Possible better UV filter stack?

[Stefano]

Almost did it. Four layers, 69.02% peak, OD 4.966, 4.6 mm total thickness. This filter is absolutely theoretical, and pure madness. 0.5 mm Hoya U-340. BG25 at 0 mm (it's not there). Also odd thicknesses.

 

 

 

In general, to have a high transmission, you have to make the U glass thin and the BG glass thicker. It may seem counterintuitive, but if you try it actually works. We need a better BG glass, but I can hardly imagine something better than the S8612. The limiting factor, both for peak transmission and for bandwidth (especially at shorter waves) is the BG glass. Even something similar to a BG3, but with a cut-on in the infrared shifted 20 nm into longer waves, would make a huge difference.

[dabateman]

Do you needthe b410 in that stack. It looks like the bg3 would cut the visible leak and the S8612 cuts the IR leak.

 

[Cadmium]

"May transmit UV better..." ?

Or, may not transmit UV very well at all, unless you know.

It might be advisable to invest in other equipment of a known quantity first before concentrating on the filter which you gain nothing from with most lenses.

 

Your plot above transmits visual starting with 5% at 400nm, and doesn't go below OD3 until 500nm!

So you have a fairly significant visual range leak, probably get blue grass with that filter like the PrimaLuceLabs Venus U does...

 

Another thing to point out about the 330WB80 I see on eBay, they seem to have OOB leaking in the IR range, according to their OD graphs.

152895097765, and 152779229306 (looks best, but how much above 800nm, unknown with that one).

[Andy Perrin]

The 330WB80s don’t leak much IR in practice with sunshine, although they will under halogen. I know Cadmium doesn’t like it but it’s worked fine for me and I’ve been happy with it. However I am a little worried because your lens cuts a lot of shorter UV so that would make any out of band issues worse.

[dabateman]

If you get a 330WB80, it needs to be the improved version. That one has the least IR leak and I use it to stack with to reduce IR leakage from other filters.

 

This one:

https://www.ebay.com/itm/Optical-Filter-330WB80-24-5mm-Improved-T-and-Blocking-to-Excite-Fluorescence/152779229306?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

 

I am not sure about other wide band filters.

But the other one I like that would work for you is the 390bp25.

 

https://www.ebay.com/itm/Optical-Filter-390BP25-25-4mm-blocking-to-1100-Very-High-Transmission-65/152857013053?hash=item2396fceb3d:g:hXgAAOSwiUpa50Te

[Andy Perrin]

I agree about getting the 330WB80 improved version if you go that route.

[Cadmium]

The 330WB80s don’t leak much IR in practice with sunshine, although they will under halogen. I know Cadmium doesn’t like it but it’s worked fine for me and I’ve been happy with it. However I am a little worried because your lens cuts a lot of shorter UV so that would make any out of band issues worse.

 

I didn't say anything about not liking the filter. I am just pointing out important aspects to be aware of according to the graphs shown, and for people with limited UV lenses.

One of those filters, which it seem you both agree with me about, is not appropriate given the OOB leak.

The other filter graph is limited to 800nm, but the trend doesn't look good to me, but if it works for you...

The other thing I was pointing out is that for some people who have a more limited lens transmission the removal of higher 380-400nm UV would not be advisable.

It would depend on the upper UV cutoff point.

These are things people should be aware of.

 

Here is the 330WB80 I have, which shows less upper UV. This was using a good UV transmitting Kuribayashi 35mm lens. Disregard 460DF20, it leaked.

[Andy Perrin]

My copy at least seems to work fine in sunshine. Dandelion centers are dark, no signs of red after white balance, etc. But that’s using my EL-Nikkor 80mm and also the quartz KSS lenses. I see what you mean about the issue with a lens that has a cutoff at ~370nm

[Cadmium]

Good, because you guys are into UVB, so a 380nm and below filter will work for you, but most people don't do much of that, they don't have lenses and equipment for that, and it is not a beginners thing,

UVB is advanced, and not something many people are going to be interested in, because they can do UV much more simply and satisfactory with higher UV.

UVB is not something you can do with the usual lenses and filters and the camera is not very sensitive to that either.

I don't want people to get the idea that they are going to use such a filter on anything other than a very good UV lens, using extended exposure times, and/or special lighting...etc..

Now you want to talk Baader U, then I have no problem but the filter you are recommending for Stefano and 95% of the rest of the people reading this is not the filter to recommend.

I don't recommend it to anyone for usual UV photography, that is the deep end of the pool, teach the people how to swim first, they will read your UVB stuff as times goes by.

Would I recommend a U-340 8mm thick filter to someone, no, it is advanced, it needs a very good lens, it is not usual, and it doesn't produce the usual reflected UV photos.

How many people come on here, talking about UVB before they even take a UVA picture.

That is where I am coming from, and I think it is realistic and helpful for people. I want to be real with people's expectations.

I have nothing against UVB, but it is not the realm of the usual UV photographer, especially a beginner with less specialized and prolific equipment.

[ulf]

There is not that much lost in the upper parts of UV-A by the Omega 330W80 improved.

1. Black : U-360/S8612 2+2mm
   
2. Magenta: Baader U
   
3. Green: Omega 330W80 improved
   
4. Orange: PtimaLuceLab UV
   

 

These measurements are normalised to a 100% peak maximum for proper cutoff comparison.

The measurement is some years old and I do not have the un-normalised measurement file anymore.

All these filters have a rather good peak transmission, so the relative transmissions are not that far off.

 

The OD of the Omega 330W80 improved seams to vary quite a lot and my sample is far from as good as the one Cadmium tested above.

My sample leaks slightly more than my Baader U.

[Stefano]

Your plot above transmits visual starting with 5% at 400nm, and doesn't go below OD3 until 500nm!

So you have a fairly significant visual range leak, probably get blue grass with that filter like the PrimaLuceLabs Venus U does...

Don't be confused. That transmission comes from my choice of writing 0.0000001 transmission (at 2.5 mm thickness) in the empty areas where data isn't provided by Hoya. At 0.5 mm thickness, that rises to 5%.

[Stefano]

Anyway, @Cadmium, how thin can you grind filters? I know you can reach 1 mm, but can you go below that?

[dabateman]

Good, because you guys are into UVB, so a 380nm and below filter will work for you, but most people don't do much of that, they don't have lenses and equipment for that, and it is not a beginners thing,

UVB is advanced, and not something many people are going to be interested in, because they can do UV much more simply and satisfactory with higher UV.

UVB is not something you can do with the usual lenses and filters and the camera is not very sensitive to that either.

I don't want people to get the idea that they are going to use such a filter on anything other than a very good UV lens, using extended exposure times, and/or special lighting...etc..

Now you want to talk Baader U, then I have no problem but the filter you are recommending for Stefano and 95% of the rest of the people reading this is not the filter to recommend.

I don't recommend it to anyone for usual UV photography, that is the deep end of the pool, teach the people how to swim first, they will read your UVB stuff as times goes by.

Would I recommend a U-340 8mm thick filter to someone, no, it is advanced, it needs a very good lens, it is not usual, and it doesn't produce the usual reflected UV photos.

How many people come on here, talking about UVB before they even take a UVA picture.

That is where I am coming from, and I think it is realistic and helpful for people. I want to be real with people's expectations.

I have nothing against UVB, but it is not the realm of the usual UV photographer, especially a beginner with less specialized and prolific equipment.

 

I was also recommending the 390bp25. That is probably the best starter UV filter. It cuts off at 403nm. I don't see the mercury 405nm line. And it has high transmission, which allows it to be used on most lenses and even non converted cameras.

I guess if I was Omega or Reed I would really be marketing that filter.

[Andy Perrin]

I certainly wasn’t talking about using the 330WB80 for UVB. I did mean UVA imaging and it’s worked on my case just fine for that using the EL-Nikkor 80mm/5.6 metal. I know I’m losing a tiny bit of the 390nm region but the camera gain (and UVA in sunshine) is so much higher there that I don’t think I lose more than the first 10nm (390nm to 400).

 

Stefano’s case is a little odd because he can’t change lenses to something like the EL-Nikkor.

[Stefano]

Hoya also makes a 365 nm peak filter, the UL365S. It is a very "diluted" filter (like the UG2A compared to the UG1). Leaks a good amount of violet, up to 420 nm.

 

The stack below peaks at 370 nm with 73% transmission, and has a OD of 5.027 at 710 nm.

 

 

 

Below, Hoya UL365S, 1 mm (alone). It transmits UV/violet below 420 nm and red/IR above 680 nm. I still haven't finished to put the data in, and I arrived at 800 nm, but it should still give an idea.

 

 

Regarding the data, this is the source: http://hoyaoptics.co...ing-filters-u/. There you should find the updated data for the other Hoya filters. Notice that the data refers to total transmission (T), not internal transmission (Ti). As reflection factors, I used:

 

0.911 for ≤400 nm

0.914 for 410-500 nm

0.915 for 510-600 nm

0.916 for 610-700 nm

0.917 for ≥700 nm

[Stefano]

Regarding the green leak of the Hoya U-340, lets say it peaks at 0.5% at 1 mm thickness. This means that a 2 mm thick Hoya U-340 (a commonly used thickness) has a OD of 4.6 at 550 nm. This means that your supposedly OD 5 filter stack actually isn't OD 5 in the green region. This of course doesn't affect photography, but take it into consideration.

[Cadmium]

I have been recommending the U-360 2mm + S8612 1.5mm to 2mm filter stack. Not U-340. I prefer UG11 to U-340 for optical filter use.

I have witnessed a leak in the 500nm range using 1mm thick U-340. I don't know how much it is, and I will not "let's say" what it might be.

Don't take my word for it, try it your self, get some UG11 1mm and some U-340 1mm from someone who actually sells real Schott and Hoya glass, and compare.

Let us know know if you see a difference, but you still can not "lets say" what the transmission is with your eyes.

So then, get a spectrometer, test each to compare the transmission difference and the OD of each.

Then you can "let's say" what we should say, because like I said, I can't say, and I won't say what I can't say, and I will not base any OD numbers on what I can't say and don't know.

See what I am saying?

 

Once again, I would not advise using U-340 thinner than 1.5mm or 2mm depending... based on my experience.

If you want to use 1mm, then UG11 works better at thin thicknesses than U-340. I have not detected the leak with thin UG11.

Using the Schott filter program, or looking at the Schott UG11 data sheet, you will see that UG11 has an OD5 data limit, using 1mm as its reference thickness.

We can actually calculate the OD for UG11. However we can not calculate the visual range OD for Hoya U-340, because they don't provide data for that range.

Schott UG11 has at least OD5 at 1mm thick from 411nm to 650nm. Still you will need to use 1.5mm thick UG11 to maintain OOB OD5. You can stay under OD4 with 1.2mm, or even 1mm.

You could also adjust the stack to be UG11 1mm + S8612 2.5mm thick, and that would keep OD%, but it would also reduce the UV peak amplitude and the UV range width a little more.

Once again, my favorite stack is the first sentence of this post.

[Cadmium]

As for the Hoya UL365S, it is rather hard to say what animal it is, since it is missing all of the data for the 410nm to 680nm range. Same old thing, and that is pretty bad when it is obviously a filter that contains UV+some visual.

And where is the Ti data? They are only including the T data for that filter, so we will need to find a way to convert that to enter it into the Schott program, did you do that for your graph?

We know that filter has a visual violet leak, even at the 2.5mm reference thickness, it is 10.4% T at 400nm, and 0.7% T at 410nm.

I thought one of your original criteria was under 400nm? Not anymore I guess?

http://www.hoyacandeo.co.jp/english/products/eo_pdf/UL365S.pdf

[Stefano]

I have been recommending the U-360 2mm + S8612 1.5mm to 2mm filter stack. Not U-340. I prefer UG11 to U-340 for optical filter use.

I have witnessed a leak in the 500nm range using 1mm thick U-340. I don't know how much it is, and I will not "let's say" what it might be.

Don't take my word for it, try it your self, get some UG11 1mm and some U-340 1mm from someone who actually sells real Schott and Hoya glass, and compare.

Let us know know if you see a difference, but you still can not "lets say" what the transmission is with your eyes.

So then, get a spectrometer, test each to compare the transmission difference and the OD of each.

Then you can "let's say" what we should say, because like I said, I can't say, and I won't say what I can't say, and I will not base any OD numbers on what I can't say and don't know.

See what I am saying?

 

Once again, I would not advise using U-340 thinner than 1.5mm or 2mm depending... based on my experience.

If you want to use 1mm, then UG11 works better at thin thicknesses than U-340. I have not detected the leak with thin UG11.

Using the Schott filter program, or looking at the Schott UG11 data sheet, you will see that UG11 has an OD5 data limit, using 1mm as its reference thickness.

We can actually calculate the OD for UG11. However we can not calculate the visual range OD for Hoya U-340, because they don't provide data for that range.

Schott UG11 has at least OD5 at 1mm thick from 411nm to 650nm. Still you will need to use 1.5mm thick UG11 to maintain OOB OD5. You can stay under OD4 with 1.2mm, or even 1mm.

You could also adjust the stack to be UG11 1mm + S8612 2.5mm thick, and that would keep OD%, but it would also reduce the UV peak amplitude and the UV range width a little more.

Once again, my favorite stack is the first sentence of this post.

I don't have neither a spectrometer nor a U-340 filter. I said 0.5% because I found this number on this forum (maybe it was from you).

[Stefano]

As for the Hoya UL365S, it is rather hard to say what animal it is, since it is missing all of the data for the 410nm to 680nm range. Same old thing, and that is pretty bad when it is obviously a filter that contains UV+some visual.

And where is the Ti data? They are only including the T data for that filter, so we will need to find a way to convert that to enter it into the Schott program, did you do that for your graph?

We know that filter has a visual violet leak, even at the 2.5mm reference thickness, it is 10.4% T at 400nm, and 0.7% T at 410nm.

I thought one of your original criteria was under 400nm? Not anymore I guess?

http://www.hoyacandeo.co.jp/english/products/eo_pdf/UL365S.pdf

Removing violet light above 400 nm is still my goal, I just posted an alternative filter.

[Cadmium]

Too bad about the newer data Hoya shows, with no Ti, which is what I have always used to enter into the Schott program, not sure how to enter the data now.

Except to use the Ti data from their older graphs.

When I said "a perfect OD5", I just mean that the plot (a plot) has exactly OD5 at the suppression point (700nm range), I am not talking about a flat out of band OD,

I am talking about the highest OD out of band, usually in the 700nm range.

As far as the visual 500nm range leak that I can detect with my eyes, I have no way of knowing the % of that leak.

One would assume it is not there, based on the graph of the UG11, assuming the U-340 1mm is quite similar to the UG11 1mm at OD5, but I can see a slight leak.

I can probably demonstrate that visual leak with some photos sometime, I just have never done that.

That is the reason I stay away from thin U-340. However that leak is not detectable when using thicker U-340, I think 1.5mm and above should be safe.

 

About your "lets say it peaks at 0.5% at 1 mm thickness" idea:

Looking at the data for U-340 data, for the 2.5mm thick reference thickness 680nm is the higher end of the provided visual data, which is 4.00E-04, which is just barely under OD3.

We don't know where it goes exactly below that point.

If we decrease the thickness to 1mm then that same 680nm transmission moves all the way up to 4.37E-02 which is just under 5%, that is a leak, however at the 680nm point the S86712 2mm is keeping that

4.37E-02 suppressed down to 4.32E-04, so under OD4.

However, the S8612 doesn't suppress the 500nm range, and we have no way of knowing what transmission the U-340 has in that range,

but it we were to pretend that the U-340 500 range was even half of the %5 then we still have a leak.

I guess I can't really guess at what the U-340 transmission would be in the 500 range, we just have no way of knowing.

You can see it though, with your eyes, it is pretty faint, but it is there. It doesn't really show in photos in most cases, but it it there to some extent.

[Stefano]

I had to manually calculate the Ti data, using reflection factors as I described above. It is time-consuming, but you can do that.

 

A 0.5% green leak is definitely detectable if you shine a white (better green) LED through the filter. Our eyes are quite sensitive to light.

 

BG3 is a must-have filter for stacks using very thin U-340. At 1 mm thickness, it strongly reduces green, barely affects UV and also reduces a lot deep red/infrared leakage before ~700 nm (especially at 2 mm thickness). This is even better considering that in general U-340 glass (Hoya U-340, UG11, ZWB1, etc.) has an infrared leak shifted towards shorter wavelengths (starting at about 680 nm and peaking at ~710-720 nm), as opposed to the leaks of U-360 type glass (Hoya U-360, UG1, ZWB2, etc.), which starts leaking almost at 700 nm and peaks at ~740-750 nm.

[Cadmium]

I see now, sorry I missed that part. What is your formula for the conversion?

 

I wouldn't use LED's, they will often time penetrate beyond normal. Use some kind of normal light. I have seen 365nm LED's penetrate through 830nm...

Weird, but true. Testing things with LED's can often times give weird results.

 

You still have not convinced me to buy or stack a UV stack with a BG3 or B-410 filter.

I think a good UV stack is all you need.

[Stefano]

Regarding the data, this is the source: http://hoyaoptics.co...ing-filters-u/. There you should find the updated data for the other Hoya filters. Notice that the data refers to total transmission (T), not internal transmission (Ti). As reflection factors, I used:

 

0.911 for ≤400 nm

0.914 for 410-500 nm

0.915 for 510-600 nm

0.916 for 610-700 nm

0.917 for ≥700 nm

This Is the method I used. For example, the T transmission at 360 nm is 80.1%. The reflection factor I used there is 0.911. So I divided 80.1 by 0.911, and then by 100 (what I actually did was 80.1*0.0109769484, which is the same thing as before (I multiplied 80.1 by ((1/0.911)/100)), so I obtained the Ti value in a 0-1 interval (instead of 0-100). The result is 0.8792535668 (and yes, I put every digit in the Excel file, even if it is useless). Do this for every point...

[Cadmium]

Stefano, Thanks for that formula! I will try it out later.

 

Basically just T divided by Reflection factor.