• Ultraviolet Photography
  •  

Transmission of ZWB1 (2mm), ZWB3 (1.5mm) and BG39 (2.5mm)

10 replies to this topic

#1 JMC

    Member

  • Members
  • 128 posts
  • Location: UK

Posted 12 August 2017 - 08:35

I recently bought a few filters from Igoriginals on eBay and wanted to share the transmission data I measured for them. The filters were ZWB1 (2mm), ZWB3 (1.5mm) and BG39 (2.5mm). All in 52mm filter thread and priced at 25USD each. I ran them on a Perkin Elmer Lambda 650S UV-Vis spectrometer (150mm integrating sphere) to measure transmission between 250nm and 800nm (1s collection time per nm). The graphs are shown below.
Attached Image: ZWB1 3 BG39 graphs.jpg

Edited by JMC, 12 August 2017 - 08:36.


#2 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,094 posts
  • Location: USA

Posted 12 August 2017 - 16:43

Thank you, Jonathan, for these transmission charts!

I'll be adding them later to the pinned Filter Transmission Charts.

Stacking with the BG39 does cut transmission and bandwidth considerably for the ZWB1 !! You might enjoy some S8612 blockers?

I don't see any visible green leak for the ZWB1 (2mm) as is evident in some other charts we've seen for thinner ZWB1. Good! Also note that thin Hoya U-340 leaks some green. It is just seems to be the nature of thin 340-peak glass? Go thick as possible with ZWB1 or U-340.
EDIT: The adjective 'green' should not have been there.

I was wondering if you could comment on the quality of your ZWB filter glass?? (Thanks in advance.) As noted in the Filter link above, a bubble or two is always acceptable in glass filters or even in lenses. But any pits or striations or other flaws are not good.
Andrea G. Blum
Often found hanging out with flowers & bees.

#3 Øivind Tøien

    Member

  • Members
  • 16 posts
  • Location: Fairbanks, Alaska

Posted 13 August 2017 - 00:15

A 2mm thick ZWB1 (20mm diameter) is what I bought from ebay seller mei-2014 for my Tank 007 TK-566 UV LED light. The result above is definitely better than the spectrum provided in the auction with respect to IR leakage:
Posted Image

Ball bearing test, first the visible light shot (flourescent tube lighting) with my D40x IR-720, f/8 20s ISO 100:

Posted Image


Then the ZBW1 on the Tank 007 UV flashlight, D40x IR-720nm, f/8, 20s, ISO 100, flashlight held at 7-10 cm distance.
There is hardly any IR leakage detectable:

Posted Image

Edited by Øivind Tøien, 13 August 2017 - 02:21.

Øivind

#4 JMC

    Member

  • Members
  • 128 posts
  • Location: UK

Posted 13 August 2017 - 08:35

None of the filters had air bubbles or inclusions in the glass Andrea. the ZWB1, 3 and one of the BG39 filters looked flat with no visible distortion in the glass. However the second of the BG39 filters did exhibit a slight striation/waviness towards one side of the filter which covered about 2-3mm of the surface (easier to see with a diagram).
Attached Image: BG39 ripples.jpg
I don't believe this will be a problem for me as it'll be used on a 85mm Asahi lens, so the edge of the filter wont be seen, however I must admit if you hadn't asked for the info Andrea, I wouldn't have noticed it.

Yes at some point S8612 filters will be in order, however this will have to make do for now, given they were about a quarter of the cost.

Oivind - that graph appears in various places online and I think may well be for a 1mm thick ZWB1. Also the ZWB3 in that graph is for 2mm, hence I see relatively more IR in mine with it only being 1.5mm. If that's true then it's not surprising that the relative IR transmission through the ZWB1 2mm is that much lower in relation to the UV.

#5 Daryll

    Member

  • Members
  • 12 posts
  • Location: petersfield . UK

Posted 13 August 2017 - 16:03

Do I need the 2mm ZBW1 for the front of my convoy S2 UV torch Claims to be 365 nm , this is for UVIVF photography

Daryll

#6 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,094 posts
  • Location: USA

Posted 13 August 2017 - 16:36

In rigorous fluorescence photography a filter is always placed over the illumination source to ensure no leaks outside the desired range and over the taking lens to ensure no stray contaminating light hits the sensor. (And UVIVF fotos are made in the dark.)

The general contaminant in 365nm UV torches is a violet/blue visible leak which can pass through the UV/IR blocking filter on your lens. You can test for this with the Shiny Metal test, q.v. in <> Sticky :: UV-Induced Visible Fluorescence <>

How much does a violet/blue visible leak contaminate a UVIVF photo? Hard to say unless you know the fluorescence peak of the subject being photographed. If, say, the subject fluoresces red under UV, then violet/blue leak might interact with the red and cause a more magenta tone. If the subject has, say, no fluorescence and the violet/blue leak is significant, then you might get an appearance of blue fluorescence when there actually is none.

Given that the ZWB1 peaks around 340 nm, you would probably be best served by using a U-360 type filter which peaks nearer to 365 nm. The ZWB1 will cut too much of the left-hand slope of your torch's transmission curve.
Andrea G. Blum
Often found hanging out with flowers & bees.

#7 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,094 posts
  • Location: USA

Posted 13 August 2017 - 16:41

The second of the BG39 filters did exhibit a slight striation/waviness towards one side of the filter which covered about 2-3mm of the surface.

Why am I not surprised? :D This stuff is cheap for a reason.

Test whether this waviness causes problems by shooting a sheet of newspaper taped to the wall. Camera on tripod, lined up square to the wall. Do you see any distortions in the resulting photograph? To confirm, rotate the filter slightly and reshoot. Did the distortions move with the filter rotation? Please let us know how this test turns out.
Andrea G. Blum
Often found hanging out with flowers & bees.

#8 Cadmium

    Member

  • Members(+)
  • 893 posts

Posted 13 August 2017 - 19:38

I would need to point out that any BG39 filter that is a "quarter of the cost" is in fact not BG39, but instead a Chinese version that has been labeled as Schott.
QB39 is the Chinese version of Schott BG39, and it easily costs less than "one quarter the price".
Unless you have them side by side, and actually shoot photos through them to compare, you will not know the difference. I say this because I have tried it.
S8612 and BG39 both suppress Red/IR exactly the same amount at the same thicknesses, however, S8612 transmits UV stronger and deeper, whereas BG39 cuts off more UV transmission.
So I never see any reason to use BG39 or any imitation thereof, because it reduced UV transmission, has no better longevity, and imitation versions are not even as efficient.

Keep in mind that graph #2 above is a presentation by those selling Chinese glass, I don't trust it.
Also, keep in mind that most graphs you see, including most graphs made with spectrometers are presented in linear form, thus not showing out of band OD in any detail.
Detail of OD is important for actual comparison and functionality.
Actually stacking it and shooting works to see if there are leaks, but even then you will not know the difference unless you compare it to Schott or Hoya (much of which now is mislabeled as such).

Andrea?
"I don't see any visible green leak for the ZWB1 (2mm) as is evident in some other charts we've seen for thinner ZWB1."
Please point me to any of these charts you are referring to that show ZWB1 leaking green.
You may be mistaken about that.
I have posted in the past that comparing UG11 (1mm), U-340 (1mm), and ZWB1 (1.5mm), that visual leaking is encountered in the 500nm range and above with U-340 and ZWB1,
but this doesn't show up on any graphs, even diabetic graphs showing OD's, Hoya lacks data for the visual range with U-340, and the Chinese graphs are probably concocted from thin air to look like whatever they want them to look like.
Schott has the best and most inclusive filter glass data.
UG11 (1mm) suppresses stronger in the visual and 500nm+ range than U-340 (1mm), and even more so than ZWB1 (1.5mm).
Andrea: I will be sending you a ZWB1 (1.5mm) so you can compare and test this and see this with your eyes. You will find that ZWB1 @1.5mm thick leaks visual light, you can SEE it with your eye, hold it up to a fluorescent light bulb, stacked with S8612 (ANY THICKNESS, even 3.5mm thick!) and you will see some visual light), you will not see that light with the same stack using UG11. If you use U-340 1mm, you will also see the light.
WHY do we see light when we stack with S8612 that is 3.5mm thick?! Because it will not suppress the 500nm+ range. Any glass you find that will suppress the 500nm+ range will not be designed to transmit UV.
I never use or recommend U-340 anymore at any thickness under 2mm thick.
The thicker you make the filter the longer the exposure time, and the wind is not kind to thick filters.

Edited by Cadmium, 13 August 2017 - 19:53.


#9 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,094 posts
  • Location: USA

Posted 13 August 2017 - 21:02

"I don't see any visible green leak for the ZWB1 (2mm) as is evident in some other charts we've seen for thinner ZWB1."
Please point me to any of these charts you are referring to that show ZWB1 leaking green.
You may be mistaken about that.


Steve, yes, you are absolutely correct. I don't have a proofreader, so sometimes I goof up and don't catch things in my own writing. I need to go back and cross out the word "green". I had just been writing about a thin U-340 elsewhere. I was attempting to refer to general visible leak in the ZWB1, not a particular color visible leak. The charts we currently have are in the Filters Sticky referenced above.

I'm looking forward to receiving my filter order!!




Daryll -- I forgot to add above that there is no absolute necessity for you to filter your UV torch for hobbyist UVIVF work. We will always try to describe the professional protocols for UVIVF or UV or IR photos, but everyone is free to experiment and play. You can get lovely UVIVF fotos without filtering your torch.
I try to always label my UVIVF fotos to indicate the types of filtration used -- or not. That way no one is mislead as to what is shown. I didn't filter my own UV-LED torch the first couple of years I used it for UVIVF because I didn't know about potential visible light leakage!! :D
Andrea G. Blum
Often found hanging out with flowers & bees.

#10 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,094 posts
  • Location: USA

Posted 13 August 2017 - 21:30

Found the ZBW charts link: What is the deal with Optima & ZBW glass?

The good ZBW charts are in that link. But look above the graph at the transmission data in the boxes. For example from 400 - 660 nm, a 1 mm thick filter of ZBW1 transmits 1% visible light. There is no way to block that. And between 680-700 nm the red visible leak varies from 11% to 30%. You have to pick a very good IR blocker that cuts off BEFORE 680 to block that red leak.

Let's just say that one should go for as thick a ZBW1 filter as possible in hopes of cutting that visible leak down somewhat.
Andrea G. Blum
Often found hanging out with flowers & bees.

#11 JMC

    Member

  • Members
  • 128 posts
  • Location: UK

Posted 14 August 2017 - 07:51

View PostAndrea B., on 13 August 2017 - 16:41, said:

The second of the BG39 filters did exhibit a slight striation/waviness towards one side of the filter which covered about 2-3mm of the surface.

Why am I not surprised? :D This stuff is cheap for a reason.

Test whether this waviness causes problems by shooting a sheet of newspaper taped to the wall. Camera on tripod, lined up square to the wall. Do you see any distortions in the resulting photograph? To confirm, rotate the filter slightly and reshoot. Did the distortions move with the filter rotation? Please let us know how this test turns out.

Yep I can do this, and will post the results when it's done.

Fully get that these are a compromise - you get what you pay for and all that. However finding good suppliers in the UK is not easy, and importing stuff adds to an already compromised exchange rate. As such while things may not be ideal, sometimes they can be good enough to prove a concept.