UV filters Compared with different Compact fluorescent lights

[dabateman]

I was interested in this test when two forum members were arguing over the different UV global levels and how it would change the output from a UG2a filter. I don't have a UG2a filter, but wondered what the effect of different amounts of UV from controlled light bulbs would have.

 

For this test I imaged the same flower with the same lens settings Pentax UAT at F8. The ISO and shutter speed changed to get a decent exposure for each filter and each light. Each image has been white balanced using the PTFE cube in the shot.

 

A 2mm S8612 filter is used for all the shots stacked with each of the filters.

 

The lights are 365nm LED bulb, Exxoterra UVB 26W compact fluorescent, Black light compact fluorescent, regular 26W white compact fluorescent bulb.

 

[Cadmium]

I think everything make sense with your comparisons above. Those shots are all what I would expect.

Since most of those stacks transmit some mix of UV+Blue+Green light, then I would be shooting with sunlight which would be closest to your White CF.

I have never tried UG5 thicker than 2mm, only .75mm, 1mm, 1.5mm, and 2mm thick, and I would usually use 1.5mm thick because it seems to be the best for such a stack.

 

The UG1 1mm and 2mm are mostly just UV +/- a little bit of visual violet, but not much really.

 

Hoya B-370 works almost identically to BG3 at the same thickness, hard to see any difference.

I use BG3 1.5mm + S8612 1.5mm, but you can use either/or both at 2mm also, but I see some subtle differences with the 1.5mm+1.5mm.

[dabateman]

I am surprised how monochromatic the compact fluorescent black light was, I would have expected more 405nm or 435nm. However, the images all look like the 365nm led.

Now I wished I had tested a hallogen bulb and a white led to get different UV/visible levels.

 

Yes the thickness of the UG5 is a little odd, I got it from a microscopy set. It also included BG12 at 1.5mm, 3mm and 5mm. They too are fun. Too bad the 0.5mm UG1 was broken when I got it. That with the 1mm UG1 would have flanked the UG2a.

[Cadmium]

UG2A 2mm = UG1 .66mm

UG2A 1mm = UG1 .33mm

 

Making filter glass thinner than 1mm gets to be hard to do without breaking it and loosing glass,

and even 1mm thick glass requires a little more care, and can be fragile to use.

 

 

This shows all the various glass types you used above, but these plots don't include the S8612 2mm stacked.

I also added UG2A 1mm, UG2A 2mm, UG5 2mm, and BG3 2mm.

[Cadmium]

Here is a graph comparing all of your stacked versions.

I also added UG2A 1mm, UG2A 2mm, UG5 2mm, and BG3 2mm stacks.

All of these plots use S8612 2mm thick for Red/IR suppression (also shown on graph).

 

[dabateman]

Thanks for the plots. Looks like 2mm S8612 is a good standard to knock down the IR. The ExoTerra UVB bulb has a strong 708nm line. I guess with a 1.5mm S8612 filter, IR would show through using that and the UG5 1mm.

[Cadmium]

I am not sure where you are seeing the strong 708nm line of the Exo Terra.

https://www.ultravio...__fromsearch__1

 

Those stacks using S8612 2mm are all OD4 at 700nm and OD5 at 720nm.

 

S8612 1.5mm works good for UG1, UG11, U-360, U-340, 2mm stacks, resulting in OD5, quite strong suppression for those.

 

By the way, the whole idea of UG5 1mm and U-330 1mm is to include some red, UV+Blue+Green+Red.

(example comparison below)

 

UG5 1mm + S8612 2mm (left), UG5 2mm (or 1.5mm) + S9612 2mm (right)

[Andy Perrin]

Ulf doesn’t feel 1.5mm offers enough protection with 2 mm UG11, I thought. I think it works 95% of the time and that’s good enough for me. YMMV and all that. Can always retake the photo with more blocking, given most of our subjects don’t move.

[Cadmium]

It was only my recommendation. That is up to the individual, you can decide what you need.

You can do a leak test photo to test if you are using enough OD suppression for your situation.

Such a test will be at least as valuable as a graph or spectrometer, and is easy for anyone to do.

For me, I find that anything stronger than OD4 shows no leak for strong outdoor sunshine.

Between OD3 and OD4 is another situation.

If you are seeing any image via a leak test image (using the same exposure time as the UV shot), then get a thicker version of S8612.

S8612 1.5mm thick should be enough for any of those, however if you want to use more nails, get 2mm.

 

Sorry, A little hard to read this graph, best I could do with the close proximity.

[Cadmium]

Separated for easier viewing comparison.

 

 

 

[ulf]

Ulf doesn’t feel 1.5mm offers enough protection with 2 mm UG11, I thought. I think it works 95% of the time and that’s good enough for me. YMMV and all that. Can always retake the photo with more blocking, given most of our subjects don’t move.

If S8612 1.5mm is what you have and it works well for you Andy, that is OK.

I think there is less problems with IR-contamination with UV-landscape images as it might not be that common with very UV-black parts in such motifs.

The NIR-contamination appears first in parts with very low UV-reflectance and at the same time high IR-reflectance. I have mainly seen that in some flowers.

 

I would strongly recommend anyone considering buying a S8612 for general usage to get the 2mm version.

 

The loss in longer exposure time with a 2mm filter is only a few percent while the IR attenuation is more than a magnitude better (10x), compared to the version with 1.5mm thickness.

A stack with a 2mm S8612 filter would almost never ever give problem with IR contamination. Then there is reason to fear having a NIR-contamination.

The only exception is with very light sources with a very strong NIR-component.

 

I can only see two reasons to use a thinner S8612 than 2mm.

1.) If for some reason the total thickness of the filter glass stack become very important, like in a cemented stack where the filter cell ( ring ) cannot handle a 4mm thickness.

2.) If there is a need to be able to state a 3% higher transmission in marketing documentation.

[Andy Perrin]

I have 1.5mm and 1.75mm S8612. Sometimes I use the 1.5mm shooting handheld, usually landscapes. 3% is 3%. I’ll take it.

[dabateman]

I am not sure where you are seeing the strong 708nm line of the Exo Terra.

https://www.ultravio...__fromsearch__1

 

 

Thats what I see. Your link has a weak bump, follow the bump in the purple curve down in the green plot and you see it. In your response with manufacturers plot its the large red bump after 690nm. That is a 708nm mercury line. For me its intense enough or my cameras are more IR sensitive that it can show up and I use it for my 70pnm IR shots.

[Andrea B.]

I would have liked to have seen David's original chart in its raw colours. False colour can be very tricky and misleading. I think Raw Digger is a necessity for the UV-gear enthusiast in order to determine what they have really captured.

 

[After statements like that I always have to add that neither I nor Birna nor UVP have any association with the Raw Digger app except that one of us (me) paid for it and uses it.]