new method for infrared tri-color (using Tangsinuo GRB3)

[Fandyus]

I have already tried several ways of achieving this. This time I got the best results though,  I think.

• the blue channel is a 720nm longpass filter and the GRB3
• the green channel is an 850nm longpass and the GRB3
• the red channel is a 950nm longpass

 

Here's the transmission curve for GRB3 as published by Tangsinuo themselves.

As you can see, stacking this with either 720nm longpass or 850nm longpass would create a distinct peak by itself.

 

Here's a hypothetical and not-at-all scientifically accurate representation of how this (most likely) works:

 

 

The only disadvantage here is that the transmission of the green channel stack is very low, so the images take a long time to expose properly, otherwise though I think it works perfectly. I tested this by taking a picture of a Guiness World Records book that is pearlescent in the visible to see if it persists in IR. See for yourself.

 

Visible (phone photo)

 

IR Tri-color

You still see red, yellow, green and blue, so I think the spectrum is pretty good.

 

 

Here are some more examples of images made with this method.

 

Visible: Assorted Lightsources

IR Tri-color

 

 

 

Visible:  Glassware and Liquids

(the yellow-green one is uranium glass)

IR Tri-color

 

 

Visible:  Filters, gemstones, remote control, minerals.

(I unfortunately bumped the tripod on the last photo so I couldn't align properly)

IR Tri-color

 

Visible:  gemstones alone

IR Tri-color

 

 

 

IR Tri-color Bonus:  Fruits

(focused on the damaged apple area on purpose)

IR Tri-color Bonus:  jars with various foodstuffs

IR Tri-color Bonus: Sony DSC-F828 digital camera

The camera is not very interesting truth be told. The plastic is very persistently black.

 

Next thing I would like to try is capturing infrared fluorescence this way to see if it is as "colorful" as the visible sort.

[lukaszgryglicki]

Looking really good.

That would also be interesting in UV.

[Lou Jost]

34 minutes ago, lukaszgryglicki said:

Looking really good.

That would also be interesting in UV.

Yes, really good, and really interesting to think about doing it in UV too. A very nice advance.

[Fandyus]

36 minutes ago, lukaszgryglicki said:

Looking really good.

That would also be interesting in UV.

Thanks.

I has been done in UV too, but it is much more demanding, both filter and lightsource-wise. You definitely have the gear for it though. A UV specialized lens would help a lot with the blue channel photos which are preferably around 315nm.

[Fandyus]

4 minutes ago, Lou Jost said:

Yes, really good, and really interesting to think about doing it in UV too. A very nice advance.

Thank you.

[Stefano]

Very nice. Reminds me a lot of my photos, but I used LEDs for mine.

 

Bernard did IR fluorescence too (link).

 

For anyone interested in seeing more examples of this techique I suggest you to search topics tagged "TriColour". I think this one should be tagged too.

[Fandyus]

27 minutes ago, Stefano said:

Very nice. Reminds me a lot of my photos, but I used LEDs for mine.

 

Bernard did IR fluorescence too (link).

 

For anyone interested in seeing more examples of this techique I suggest you to search topics tagged "TriColour". I think this one should be tagged too.

Thanks.

Also damn, Bernard beat me to it. Truly one of my favorite posters on this forum. It's a shame he has not posted during my time here (I think).

I tagged the post. Thanks for the reminder.

[Andrea B.]

(Thanks for tagging. I added some labels to clarify what viewers are seeing. They may not all immediately recognize the stuff we do here.)

 

Nice work!! 

 

Fandy, it would be useful if you gave some specs and/or equivalencies for the GRB3 glass? Just in case not all readers understand what that is. (Like me, duh! I had to go look it up.) I am wondering how the GRB3 contributes to this tri-colour? I'm curious how the images would look without using GRB3?

I do get that the GRB3 cuts off some IR. So using GRB3 in combination effectively gives you an IR band-pass filter, yes? That is probably the answer to my first question. 

[Fandyus]

3 hours ago, Andrea B. said:

(Thanks for tagging. I added some labels to clarify what viewers are seeing. They may not all immediately recognize the stuff we do here.)

 

Nice work!! 

 

Fandy, it would be useful if you gave some specs and/or equivalencies for the GRB3 glass? Just in case not all readers understand what that is. (Like me, duh! I had to go look it up.) I am wondering how the GRB3 contributes to this tri-colour? I'm curious how the images would look without using GRB3?

I do get that the GRB3 cuts off some IR. So using GRB3 in combination effectively gives you an IR band-pass filter, yes? That is probably the answer to my first question. 

Thanks a lot.

Yes, sorry for not explaining myself. GRB3 is like conventional IR blockers but it cuts off deeper into the IR spectrum. This results in it being able to form two distinct transmission peaks when stacked with either the 720 or 850nm longpasses. It transmits very little around the 850nm mark though, so the output is very dim when stacked that way. I suppose I could repost the transmission charts given by Tangsinuo themselves but I don't know if that's okay to do on your forum, since I am not the one who made these charts.

[Andrea B.]

I think transmission charts have always been ok to repost. We've had transmission charts from Tangsinuo, Schott, Hoya and others posted here. 

 

Thanks for the explanation about the two distinct peaks.

[ulf]

I have a GRB3, 2mm that I can measure when I get some free time.

AFAIK the GRB3 is supposed to be similar to KG3 

[ulf]

An alternative stack for a better peak around 700nm is to stack an orange or red filter with a ZWB1 or ZWB2. They might be thinner than 2mm for a batter transmission.

The peaks created are those we fight to block with BG-glass when creating a good UV-Pass filter.

 

For peaks higher up in NIR I think it is better to use a dichroic bandpass filter like the one Fandy found for the black water effect a few years ago.

Such filters are ar rather leaky in VIS so they have to be combined with a IR filter ≥720nm.

They are often also rather small in diameter and works best as ream mounted on lenses with a small rear element.

 

I bought the same filter 960nm? and it fitted well at the back of my Novoflex 35mm lens.

 

The same seller had similar filters closer to VIS too. They were very cheap.

[Foxfire]

Nice ones, liked the one with uranium glass but these all are great. Am interested did you use something to align the images(channels)? 

[lukaszgryglicki]

I would like to have something made from U-glass.

I do have some U-238 pieces (deplated).

 

[Fandyus]

On 8/3/2023 at 9:12 AM, Andrea B. said:

I think transmission charts have always been ok to repost. We've had transmission charts from Tangsinuo, Schott, Hoya and others posted here. 

 

Thanks for the explanation about the two distinct peaks.

No problem, I will repost the official transmission chart and edit it into the post.

[Fandyus]

22 hours ago, ulf said:

An alternative stack for a better peak around 700nm is to stack an orange or red filter with a ZWB1 or ZWB2. They might be thinner than 2mm for a batter transmission.

The peaks created are those we fight to block with BG-glass when creating a good UV-Pass filter.

 

For peaks higher up in NIR I think it is better to use a dichroic bandpass filter like the one Fandy found for the black water effect a few years ago.

Such filters are ar rather leaky in VIS so they have to be combined with a IR filter ≥720nm.

They are often also rather small in diameter and works best as ream mounted on lenses with a small rear element.

 

I bought the same filter 960nm? and it fitted well at the back of my Novoflex 35mm lens.

 

The same seller had similar filters closer to VIS too. They were very cheap.

You are correct about the dichroic filters. The problem I have with this method is that I have to putty mount the filter to the back of the lens, so I would basically need three separate lenses that I'd have to switch for every photo and that is a huge nuisance. This is way more practical to me truth be told.

As for the ZWB1/ZWB2 method, I have actually used a ZWB2+510nm longpass for the IR (red) channel in my recent aerochrome emulation attempts (haven't posted yet) and I felt like these looked like just normal IR photos. Do ZWB1 and 2 filters start blocking again in deeper NIR?

20 hours ago, Foxfire said:

Nice ones, liked the one with uranium glass but these all are great. Am interested did you use something to align the images(channels)? 

Thanks, yes I used photoshop to align the channels but I usually have to do some manual alignment afterwards as well to minimize color fringing.

[Fandyus]

16 hours ago, lukaszgryglicki said:

I would like to have something made from U-glass.

I do have some U-238 pieces (deplated).

 

Well, there is a lot of uranium glass floating around on online auctions, but very pricey. I have gotten my pieces (I own a few more) from flea markets, where they are a lot cheaper sometimes. I have also gotten some other fluorescent pieces of glass. The one with the orange top is a good example, might not be visible but it's fluorescent the way uranium glass is.

[lukaszgryglicki]

Yeah, but I'm also a bit afraid of fake ones. Anyway, this is an OT.

[Lou Jost]

22 hours ago, ulf said:

An alternative stack for a better peak around 700nm is to stack an orange or red filter with a ZWB1 or ZWB2. They might be thinner than 2mm for a batter transmission.

The peaks created are those we fight to block with BG-glass when creating a good UV-Pass filter.

 

For peaks higher up in NIR I think it is better to use a dichroic bandpass filter like the one Fandy found for the black water effect a few years ago.

Such filters are ar rather leaky in VIS so they have to be combined with a IR filter ≥720nm.

They are often also rather small in diameter and works best as ream mounted on lenses with a small rear element.

 

I bought the same filter 960nm? and it fitted well at the back of my Novoflex 35mm lens.

 

The same seller had similar filters closer to VIS too. They were very cheap.

While a filter with "natural" peaks is probably best, we can do tricolor imaging with ordinary longpass (LP) filters that have different cut-off points. If you have a Hoya R72 (720nm cutoff), and one that cuts off at 800nm, and another that cuts off at 900nm, for false RGB you can get the "R" channel by subtracting the image taken with the 800nm LP filter from the image taken with the Hoya R72 LP fiter.  The "G" channel can be obtained by subtracting the 900nm LP image from the 800nm image. And the 9i00nm channel can be obtained directly from the 900nmLP. Depending on the filters used. it may be necessary to normalize the images (equalizing images of a teflon target tape under the same light source as the real images) before subtracting them, since their transmission percentages will probably differ.

[Fandyus]

7 minutes ago, lukaszgryglicki said:

Yeah, but I'm also a bit afraid of fake ones. Anyway, this is an OT.

Not sure what you mean by OT. Also, I have never heard of fake uranium glass. Is that a thing? I had it that if it glows greenish in daylight and is heavy and cold feeling like glass, that's uranium glass. I have not heard of any other "conventional" materials that would do this. Obviously, there are many plastics that do this but you could tell plastic just from the way it feels in your hand.

[lukaszgryglicki]

By OT I mean "off-topic" and by fake I mean buying online where I cannot touch it.

 

[ulf]

2 hours ago, Fandyus said:

As for the ZWB1/ZWB2 method, I have actually used a ZWB2+510nm longpass for the IR (red) channel in my recent aerochrome emulation attempts (haven't posted yet) and I felt like these looked like just normal IR photos.

 

Do ZWB1 and 2 filters start blocking again in deeper NIR?

Thanks, yes I used photoshop to align the channels but I usually have to do some manual alignment afterwards as well to minimize color fringing.

Even a single ZWB1 or ZWB2 can work as a mainly IR filter as the camera sensitivity in UV is not that high, but if you include the sky it will be purple or bluish.

Stacking with a long pass filter would eliminate that.

Then you get a rather distinct peak only,  around 720nm / 750nm,

 

The peak of ZWB1 and UG11 is closer to VIS and it is a peak, meaning that you rather soon have little transmission further up in the NIR.

Here is some transmission graphs for 1mm thick filters:

https://www.pgo-online.com/intl/curves/optical_glassfilters/UG1_UG5_UG11_BG24A_kurve.html

With 2mm thick filters the attenuation of the longer wavelengths will be much stronger.

Just square the ration between the peak and values further up.

The UG1/ ZWB2 benefits more of a thicker glass.

 

The UG11 / ZWB1, 2mm will have a quite clear peak already at 720nm and less influence of UV/ purple light.

There might not be any sky effects at all. I do not remember if I tested that 

[Fandyus]

@ulfOkay, that's a huge shame then, since I would love to try it but I can't really. I own three ZWB1 filters, two have striations, two have developed oxidation I can't remove, not even with Cerium oxide. Neither of the three is without fault. If only they worked. Not only would I be able to use them for said experiment here, but also I would be able to stack the two of them and my QB39 and get UVB-ish images with my reptile bulb. But oh well. Once I get a job and some income, I guess I'll have to place another order. Doesn't help that the ZWB1 is the most expensive out of the ZWB substrates.

It's odd, I never had problems with any other Tangsinuo glasses but the ZWB1s are just unviable. It's even stranger when I consider that my ZWB2 has been basically constantly exposed to the elements since I got it two years ago, and yet it is pristine. I don't take it off the Industar I use it with. The ZWB1s all sat comfortably on my shelf in their packaging and they all degraded. The substrate must be really unstable.

Though I am planning to try and use the diet coke you see pictured above to remove the oxidation with the phosphoric acid it contains. We'll see.

Edit: I suppose I could still try substituting the blue channel channel with just a ZWB2 stacked with something that blocks UV. Might test this tonight. Perhaps with the pearlescent GWR book.

[Andrea B.]

Interesting to hear about the degradation of the ZWB1 filter glass. It might have been old when it was sold?? Maybe I should put a warning to immediately inspect and clean any ZWB1 glass when the purchase arrives.

[Andrea B.]

The "problem" I had when I tried Tricolour IR is that the three files were so similar that I didn't get much color. I'll go find the link and put it here --> LINK

 

I did not do a good job of testing those MaxMax IR bandpass filters. I should try again. They are beautifully made filters, FWIW.