Are UV false colors possible even with LED lighting?

[Kai]

The following experiment shows by way of example that false colors are possible even with narrow-band illumination using LEDs and that the transition from yellow to blue is not always at the same wavelength. 

 

Experiment

To investigate, I recorded and compared the spectra of the light sources using a simple method:
LED flashlight, metal slit (1.5mm), 1m distance, DVD grating, Baader-U plus 2mm QB21 (to avoid NIR artefacts), f 50mm enlarger lens, Canon 500 D-FS 
WB in each case on PTFE, identical color profile, which produces yellow and blue false colors with sunlight. 
LED flashlights: peak values at about 375 nm and 405 nm.
Comparison/wavelength calibration: sunlight.
RAW conversion: DxO PhotoLab 3, WB sunlight, color rendering "generic" plus "neutral colors, neutral tone, gamma 2.2" 

 

 

Results 
- Even with narrow-band light sources, yellow tones are shown in the shorter-wave range, blue-violet tones in the longer-wave range. 
- The position of the color transition from yellow to blue is therefore dependent on the light source. 
- The color rendering is slightly different even with identical color profiles if the used spectral range is different.  

 

Additional comment 
The Baader-U cuts off at below 400 nm. Accordingly, only the outermost part of the short-wave flank of the emitted light of the 405 nm LED is shown. Since only very little light is emitted here, the exposure time for the 405nm LED is one second - for the 375nm LED it is only 1 / 350s. 

 

 

[Andy Perrin]

Yes, this was previously known? Although I guess maybe not to everyone! You get LESS false color, but not zero. Usually if you increase saturation you can bring it out.

[Kai]

Here is a picture of dried up plant remains.
The RAW development is neutral.
Without and with color enhancement:
 

 

[Andy Perrin]

The color can be brought out to some degree, but having a wider spectrum light source (and a lens with wider bandpass) is still better because increasing saturation also increases noise, and previously saturated regions may hit the ceiling when saturation is increased further.

[Guest]

Kai, I think the photos look clean. Looks like you found some dried flowers for the yellow. 

 

[Kai]

1 hour ago, Blazer0ne said:

Kai, I think the photos look clean. Looks like you found some dried flowers for the yellow. 

 

 

Better to wait a little longer. The flowers were "freshly dried";)

In the coming days I am expecting a color card with 140 different watercolor samples.
I'm curious whether there are also UV-yellow pigments there ...

 

Otherwise I have to look for pattern books for (yellow) textiles. In street shots, UV-yellow clothes are always noticed. Unfortunately, the intensity is not as high as with colored flowers.

 

I'll stay tuned! 

[colinbm]

8 hours ago, Andy Perrin said:

Yes, this was previously known? Although I guess maybe not to everyone! You get LESS false color, but not zero. Usually if you increase saturation you can bring it out.

I don't understand how a UVA false colour response can move on the wavelength scale, all things being equal, except maybe with White Balance changes ?
OR perhaps with camera changes, where the CFA's response to UVA might be different ?

[Cadmium]

Kai, Nice test! I never get much color at all when using torches/LED's for reflected UV.

[ulf]

2 hours ago, colinbm said:

I don't understand how a UVA false colour response can move on the wavelength scale, all things being equal, except maybe with White Balance changes ?
OR perhaps with camera changes, where the CFA's response to UVA might be different ?

It is the WB that move the colours over the wavelength scale.

 

There has never been a strict relation between wavelength and colours.

That is clearly true when when using different UV-pass filters, but also valid for light sources with wavelengths where there are some reasonable sensor-sensitivity in several of the bayer channels and the lightsource is covering a reasonable range of wavelengths.

 

[colinbm]

28 minutes ago, ulf said:

It is the WB that move the colours over the wavelength scale.

 

There has never been a strict relation between wavelength and colours.

That is clearly true when when using different UV-pass filters, but also valid for light sources with wavelengths where there are some reasonable sensor-sensitivity in several of the bayer channels and the lightsource is covering a reasonable range of wavelengths.

 

Thanks Ulf

[Guest]

I always shoot Sony picture profile standard, but now that I think about it some of the other profiles are geared toward enhancing certain parts of the visual spectrum like skin tones, plant life or landscape. Is it best practice to use neutral profile with UV and IR? Can the picture profile create adverse effects to the tone curve?

[colinbm]

The only thing I trust is a Custom White Balance.

[Guest]

34 minutes ago, colinbm said:

The only thing I trust is a Custom White Balance.

 

With the Sony raw editor called Imaging Edge and within the camera it is called Creative Style. This setting is separate from white balance. So, I can set the white balance and the creative style independent of each other. If I have white balance set to some value that I captured using PTFE, I can still change the creative style to one of several choices and that changes the tone curve (neutral, standard, vivid, landscape, black and white, sepia, portrait, deep, clear, light, sunset, night view, autumn leaves, clear.)  It might add more red or more green or remove contrast overall or add contrast only in a channel with a curve and it's not linear as far as I can tell. I don't know how this would translate to other raw editors like pixel ninja or photo samurai or dark knights or the round table or whatever. Just a question of practice. I assume standard is what Sony intended for normal everyday photos but I am sure it adds something local to what people perceive as better in the visual world. Could that negatively impact UV and IR photos?

[Kai]

5 hours ago, colinbm said:

I don't understand how a UVA false colour response can move on the wavelength scale, all things being equal, except maybe with White Balance changes ?
OR perhaps with camera changes, where the CFA's response to UVA might be different ?

This is a consequence of the WB. When I do the WB on sunlight, the light of the 370nm LED appears of course completely yellow and the (partial) light of the 405 nm LED appears blue-violet. 

[Kai]

5 hours ago, Cadmium said:

Kai, Nice test! I never get much color at all when using torches/LED's for reflected UV.

Thank you Cd ;)

I explain it like this: The colors are created by the different sensitivities of the RGB pixels depending on the incoming wavelength. This results in exactly one RGB intensity combination (= color tone) for (almost - because below approx. 330 nm only red is sensitive) for each wavelength.

In this test, the sensor has very pure wavelengths available through the grating and the slit. Therefore the differentiated and strong colors.

In nature there seems to be practically no material that reflects this selectively. Except maybe structural colors, e.g. the thin interference layers (but wavelength combinations are also often created there = mixed colors).

[Kai]

On 10/27/2021 at 9:08 AM, Blazer0ne said:

 

With the Sony raw editor called Imaging Edge and within the camera it is called Creative Style. This setting is separate from white balance. So, I can set the white balance and the creative style independent of each other. If I have white balance set to some value that I captured using PTFE, I can still change the creative style to one of several choices and that changes the tone curve (neutral, standard, vivid, landscape, black and white, sepia, portrait, deep, clear, light, sunset, night view, autumn leaves, clear.)  It might add more red or more green or remove contrast overall or add contrast only in a channel with a curve and it's not linear as far as I can tell. I don't know how this would translate to other raw editors like pixel ninja or photo samurai or dark knights or the round table or whatever. Just a question of practice. I assume standard is what Sony intended for normal everyday photos but I am sure it adds something local to what people perceive as better in the visual world. Could that negatively impact UV and IR photos?

That's how it looks with me too.

The color profiles change the contrasts and the color tones, not the WB:
(Canon EOS 500D-FS, Baader-U, sunlight; DxO PhotoLab3)

 

A general problem:

With the removement of the hot mirror, the color modes are certainly no longer "standard".
If you want uniform results for material recordings (scientific documentation including contrast, brightness and color), it would be best to create a "UV color checker" as reference. You could then assign false color values to this as reference values and then use them to create individual ICC profiles for the individual light source-filter-lens-sensor combinations. This should make it possible to achieve the same color results with different equipment.
Unfortunately there is no such reference color chart yet ... 

[colinbm]

Different cameras & CFA may vary slightly from this.

[Guest]

Actually that is where the lodge is located on the left. On the right is the end of the beginner ski slope where you take the lift back to the top. 

[Kai]

2 hours ago, colinbm said:

Different cameras & CFA may vary slightly from this.

Hello Colin,
Such a chart shows the sensitivity of the sensor.
It would be exciting to see the left part of the curve with a higher resolution and into deeper UV.
I made a few experiments to this some time ago.
I can post the results as a seperate post afterwards.

[Andy Perrin]

Kai, Jonathan did this already if you do a search. I’m on my phone so it’s hard for me to find it for you at the moment. Obviously independent measurements are good but you might want to know how he did it. 

[Kai]

3 hours ago, Andy Perrin said:

Kai, Jonathan did this already if you do a search. I’m on my phone so it’s hard for me to find it for you at the moment. Obviously independent measurements are good but you might want to know how he did it. 

Hello Andy,
Thanks for the hint!
Unfortunately I found nothing suitable. JMC also has 1400+ posts ;)

Since I've already done my investigation and Canon equipment may also give slightly different results, a certain redundancy of the investigation may not be bad at all...
 

[Andy Perrin]

22 minutes ago, Kai said:

Hello Andy,
Thanks for the hint!
Unfortunately I found nothing suitable. JMC also has 1400+ posts ;)

Since I've already done my investigation and Canon equipment may also give slightly different results, a certain redundancy of the investigation may not be bad at all...
 

Here's the thread. It does help that I knew what I was looking for specifically:

https://www.ultravioletphotography.com/content/index.php?/topic/2813-camera-sensitivity-chasing-ghosts-in-spectral-sensitivity-measurements/

 

He tested a Eos 5DSR. His first set of measurements had an issue that was corrected down-thead I think.

[dabateman]

16 minutes ago, Kai said:

Hello Andy,
Thanks for the hint!
Unfortunately I found nothing suitable. JMC also has 1400+ posts ;)

Since I've already done my investigation and Canon equipment may also give slightly different results, a certain redundancy of the investigation may not be bad at all...
 

These sites may be what your after:

https://jmcscientificconsulting.com/imaging/project-mirrorless-part-2-the-camera-conversion/

 

https://www.ultravioletphotography.com/content/index.php?/topic/3608-sony-a7iii-conversion-to-multispectral/&do=findComment&comment=31603

 

https://www.ultravioletphotography.com/content/index.php?/topic/3730-multispectral-monochrome-nikon-d850-testing-and-shakedown/&do=findComment&comment=33297

 

https://www.ultravioletphotography.com/content/index.php?/topic/2580-build-thread-at-home-measurement-of-camera-uv-spectral-response/

 

[Kai]

@ dabateman and Andy:
Thank you for that information! JMC presents impressive work here !!!
Such a monochromator is certainly a fine thing :)
My simple attempts do not contradict this either, they are perhaps more oriented towards practical photography.
I usually take photos in sunlight through a Baader-U or filter with similar characteristics.
But see for yourself ...

 

 

[microbat52]

Gah i cant get ANY false colour in UV with my setup :/