• Ultraviolet Photography
  •  

A Diffusion filter to remove UV hotspots

Filters UV Lighting
19 replies to this topic

#1 dabateman

    Da Bateman

  • Members+G
  • 2,044 posts
  • Location: Maryland

Posted 23 March 2019 - 11:42

In my test packs, I saw this #750 Lee filter, the Durham Frost filter:

Attached Image: DurhamFrost.jpg

It works very well to remove hots spots from lights and works well for UV-A lights as well.
Attached Image: DurhamFrost_Filter.jpg

Image of African violet with 365nm trans iluminated without #750 filter
Attached Image: DurhamFrost_Flower_Nofilter.jpg

Image of African violet with 365nm trans iluminated with #750 filter
Attached Image: DurhamFrost_Flower_filter.jpg
Image of African violet with 365nm trans iluminated without #750 filter

#2 Mark

    Member

  • Members(+)
  • 566 posts
  • Location: Massachusetts, USA

Posted 23 March 2019 - 12:22

@dabateman: In your examples above the non-diffused image is exposed more than double the duration as the diffused image. It would be good to see a direct comparison with equivalent exposures. I mention this because I have also tried to diffuse UV, using a UV quartz diffuser (post). I expected it to have have excellent transmission, alas it significantly reduced UV transmission to a degree I did not find acceptable. In fact, the filter you tested above has a greater transmission profile than the quartz diffuser I tried (at least for λ above ~325 nm). I wonder just how effective the Lee filter diffusion is, given it is very thin. If it is effective, with high UV transmission, this is a great find.

#3 dabateman

    Da Bateman

  • Members+G
  • 2,044 posts
  • Location: Maryland

Posted 23 March 2019 - 14:50

Yes it doesn't cut much above 340nm

Here is No filter over single 365nm Light, F11, ISO 400, 10 seconds, Baader Venus U on Camera:
Attached Image: NoFilter_400_10sec_F11.jpg


Here is #750 over single 365nm light, F11, ISO400 10 seconds, Baader Venus U on Camera:
Attached Image: Filter_400_10sec_F11.jpg

Edited by dabateman, 23 March 2019 - 14:55.


#4 Mark

    Member

  • Members(+)
  • 566 posts
  • Location: Massachusetts, USA

Posted 23 March 2019 - 22:33

I'd say that #750 produces a remarkable improvement. I have some Roscolux filters which include some diffusers, although none are listed as equivalent to the #750. I'll give some of them a try and see if I can find similar results. I hope one of them works - I love when low-tech solutions work better! Except in my case... I already spent the money to buy a UV quartz diffuser :(

#5 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 24 March 2019 - 00:37

Yep, I tried it here also, works great. Even at close proximity, so it could even be installed in front of the torch filter if wanted.
Definately removes the hot spot. Great find once again!
This place has the 21" x 20" sheet for $7.25 (available somewhere in roles too I suppose).
https://www.pnta.com...ASABEgKdevD_BwE

Edited by Cadmium, 24 March 2019 - 00:38.


#6 UlfW

    Ulf W

  • Members+G
  • 1,172 posts
  • Location: Sweden, Malmö

Posted 24 March 2019 - 07:14

View PostMark, on 23 March 2019 - 22:33, said:

I'd say that #750 produces a remarkable improvement.

I must be blind for the improvement.
When I look at the flower only, it is difficult to tell the difference.
Outside the flower the dust and reflections in the material the flower is resting on is masked, but I cannot see anything else.

Please help me find the improvement in imaging the flower.

Edited by UlfW, 24 March 2019 - 10:05.

Ulf Wilhelmson
Curious and trying to see the invisible.

#7 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 24 March 2019 - 07:33

Ulf, I understand what you mean. I am not sure I understand the photos above.
However, I tested it, and the 750 defiantly diffuses the hot spot, and with no real apparent dimming in overall light.
A year or two ago I made some U-340 2mm thick, I tried two things, sand blasting and frosting, and both dimmed the light drastically.
This stuff works better than anything I have seen.
What is kind of amazing to me is that it transmits as low into UV as it does.

For the $7.50 price, I think it is more than worth a try.
I thing he has really found something here that will be useful.

Edited by Cadmium, 24 March 2019 - 07:35.


#8 dabateman

    Da Bateman

  • Members+G
  • 2,044 posts
  • Location: Maryland

Posted 24 March 2019 - 12:25

Ulf,
I don't think Mark was referring to my image. But an improvement compared to his expensive frosted quartz. Others I have tried cut the light by at least half.

This diffuser seems amazing to transmit as much UV as it does. I was able to see at 335nm, but didn't look quantitatively. So can't say how much loss I was getting. At 365nm, where most Uva images are taken, there seems to be really no loss of light.

#9 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 24 March 2019 - 19:35

Yep, that would be the word I would use, amazing.
It is true, any UV glass with surface diffusion I have made cuts the light a lot.
This doesn't cut the light, it just removes the hot spot, first time I have ever seen that.
Sorry, guess I already said that. ;-) But I am so impressed.

Did you see the reflector samples? Those look like that might be really handy for making flash and lighting reflectors in different situations.
Anyone who doesn't have one of these swatch sets, you should get one, they cost very little, get the numeric one, it contains all filters, some swatch kits don't.

Edited by Cadmium, 24 March 2019 - 19:37.


#10 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 25 March 2019 - 02:48

David, Did you test any of the other Lee Diffusion filters?

#11 dabateman

    Da Bateman

  • Members+G
  • 2,044 posts
  • Location: Maryland

Posted 25 March 2019 - 04:26

Cadmium,
Not yet. I got the diffusion edition swatch pack used off ebay which has a lot of diffusion filters in it and size is 76mm x 90mm. So a lot to play with. Most of these do not have any spectra with them. But quite a few look to be various weaved fabrics.
I also ordered direct from Lee the numerical edition and designer's edition. These are 38mm x 89mm, but no quick page reference guide. I am still playing with them.

Edited by dabateman, 25 March 2019 - 04:56.


#12 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 25 March 2019 - 05:57

Another one I plan to try is #116, looked at the swatch graph today, looks close to #729, with slightly more 500nm peak amplitude, which might translate into slightly better exposure time,
but hard to see if the colors will be the same.

Looks Like David Twede may have used it also, he recently posted the graph for both the #729 and the #116.
Thanks again for originally sharing his link here. :)
https://next-eyes.bl...tal-ir.html?m=1

#13 dabateman

    Da Bateman

  • Members+G
  • 2,044 posts
  • Location: Maryland

Posted 25 March 2019 - 06:35

In the first Lee post I posted an image using the #116 filter. I didn't see that David Twede also had that one in mind. Interesting. I actually liked that filter straight out of camers more than the 729. Its not red, but more orangy

#14 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 25 March 2019 - 08:03

Yeah, I forgot about that, I will have to go look, one of them was a very slight amount more red, kind of?
I tried some other close ones today, but they didn't quite look the same, very close though.
Swatches are almost too small for me to test, so I just ordered a sheet.
Yes, he recently post a pic of both graphs.

#15 Mark

    Member

  • Members(+)
  • 566 posts
  • Location: Massachusetts, USA

Posted 27 March 2019 - 23:54

@dabateman: So, I tried some of the Roscolux filters I have (from an old swatch booklet, with some filters missing from old projects). To see if any yield results comparable to your Lee #759 I set up a test scene with the UV source (MTE-U303 + U340) at a fixed position & distance. In addition to my UV quartz diffuser, I tried these 25 candidates:

#101, #102, #103, #104, #105, #106, #111, #112, #113, #114, #115, #116, #117, #119, #120, #121, #124, #125, #126, #127, #132, #140, #160, #162, #163

Out of all those hopefuls, most either cut too much UV or just didn't diffuse much at all; or both. There were only two which showed both decent UV transmission and appreciable diffusion. Those were the #101 and the #163, each of which were very nearly comparable, with some negligible trade-off between diffusion and transmission (results shown below).



While neither diffuses as well as the UV quartz, they are significantly brighter, cheaper (!), and appear to diffuse just well enough to ameliorate the hot spot. I'll put one of these to a real test during my next shoot.

Thanks for the idea on this!

#16 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 28 March 2019 - 02:18

Mark, How about if I send you a piece of Lee 750 so you can compare it directly?
I got extra. Be happy to send you one.

Edited by Cadmium, 28 March 2019 - 02:19.


#17 Mark

    Member

  • Members(+)
  • 566 posts
  • Location: Massachusetts, USA

Posted 30 March 2019 - 12:17

@Cadmium: I appreciate the offer, but I think the filter options I have will work quite well enough for my purposes. This has also got me thinking that this worked (these optical lighting filters) probably due in part to the filters' thinness, especially compared to my quartz diffuser. So I'm going to change up my next test... I'm going to see how a sanded/roughened very thin piece of common float glass compares.

#18 Cadmium

    Member

  • Members+G
  • 3,767 posts

Posted 30 March 2019 - 20:19

Just thought there might be some difference between the 750 and the 101 and 163. I should look to see if those are in the Lee book, if so compare... will do.

#19 Mark

    Member

  • Members(+)
  • 566 posts
  • Location: Massachusetts, USA

Posted 31 March 2019 - 02:24

I got to make a couple homemade diffusers today, using some sandpaper, music, and elbow grease. One is made from the clear glass removed from an MTE-U303 torch; the other is made from a piece of Schott type WG 295 nm cut-on glass. Here's a UV image of them before (top) and after (bottom) modifications (the UV quartz diffuser was not modified).

(UV source: 4 x fluorescent BLB; lense filter: AndreaU-MkII)

And then a test of the diffusers, using this highly technical, precise test scene...


And from these results I conclude exactly what I've suspected almost from the start, that is, an inverse relationship between diffusion and UV intensity. Simply, more diffusion = less UV. Of course, this could be skewed with more expensive equipment (e.g., UV grade holographic diffusers), but keeping things practical (i.e., affordable) in my hands this relationship seems to hold true.

(UV source: 1 x MTE-U303+U340; lense filter: AndreaU-MkII)

This isn't necessarily a bad thing. Sure, ideally there would be a filter which has 100% diffusion with 100% transmission to evenly light up my scenes and get rid of hot spots - but even that is in part misleading, as it wouldn't be possible to yield the equivalent intensity over a larger scene without increasing the supply power accordingly. The good news is, now I have options from which I can pick either more diffusion or more transmission, as needed for any given shot.

Cheers.

#20 dabateman

    Da Bateman

  • Members+G
  • 2,044 posts
  • Location: Maryland

Posted 31 March 2019 - 02:48

I realky like the 3x2 grid of images. Perfect and in what looks like rank order of diffusion. I think your really getting how I like to see things. Perfect.

I agree, But the #750 is a magical diffuser. May still be the best for you. It might be a woven fabric, which diffuses, yet passes light. I think that is its magical trick. Rather than being a sanded uniform surface.

Also, your highly technical construction erraser puzzle dude is really cool. The 163 seems to be the winner.

Edited by dabateman, 31 March 2019 - 02:53.