What is the deal with Optima & ZBW glass?

[Andrea B.]

I've been trying to track down information about ZBW filter glass reportedly made by the company named Optima.

 

Turns out Optima is a Japanese company: http://www.optimajp.com/

 

So why are there so many Chinese optical companies offerine ZBW glass? This is how we usually see ZBW glass on Ebay - from Chinese vendors. Are they "knocking-off" ZBW glass? Does anyone have any information?

 

Currently I have had to put a Warning in our Filter Sticky not to buy ZBW filters for UV-only photography due to high leakage of visible light and also because of unacceptable levels of flaws. But maybe there is a "real" version of ZBW which is of better quality??

[Andrea B.]

Found these charts finally on the Optima pages.

 

Does not look good for ZBW glass for UV-only photography because of high visible transmission?? Of course, I need some tests to show how this actually affects UV-only photography. We don't mind some violet transmission on the 400nm end of things. But these filters show more than that!

 

I will need to update the Filter Sticky.

 

 

 

 

EDIT: Duplicate post removed.

[nfoto]

These filter might be a candidate for "BUG" captures, when combined with a suitable IR-blocking filter.

[Cadmium]

Optima ZWB1 is their 'equivalent' to Schott UG11 and Hoya U-340 .

Optima ZWB2 is their equivalent to Schott UG1 and Hoya U-360.

Optima ZWB3 is their equivalent to Schott UG5 and Hoya U-330 (which could be used for UV+Blue+Green, BUG, but ZWB1 or ZWB2 would not be used for that).

 

It is my understanding that Optima makes their glass in China. I don't know where Hoya makes their glass, but even Schott, a German company, makes their filter glass in Malaysia.

Equivalents are not always equivalent. The ZWB1 I have does not block visual completely even at 1.5mm (when stacked), neither does the Hoya U-340 at 1mm (when stacked),

yet the Schott UG11 1mm blocks visual completely (when stacked).

It only really matters how it works.

 

Generally I consider Schott brand filter glass the best. There are a few types of glass that Hoya has that are either unique from any offering Schott has, or are better.

One type of filter glass that I feel is better than the Schott version is U-360 which out performs UG1.

I trust Schott graphs and data the most, and Schott has data for every 1nm (in their program), where as Hoya has data for every 10nm, and Optima even less. I don't trust Optima graphs or data.

 

There is some 1mm thick filter glass being sold to people that they will never know is inferior. Even some knock off filter glass being sold as Schott and Hoya brand glass. :(

There is no way to know who is doing what. Price can be an indicator.

One thing I know, if you have any UG11 filter glass that is 1mm thick, and you stack it with an appropriate thickness of BG## or S8612 filter glass for Red/IR suppression (say S8612 2mm or more, for example),

and you look at a common household light through this stack, and if you see any dim light coming through it, then it is not UG11, it is something else.

You can try this test with UG11 1mm, U-340 1mm, or ZWB1 1.5mm.

The common idea is, "I will just stack this stuff with thicker BG glass", but It will not matter how thick the BG/S8612 glass is, if the U-glass is not suppressing the visual range adequately.

I have stacked U-340 1mm and ZWB1 1.5mm with 6mm ( ! ) of S8612, and the dim light is not removed, because it is in the 500nm range and upward.

If U-340 or ZWB1 is thick enough, then it is OK, but you will need to use 2mm U glass, and this reduces the UV band width and also lengthens exposure time,

and 'the wind' will follow you around, because it knows a thick piece of U-Glass when it sees one. ;)

Also, the ZWB1 I have is not as efficient, meaning even though the ZWB1 graph may show the same peak UV transmission strength as UG11 and U-340, and show the same UV bandwidth as UG11 and U-340,

my tests have required more exposure time. Results will vary, but these are my experiences.

[rfcurry (1950-2024)]

I believe that the ZWB* glass is a designation of a chemical formula or recipe for the melt. Quite a number of different Chinese companies produce the ZWB*. As with any ionic glass, you need to look upon the transmission characteristics -- and bubbles, striations, etc. - as the properties of the individual melt. If you buy a large enough quantity, the manufacturer will give you the testing data for that melt. Unfortunately, this is akin to asking for your dog's ashes from the vet's office. :( (That may be Fido's carbon... or Fluffy's, or both of the above)

[Andrea B.]

That makes sense about ZBW being a designation for a glass melt given the various references I've turned up online.

 

And I'm sure there are some good versions of this ZBW glass "out there". It does not seem to surface on Ebay however.

[BruceG]

Just received my order of ZBW1 and ZBW2 glass, they do transmit a bit visible light, I did some research and it seems ZWB is a product code used in China. Countless factories are making ZWB glass, unlike U-340 which is only manufactured by Hoya. There is also ZB1, ZB2 and ZB3 which correspond with B-390 BG3 and B-370 respectively. They are dirt cheap anyway so I got a dozen. Waiting to to get IR blocking filters now for some testing.

[Andrea B.]

But UV filters should not leak visible light. There is way more visible light than UV light in sunlight, so even small leaks can contaminate a UV photo.

 

So I'm looking forward to your examples. I hope these filters work for you. But am waiting to see.

[bvf]

This old topic is of current interest to me, as I am contemplating using ZWB1 in my next project - to do full-colour IR photos using the same techniques as for full-colour UV, which has been discussed elsewhere.

 

One of the difficulties has been finding a bandpass filter in the 700-800nm range (without forking out for the extremely expensive MaxMax IR Tri-colour set). So I'm thinking of taking advantage of the IR leakage from ZWB1 by sandwiching it with an R72 to cut out the UV and Vis. This looks like it will work for the 700-800 nm component, but I've been a bit concerned about the 2nd blip of IR leakage through the ZWB1 from around 950 nm, which might interfere with my image for the 900nm+ range.

 

I was trying to figure out whether an Optima ZWB1 would be better than a cheap-from-China ZWB1. (BTW - I followed Andrea's link for Optima, but the site is competely devoid of content.) Here's a graph of the spectra through the two - using the data in the Optima table above (brown line) plus ths far less precise graphs published on ebay by the ZWB1 sellers from China (grey line):

 

 

The following graph shows the effect of combining these ZWB1s with R 72. I have also applied a factor reflecting the typical sensitivity of CMOS sensors at different frequencies. The vertical scale is different to the first graph, Optima ZWB1 + R72 is red, China ZWB1 + R72 is blue.

 

 

If the Chinese graphs are accurate, the Chinese version would be better for my application. I'm hoping that the 950nm+ transmission is low enough to not cause any problem: I seem to remember there were similar overlaps with Kodachrome and Ektachrome - and they did OK!

[Cadmium]

Don't use ZWB1. It is made in China, nothing essentially wrong with China, but it is not made by any one company or source.

It is not consistent. There are many courses in China making the same thing, not always labelled as ZWB... or Optima, but also labeled as Scott... or Hoya..., ro whatever you are wanting....

If you want consistency, then get the real thing, trust Schott and Hoya.

Filters from China are not the same.

 

ZWB1 is supposedly like Schott UG1 or Hoya U-360.

ZWB2 is suppose to be like Schott UG11 or Hoya U-340.

If you want to wear those shoes...

 

 

.

[ulf]

When comparing transmission graphs from different sources it is important to keep in mind that they might represent results from two different glass thicknesses.

They can also contain different types of measurement errors. Only data from reliable manufacturers are to be trusted.

 

The best data, iI think, comes from Schott and it represent a typical average of several melts.

Evan from them there is only garantied performance at a few essential specified wavelengths.

[Cadmium]

I agree, Schott has the best and most complete transmission data.

Hoya has the second best.

Don't trust any transmission graphs or data you get from the so called Optima/Chinese graphs.

it is just plain invented.

Schott is the best.

I am not saying that Chinese UWB... will not work, but I am saying it is not as good, it is not calculable, and it will not work as well, as efficiently, or as trustworthy.

So, "at your own risk", and don't expect me to view your results as accurately or scientifically as how I would view results made with Schott or Hoya glass.

[bvf]

Good point about thickness, Ulf - the Optima data is for a 1mm filter and the China data for 2mm, which could explain the higher transmission from the Optima. There may also be an issue in whether different sources are providing Transmission or Internal Transmission data.

 

I have considered the UG11, but it is a lot more expensive and is just one of 3 filters I have to buy for what is just an experiment. I'll probably get started with a China ZWB1, and fork out the money for a UG11 is this isn't working or I decide to progress beyond experimenting to doing some serious work.

[Cadmium]

If you don't like UG11, then get U-340, but make sure you get some from someone who isn't selling you ZBB2 instead of real UG11 or U-340, b3cause there are some who do that.

I never do that. I am honest, otherwise I would not be doing it at all.

[Cadmium]

Keep in mind that you need to use diabatic graphs when calculating OD suppression. Show me a ZWB/Optima Diabatic graph.

You need OD suppression info when calculating stacks.

[bvf]

I've got a U340, but that doesn't look useful because the IR leakage is so low - for my purpose, the more IR leakage the better, as long as the 900nm+ leakage is low.

 

I actually have 2 U340s. One says "U-340 Pro Series 49mm" on the mount. The other says " 52mm U-340" - but you can see the ring is also engraved with U-360, UG-11, BG-38, BG-39 - presumably so the vendor can put whatever glass he wants in the mount and then fill the appropriate engraving with white.

 

I did compare them once, and they seemed to give different rersults. So I guess (at least) one of them is not the real thing - but I don't know which.

[Stefano]

I always wandered how do they make this types of glass. There are three of them: the "360" type (Hoya U-360, UG1, ZWB2), the "340" type (Hoya U-340, UG11, ZWB1) and the "330 wide bandpass+violet" type (Hoya U-330, UG5, ZWB3). Normal Wood's glass is similar to the "360" type (according to Wikipedia it peaks at a nice 365 nm), and it is a "special barium-sodium-silicate glass incorporating about 9% nickel oxide". So, how do they achieve the various types chemically? What substances do they contain? It is very unusual for something to be transparent in UV and opaque in VIS. And it is also very unusual for something to be opaque in IR and transparent in VIS (S8612, KG1, etc.).

 

I have a ZWB2 (2 mm) + chinese BG39 (2 mm), and when stacked they become a nice UV-pass filter, and I didn't notice any significant leak. Of course, if 1 mm of glass has a 1% leak (OD 2), 2 mm of the same filter have a 0.01 % leak (OD 4).

 

Hoya U-340 passes more UV and less IR than UG11 at equal thickness.

 

Thorlabs and MidOpt also make their filters.

A "330" type filter from Thorlabs is FGUV5-UV, which peaks at 330 nm with a transmission of 98.307404% (the spectrometer they use must be impressive, 8 digits of precision! Or maybe it is just an average?), and they also make FGUV11-UV, which is a "340" type glass peaking at 326 nm with 90.701832%.

MidOpt makes BP324, peaking at 330 nm with 90.81% transmission (I guess it actually peaks at 324 nm) and the BP365 (which strangely leaks IR only past 800 nm, I guess it is an interference filter but I may be wrong). Also very interesting their BP250, peak at 240 nm with 35.36% and what seems a >OD 3 blocking.

[Andy Perrin]

Several things here:

1) Chinese glass has unreliable specs, because it depends on the company who makes it and there is no standardization.

2) there is no such thing as “Chinese BG39” although I understand what you meant to get across. BG39 is a Schott product.

3) Omega filters are missing from your list for some reason?

4) Thorlabs and MidOpt are dichroic filters. I have a feeling those peak values are calculated values not measured.

5) You need much better than OD3 blocking in general, at least OD4, and much more if you are using a filter in a place where your camera sensor has low QE. Like 250nm. (Don’t forget you need a quartz lens also.)

[bvf]

...

3) Omega filters are missing from your list for some reason?

...

 

I know this question was addressed to Stefano, but I did approach Bob at Omega to suggest suitable IR colour separation filters - he had come up with the goods when I was doing this with UV. Several times he said he would come up with a set for IR but this hasn't happened - so I'm looking for other solutions. I'll use a MidOpt 850 for my mid-wavelength filter, and a MidOpt L920 or Chinese IR950 Long Pass for my long-wavelength filter.

 

Of course, the ideal filter set is provided by MaxMax. But the three filters would set me back $500. I saw a used set on ebay for a mere £300, but that's still too much to spend on a whim.

[Andy Perrin]

Bernard, I actually have some Omega bandpass filters -- they aren't super wide, but I could definitely try doing the three wavelength thing with them. I have a 780BP30, a 830DF30, and a 1064BP25. I could try doing it with the TriWave, which is monochrome and has no pesky Bayer, and nearly constant gain in the NIR!

[Stefano]

Several things here:

1) Chinese glass has unreliable specs, because it depends on the company who makes it and there is no standardization.

2) there is no such thing as “Chinese BG39” although I understand what you meant to get across. BG39 is a Schott product.

3) Omega filters are missing from your list for some reason?

4) Thorlabs and MidOpt are dichroic filters. I have a feeling those peak values are calculated values not measured.

5) You need much better than OD3 blocking in general, at least OD4, and much more if you are using a filter in a place where your camera sensor has low QE. Like 250nm. (Don’t forget you need a quartz lens also.)

Yes, using filters such as the BP250 alone is not enough. In UVA you need at least OD 4, better if you have OD 5 (more than that is usually useless, it depends on the UV/IR ratio you have). I thought it was interesting because it can be stacked with other filters in that region to reduce leakage. I didn't know about Omega, I will look for that. And the filters from Thorlabs, I am sure, are colored glass. They are on the colored glass section, are "black glass" in the image they provide and that secondary IR peak is exactly what you would expect in a "Wood's glass" type filter. I know Thorlabs makes interference filters too.

 

So, does somebody know how they make (in general) the three types of glass?

[bvf]

Bernard, I actually have some Omega bandpass filters -- they aren't super wide, but I could definitely try doing the three wavelength thing with them. I have a 780BP30, a 830DF30, and a 1064BP25. I could try doing it with the TriWave, which is monochrome and has no pesky Bayer, and nearly constant gain in the NIR!

 

Those bands are too narrow to be ideal, but it would definitely be worth having a go - a sort of proof of concept. I would expect you to get a full-colour image, but without all the nuances of colour.

 

My ideal filters would have been 700-800nm, 800-900nm, 900-1,000nm - with a bit of overlap of each. I think the MaxMax filter set is close to this, but out of my experimenting budget.

 

Give it a go!

[dabateman]

Bernard,

Did you check Bjomejag store for those filters. I think I saw a set like that. Including this 800nm band one:

https://www.ebay.com/itm/Optical-Filter-850WB80-25-mm-Bandpass-NOS-Sputtered-Laser-LED-IR-Fluore-80-T/153495462231?hash=item23bd0ae157:g:~-YAAOSwcj1c4eJc

[SteveE]

This listing from Bjomejag would seem like it might fit your requirement, if $90 is not too expensive. You specify the desired NIR wavelength and width, and he will find you one within 25% of the specified width

 

Optical Interference Filter CUSTOM BUILD,NIR Wavelength,FWHM 12 to 50nm 25mm,CCD

https://www.ebay.com...CD/152161062807

[bvf]

Dave - amazingly I hadn't picked up on that filter, even though I've spent several hours trawling through the Omega shop. It's possible I saw it but had a concern with the big spike approaching 1200nm: I don't know what the long-wavelength cut-off of my camera is. As it happens, I think I have the 800-900nm range nailed. The Midwest Optical (hadn't heard of them before) BN850 looks almost perfect - https://midopt.com/filters/bn850/ . And Discount Optics on ebay is selling a 52mm one at $20 (becomes doubled when you include UK postage and import duty).

 

Steve - I hadn't though of trying the custom option. The description says "I will search our multi-million piece inventory of surplus parts for a closest match." - I guess that is what he was going to do when I contacted him, but hasn't got back to me. But this may be an option if my initial choices don't work.

 

A slight drawback of both these filters is the 25mm diameter. I had no option but to use this size for UV tricolor, and that limited me to my 50mm and 105mm lens: I'd prefer larger filters so I could use a wider range of lenses when I'm doing the IR.