Jump to content
UltravioletPhotography

UV light source shootout


Recommended Posts

Very interesting Johan & nice comparisions. Though only for fluorescing a British £20 note.

I need to read all of this a bit slower tomorrow (my bed time is coming up quick).

Do you have the sources of the lamps purchases & have you tested the wavelengths of the lamps outputs ?

But what does this UV KD lamp head 'drop into', please ?

Cheers

Col

Link to comment

All these are on ebay under various names and labels. These keep changing so rather than have outdated links I don't have direct links - but more info at http://extreme-macro...macro-lighting/

 

I don't have the eq to test wavelengths - this itself was a cheap way of testing it for me. Quite a clear way of showing that the cheap eBay torches might show some UV but there's a lot of visible contamination.

 

I think these dropins let you make your own torches, don't know much about them, hence was asking :D. But 18V is going to make for quite a beefy torch, especially if it really is 365nm as advertised!!!

 

Fwiw, Wratten 2a over the lens on all shots. I thought the shot with the UV converted flash and a Baader filter over the output is quite promising. I put 2 panes of ordinary 2mm float glass between the Baader and the flash so the Baader didn't get cooked. The float glass transmits the nichia torch just fine so it can't be eating significant amounts of UV

Link to comment

Did this series as well because the yellow panel below the orange star is a good indicator. What's interesting is the way that the cover of the BLB-T5 'bank note checker' actually does itself a disservice. The clear plastic cover over the tube it itself excited by UV and gives off blue light, which swamps the already quite thin UV fluorescent green/yellow panel.

 

http://extreme-macro.co.uk/uv-macro-lighting/uv-shootout2.jpg

Link to comment

Johan,

This is a good comparison.

I think if you combine a Schott S8612 or BG39 (on the flash or the lens) with the Kopp 9863 you would get something more similar to your Baader-U result.

- John

Link to comment

Hi Johan

In addition to John's idea, I think a IR blocking filter is needed on the lens, to block any IR leaks. Perhaps a Baader UV/IR block filter, or use a Visible light only camera.

Col

Link to comment

Thank you gents - I should have used the Baader UVIR cut on the flash shots, yes, so I'll do that tonight. Despite its magenta cast, the Kopp result though is actually very very near the 'good' ones, unlike the cheap LED shots the fluorescence isn't drowned out. In photoshop all it really takes is a bit of green addition to make it broadly the same as the 'decent' ones. This would make a certain amount of sense as green is opposite magenta in the colour table so it's like a green barrier filter. And green is of course also very close in colour to the blue S8612/BG39 that John pointed to. I happen to have a green filter that I can try so I'll see what happens with that too, otherwise I'll need to go get the schott blue.

 

The result I was really pleased with for my fluorescence purposes was actually the BLB-T5 result, which is cheap and not blue/violet contaminated. I initially dismissed this light source as too weak but that's because they're 6 inch long tubes so the light isn't focused in any sort of concentrated beam. But, on the flipside, you can buy 4 of these tubes for £10. So if I had to, making two "4 tube" panels for indoor studio use running off AC surely can't be rocket science.

Link to comment

This is better! What do you guys (& gals) make of this?!!

Flash was all 1/180, continuous was 4s.

Non modified K5 with smc-a f/4 macro on min. bellows.

 

bigger image - http://extreme-macro...v-shootout3.jpg

 

http://extreme-macro.co.uk/uv-macro-lighting/uv-shootout3.jpg

 

 

Have I got this right?

 

The BLB and 76LED light sources do not reveal UVIVLFL but instead reveal UVIUVFL?

Bear in mind this is not a converted full spectrum camera, just a standard Pentax K5.

So does that mean that the K5 actually has a poor UV filter which shows some UV?

 

Or is it the case that the top 5 all have some visible light spill and therefore they still show something

even with UVIR cut filter?

Link to comment

Looking good Johan!

However, the purpose of my suggestion of adding a Schott S8612 or BG39 (on the flash -or- the lens) in combination with the KOPP 9863 is to suppress the secondary transmittance peak of the KOPP filter which peaks at ~725nm within the pass band of the UV-IR Cut.

-John

Link to comment

I can see that the Baader UV / IR cut filter is still getting a leak in IR. Most Vis light cameras UV/IR cut filters cut at about 660nm, but the Baader UV / IR is for astro photography that want the Hydrogen alpha emission line in the shot too.

IR leaks are very powerful on silicon sensors.

The Schott S8612 is the best IR suppressor we have.

Col

Link to comment

According to Baader's literature, this UVIR cut transmits from 420-680 nm.

This is the graph:

 

http://www.baader-planetarium.de/sektion/s43b/kundenreferenz/trans-uvircut.gif

This is Kopp 9863:

 

http://www.newportglass.com/kopp_filter_std/Kopp9863_files/image001.gif

 

 

So it's that whole 660nm+ hump that's the IR. Which the unclad schott 8612 should reduce:

 

http://images.nikonians.org/galleries/data/22542/BG_S8612_non_clad_10.jpg

 

 

Is anyone able to explain the bottom two results for me? The images on the right of both the BLB and 76LED (ie unfiltered)

show signal outside 420-680 - this is the IR leak that Schott S8612 would block?

 

Colin, you say you can see that the Baader UV / IR cut filter is still getting a leak in IR. Can you elaborate - what is showing you this?

 

Thx!

Link to comment

Hi Johan

Q / Colin, you say you can see that the Baader UV / IR cut filter is still getting a leak in IR. Can you elaborate - what is showing you this?

A / The photos that have a red / pink glow are IR leaks.

 

Q / According to Baader's literature, this UVIR cut transmits from 420-680 nm.

A / Yes, but not very accurate. There are some graphs that show that it goes close to 700nm & the curve on the bottom right is important.

The line from 700nm is not flat & absolute, there are bumps / rises in transmittance that Baader has chosen to ignore & the end at 1100nm is important too as most silicone sensors are still sensitive here, any IR is very receptive to silicone sensors.

Some fine details & discussion here..... http://dslrmodificat...mod450d14a.html

 

If you place the Baader UV/IR cut, on top of the Kopp9863, there is a significant amount of IR leak around 700nm & beyond.

post-31-0-11700500-1401271807.jpg

 

The graphs you have above are great, but notice that the first Schott graphs have a different vertical scale to the second, this is important, especially the amount of suppression of IR.

A simple flat line is not enough to suppress IR, IR needs to be suppressed deep, not even tiny bumps. That is why the first Schott graph looks different, the bottom line is emphasised. At 700nm the S8612 is getting there, the BG39 is 5 x better at suppressing IR leaks.

I hope this helps you Johan. Just remember IR rules!

Cheers

Col

Link to comment

Johan,

Colin has correctly detailed the issue I stated earlier, the secondary transmittance peak of the KOPP filter, which peaks at ~725nm, is within the pass band of the UV-IR Cut. I shall also opine that alone the UV-IR cut is probably not a sufficient UV barrier filter for clean fluorescence imagery.

- John

Link to comment

Shoot in the dark with a filtered uv source (and a filtered lens) and cease worrying about minor near-ir leakage ??

Are you shooting in the dark?

That's what they taught me to do anyway. :D

 

That kopp filter chart looks exactly like a UG11 ??

 

1100nm is important too as most silicone sensors are still sensitive here

So how could UV illumination be outputting any IR, especially beyond 1100nm. Wouldn't SI become useless there anyway? That's around the place where you need to go to the IN/GA sensors, yes?

Link to comment

If you use a Nichia 365nm UV-LED for inducing fluorescence there will be no Vis or IR leakage. Klaus's Nichia chart is attached below.

Then shoot in the dark with the Nichia and your Baader UV-IR Cut on the lens and you will record Visible fluorescence only.

 

Chart by Dr. Klaus Schmitt, posted on Nikongear.com, 10 Oct 2009.

http://www.pbase.com/kds315/image/87098063/original.jpg

Link to comment

Here is a capture of Visible Orange-Red Fluorescence in Calcite using Nichia 365 UV-LED illumination in the dark and a B+W 092 IR-Pass filter on the taking lens. This filter cuts in around 650nm.

 

This particular Nichia 365 (made by Klaus) has a Baader-U diffuser on it.

 

Photo by Andrea G. Blum, posted on Nikongear.com, 09 Oct 2009.

calciteAndreaBlum.jpg

Link to comment

Let'a make a point here: When I use the Baader UV/IR-Cut filter on my taking lens while recording UV-Induced Visible Fluorescence, then that is because I am usually using my full-spectrum camera (currently a D600) to make the UVIVF shots in conjunction with some standard UV shots.

So, if you are using an ordinary, unmodified camera to record your UVIVF, then you might be able to omit the UV/IR-Cut filter on the taking lens because the cam will have its own internal UV/IR filter in place and that internal filter is, we hope, strong enough to block any UV/IR.

Whew!

Link to comment

Hmmm... given that Andrea, how to explain the last 2 rows of post #10?

 

As in with UV/IR cut filter, the frames are dark - so the reasonable hypothesis must be that the signal without is either UV or IR (or a mix of both?)

Link to comment

What can you tell me about "76 LED" ??

I don't know what this is?

 

*****

 

I'll note that you can push some UV or IR onto any unmodified camera sensor given a long enough exposure.

I don't know specifically about how strong an unmodified K5's internal filter is.

 

Also, you must close your viewfinder and any other potential source of light leaks.

Link to comment

Yes, I have an endcap on, no light leaks. I don't have an IR cut filter yet - but 3 on the way. Hopefully when I get one I can understand a bit better. Fwiw, K5 is known to have a bit of an exaggerated red response so IR rather makes sense. 76 LED is a cheap ebay torch, 390nm - this one.

 

ed - should have mentioned, all these are in dark.

Link to comment

What white balance setting were you using on the K5 when you made the 76 LED shots?

 

It seems like to me that there would have to be significant purple light from that 76 LED torch if it peaks at 390nm and maybe some blue too? If so, you should be able to see at least a faint purple/blue beam from the torch.

Yes/no on this ??

 

And so I was therefore thinking that such a purple & blue leakage could account for the unfiltered magenta photo (right-most column) from the K5 - depending partly, perhaps, on the WB setting?

 

It seems unlikely that there is any IR at play from this torch. And also you have carefully taken all the necessary precautions - shooting in the dark with covered viewfinder - to prevent stray light contamination. [Not that I thought you hadn't, but it is always better to clarify the experimental parameters by asking when they were not explicitly stated.]

 

Side note: I think I might have the name of the shortest wavelength visible colour incorrect. It called 'purple' by some and 'violet' by others. Danged if I can ever remember which is correct - sorry 'bout that!! "-)

 

Well, what do you think? Were the R and B channels tickled a bit by some purple/violet/blue visible light, thus accounting for the appearance of the unfiltered shot?

Link to comment

All the shots were at the same setting - 4250 deg.K. I havent got my head around the proper white balance yet :D I think you mean violet? At school I was taught it goes roygbiv - red orange yellow green blue indigo violet.

 

I hear what you're saying - it certainly sounds plausible! I think when I get a few more filters I'll do the whole lot again with both UV cut and IR cut filters and see what happens. Might err shed some light on the issue. Maybe I made a mistake, I'm wondering if I used Baader-U rather than Baader UVIR.

Link to comment

That kopp filter chart looks exactly like a UG11 ??

 

Yes, KOPP 9863 cross references to UG-11.

 

1100nm is important too as most silicone sensors are still sensitive here

So how could UV illumination be outputting any IR, especially beyond 1100nm. Wouldn't SI become useless there anyway? That's around the place where you need to go to the IN/GA sensors, yes?

 

Yes, the sensitivity of the Si detector on one of my spectroradiometers falls below the sensitivity of the cooled Ge at 1050nm so that is where I switch detectors when scanning across that wavelength range. My InGaAs (indium, gallium, arsenide) spectroradiometer can cover 800-1600nm so there is a more of overlap with Si.

 

- John

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...