UV fire flame (UV green ??)

[Marco Lagemaat]

I was trigged by this post where Dabateman stated that it is possible to get really deep UV from fire (-works) ;

 

 

So I took my butane gas stove and did a small experiments;

Now I am wondering what kind of green colour the top of the flame is. Is it UV green or normal visible green? Anybody knows? How to measure? 

 

 

First a VIS photo of the flame, there is no green visible. 

 

 

 

 

Here (below) you see a photo with Nikon D70s, Prinz galaxy 35mm 3.5, ISO400, 13sec, f/3.5, ZWB3, BG39. WB on PTFE see background.

 

 

 

 

Nikon D70s, Nikkor 50mm 1.8D, ISO1600, 13sec, f/9, ZWB3, BG39. WB on PTFE see background.

 

 

 

Nikon D70s, Soligor 135mm 3.5, ISO400, 13sec, f/3.5, ZWB3, BG39. WB on PTFE see background.

 

 

 

[Stefano]

Nice experiment, now we know another source for UV-green. It's a rare color.

[Marco Lagemaat]

2 minutes ago, Stefano said:

Nice experiment, now we know another source for UV-green. It's a rare color.

But! the question is still..... it UV green? and how do we (I) know?

[Stefano]

If your white balance is correctly set, you can be quite sure it is actually UV-green. If you have a 365 nm torch, it should look yellow. If that looks green, you may have a problem.

 

Do you have some polycarbonate? If you are seeing UV, it should also block the UV-green from the flame.

 

The best tool would be a spectrometer, but I don't think you have one.

 

 

[Marco Lagemaat]

21 minutes ago, Stefano said:

If your white balance is correctly set, you can be quite sure it is actually UV-green. If you have a 365 nm torch, it should look yellow. If that looks green, you may have a problem.

 

Do you have some polycarbonate? If you are seeing UV, it should also block the UV-green from the flame.

 

The best tool would be a spectrometer, but I don't think you have one.

 

 

Good one! I’ll try tomorrow the polycarbonate. 
in the background you see the PTFE I used for WB. Looks fine to me. 

[Fandyus]

You're using ZWB3 which leaks quite a lot of visible and infrared, even red. I think whatever you're seeing there is probably not ~340nm

[Andy Perrin]

There are a number of major issues here. 

 

1) As Fandyus said, leakage is likely to be a factor here. When the light source is not sunlight, the ratio of UV to visible+IR may be shifted either way, and if it's towards the vis+IR direction, leakage would probably dominate. We don't actually know for sure which way this is going here, it may be okay, or it may be mostly or partly leakage.

 

2) The white balance cannot be interpreted in a straightfoward way as indicating particular wavelengths, and this is a fantastic demo of that fact. When we white balance under sunshine and get certain colors, those colors come partly from the wavelengths present in the sunshine. If certain wavelengths are missing altogether, they won't get reflected from the PTFE, and therefore the white balance would not adjust for them. You cannot use the light from the flame to white balance with here if you want "green" to mean the same thing it does in our normal sunshine-illuminated photos. Instead you should do the white balance under actual sun off the PTFE, then take the flame photo using that previously made solar white balance.

 

3) This issue of missing wavelengths is especially important with a gas flame because when gases burn they primarily make a line spectrum which is missing a lot of wavelengths and consists of large peaks in other places. In fact, for natural gas, which is mostly methane, the spectrum looks like this:

 

 

https://www.researchgate.net/figure/Flame-emission-spectra-measured-from-laminar-methane-air-flames-at-a-U-14-08-b-U-14_fig5_233373435

 

Since it's unlikely to the be the UVB spike (which has been imaged on the forum before by Jonathan!), you are probably seeing leakage from visible or IR.

 

[dabateman]

I think you are seeing visible and IR. 

Zwb3 will also push a dark UV signature to look green.

 

This is Jonathan's blow torch test

https://www.ultravioletphotography.com/content/index.php?/topic/3221-ir-visible-and-uv-of-small-blow-torch-flame/&hl=flame&fromsearch=1

 

[Cadmium]

This is UV+Blue/Green, AKA 'bee vision' stack. You are mixing UV + visible blue and green.
There is no IR, the BG39 blocks IR.
 

[Fandyus]

1 hour ago, Cadmium said:

This is UV+Blue/Green, AKA 'bee vision' stack. You are mixing UV + visible blue and green.
There is no IR, the BG39 blocks IR.
 

Yes, but the stack will also leak red/orange. I've tried it.

[ulf]

1 hour ago, Cadmium said:

This is UV+Blue/Green, AKA 'bee vision' stack. You are mixing UV + visible blue and green.
There is no IR, the BG39 blocks IR.
 

Correct, but the very low sensitivity of the sensor in deeper UV-A, (UV-green), compared to the blue and green sensitivity will make the VIS dominant, especially if the emission spectra is anywhere near the ones showed by Andy above.

It his highly unlikely that you see any UV-green here.

[Andy Perrin]

All the hydrocarbon gases make roughly similar spectra because the chemistry is almost the same. You get a UV-B OH line, and some others in visible and IR. Not much in UV-A usually.  Compare to the spectrum Jonathan got from butane in his thread. 

[Cadmium]

23 hours ago, ulf said:

Correct, but the very low sensitivity of the sensor in deeper UV-A, (UV-green), compared to the blue and green sensitivity will make the VIS dominant, especially if the emission spectra is anywhere near the ones showed by Andy above.

It his highly unlikely that you see any UV-green here.

Yeah, But the UV-A mixed with the Visual Blue could create green...  various combinations.

You might find out more if you compare to a strictly UV pass filter shot.