflection of a subject is to use proper spectrometric gear.

If, however, one wants to estimate UV-absorption/reflection of a subject by making a reflected UV photograph, then a series of calibrated reflective standards (which are stable under UV illumination) can be included in the photograph. The degree of UV-absorption/reflection is then estimated by determining the standard which has the closest match under the particular UV illumination used. (Of course, the UV illumination adds another level of difficulty to making a photographic estimate.)

 

But to be able to see the photograph of the subject together with the standards, the photo file must be demosaiced and the usual brightness curves applied.

• The brightness curves are meant to match the way human eyes see. But human eyes cannot see/detect most UV light. So right there is the primary difficulty in defining brightness for a reflected UV photo. What does brightness mean for something we cannot see?
• Then one must ask whether the demosaicing and curve application alter the outcome of the estimate? I think it probably does not. But I do not have any direct evidence because I don't have spectrometric gear to use to test the photographic estimate. 
• IIRC our UVP member Jonathan has done that test. I'm going to go look for that LINK.

So what is my point in discussing all this? Well, here on UVP we are making photographs to see what the world might look like in reflected UV light. We make the reflected UV photo which (we hope) is not over-exposed or under-exposed. And our photo file is demosaiced and curved using the usual visible photo processing tools. And we must accept that the UV tones -- white and black and 100 shades of gray -- in our processed photo are relative (to one another) and do not depict any absolute UV-absorption/reflection. It is OK to make the photograph look good.

 

Counter-arguments, discussion, corrections all welcomed.

 

LINK: https://www.ultravioletphotography.com/content/index.php?/topic/819-a-calibrated-gray-scale-for-forensic-ultraviolet-photography/

LINK: https://www.ultravioletphotography.com/content/index.php?/topic/2607-question-about-raw-retaining-linear-data-for-image-analysis/

 

[Andy Perrin]

I'm 100% with you on this, Andrea. At best we can hope to get relative values within the same photo. Comparing between photos in different parts of the spectrum would need something like what you described, essentially a set of reflectance standards in both photos that are designed for UV and visible respectively. Jonathan did something like that once, in fact. He used alabaster, I think?