Kolarivision IR Chrome DIY Chinese version.

[Marc del Piro]

Aye up lads!

 

I have been attending this forum for a while as a reader, and following your discussions on how to imitate Aerochrome film in digital. This is clearly not what this post is about, although it put me on the right way to find out how to obtain a DIY version of Kolarivision IR Chrome filter. I would say it has been a mix of trial and error plus some rudimentary investigation on colour wavelengths. The filters tested are the chinese QB2, QB3 and QB19. A chinese factory was selling QB2 as the cheap version of IR Chrome, but QB2 is far from any resemblance by itself. They even stole Kolarivision photo from the website to promote the QB2.

 

 

 

 

The QB2 toghether with a GRB3 or KG3 from the same retailer made some difference, but still not good enough. The leakage of IR was still too strong. The tests I did with only visible light showed too much red and yellow overall. So the amount of light transmited around 500nm to 700nm was still too much compared to IR Chrome. On the images below you can see the original colour pattern I created followed by Kolari vs. QB2 vs. QB3.

 

 

 

 

All photos taken with a non-sensible NIR camera. So there is no possible NIR leakage in them.

 

QB3, as you can imagine, is far from any resemblance with Kolari's. The final image is too blueish. Although its wavelength is quite similar to Lee 729, I couldn't manage any similar results. I must say, my QB3 is just 1mm, I will come back to this later. I saw this QB19 filter after a while, which looked somehow in between QB2 and QB19.

 

 

QB19 filter is really similar to QB2, but cuts blues at 325nm aproximately. Also, does not allow that much light from 500nm to 700nm as QB2 does. I will let experts from this place to debate on it. I am not a physicist, so I am afraid I'd say some stupid stuff. 

 

My QB19 filters are 1mm and 1.5mm. Toghether with them, I tried GBR3 or KGB3 from the same retailer. These are 1mm and 2mm thick. Right now I am using 1:1 ratio. So same thickness on both filters. The good thing about using 1mm filters is that I can use one single ring for my super wideangle lenses.

 

Below you can see the results SOOC from a Fuji XT10 with 10-24mm and 15-45mm lenses. The order is as follows:

 

1. Original scene (no NIR).
2. Kolari IR Chrome.
3. QB19 + KG3 (True Color SR). 1:1mm
4. QB2 + KG3 (True Color 550nm). 2:2mm
5. QB3 + KG3 (True Color SB). 1:2mm

 

 

 

The slight difference you can see between QB19 + KG3 and Kolari's is due to adjusts made on camera white balance. I did not have a WB card, so I had to do it by visual approximation. A bit more green on the camera would solve the problem. QB2, as you can see has a bit too much red, this bothers a bit when editing and makes it hard to preserve the warm-orange-red tones from the vegetation. This combination, QB2 + KG3 is really nice to make golden vegetation.

 

The following images are in the same order, but the WB preset is a bit different.

 

 

 

I will keep posting some more images to compare True Color SR and Kolari's filter.

 

Disclaimer: I really appreciate Yann's work on IR Chrome, and this post is not by any means trying to discredit his work. In fact, I got inspired by his dedication to IR light and this is what took me here!

;)

 

 

[dabateman]

Interesting.  Is your Fuji XT20 full spectrum converted? 

The Kolari chrome looks to have a poor orange result with your camera. 

[Marc del Piro]

Oh, sorry I wrote down XT20. It is a XT10, I will correct it. Anyway, it is full spectrum converted indeed. The poor orange could be due to the preset used on camera, I think I might have added too much green on the WB preset.

 

I don't know if there is many people using Fuji in this forum, but as long as I've seen, Fuji tends to render this filter towards orange. I've managed to obtain a redish-purple tone on the leaves in certain light situations, but colour post-process is needed with this camera system most of the times.

 

[Yves W]

I have Fuji X-T20 and X-T2 and I'm happy with it.

[lukaszgryglicki]

I have Fuji GFX 50R - but not converted - also *very* happy with that camera!

 

[Marc del Piro]

This is the post processing from the RAW files.

 

Left/Top IR Chrome - Bottom/Right QB19 + KG3 (True Color SR).

 

 

 

Below, you can see two different crops from my screen in the post-process tab of Adobe LR. Note that to obtain the same result, I just needed slightly different slider adjustments. The biggest difference lies on WB adjustment. As you can see, QB19 + KG3 need lower colour temperature as well as adding more green. This is probably due to more IR light passing through my filter since it is just 2mm with 1mm KG3. Normally IR Chrome needs to slow down exposure by 1/3 to 2/3.

 

 

16 hours ago, Yves W said:

I have Fuji X-T20 and X-T2 and I'm happy with it.

I'd be so happy to see some shots from your cameras with the True Color SR filter ;)

[Yves W]

I only have UV, UVIVF and UV+B+G for insect vision

 

See what I wrote about Fuji :

Best & Worst Cameras for Conversion (Updates Welcomed !) - Page 3 - Techniques, Tests & Gear - UltravioletPhotography

[Nisei]

Thanks for posting these results.

I don't know if you've been reading my post about the same subject https://www.ultravioletphotography.com/content/index.php?/topic/5394-glass-alternative-for-lee-729-scuba-blue-filter/ but I've been on the hunt for a glass alternative to Lee 729 (and Lee 115) lately.

It's funny you got there just before me because after looking at transmission charts for the last couple of days I was about to order a QB19 as well.

Have you been looking at QB21? I'm also looking to get something close as possible to Lee 115 and wonder how QB21 will look.

[Marc del Piro]

3 hours ago, Nisei said:

Thanks for posting these results.

I don't know if you've been reading my post about the same subject https://www.ultravioletphotography.com/content/index.php?/topic/5394-glass-alternative-for-lee-729-scuba-blue-filter/ but I've been on the hunt for a glass alternative to Lee 729 (and Lee 115) lately.

It's funny you got there just before me because after looking at transmission charts for the last couple of days I was about to order a QB19 as well.

Have you been looking at QB21? I'm also looking to get something close as possible to Lee 115 and wonder how QB21 will look.

 

 

Hi Nisei,

 

I managed to get to it around the end of april, and funny thing, I also tought about the QB3 you ask for in the other post. I've been quite bussy lately, so I could not post it before :P I have not tried QB21, but I am pretty sure it will give some interesting results. So if you order it, I'd love to see the results.

 

 

2 hours ago, Nisei said:

Thanks!

Sorry, I missed these posts so I already responded to your findings in the other topic you started.

Are you buying from Nantong Foric Optical Glass company by any chance?

I usually do business with Tangsinuo Optoelectronic Tecnology company but I don't think they are selling QB19. They do have QB21 though.

 

 

 I will answer all your questions from the post you created about a glass alternative to Lee 729. It will be easier for readers to follow the conversation:

 

Nantong Foric Optical Glass is my trustful company for ordering filters. Their quality is great, not best, but the price is quite competitive. They are already selling a combination, QB19 + KG1. This combination came up after all the conversations I had with them, but I really don't know how well will this works. My take goes with KG3 and I see no point to combinate it with KG1. Feel free to send them a message and ask for a QB19 + KG3 1:1mm. I will be happy to see the results you get with the True Color Super Red (SR)! Also, if you don't mind stacking two filter rings, a 2mm each one would be suitable I guess.

 

If you want to order any other filter from them, they can personalize the thickness, which the Tangsinuo Optoelectronic Tecnology will not do unless you ask them for a bunch of the same filter at once.

 

Also, if you need other thickness, stacking two 1.5mm filters in their ring is feasible, but the cap of the lenses do not fit on the filter so you have to screw it in and out.

[JMC]

Marc. Nice results and thanks for sharing.

 

However I must point something out. You shared a transmission graph for QB19. In that graph, in that steep rise in the IR, the curve actually goes back on itself at one point which is obviously wrong. It's things like that which make me distrust provided spectra. I'm afraid without getting them measured properly its hard to know what you're actually getting.

[Nisei]

JMC,

I think both Marc and I are here because we like the aesthetic part of IR photography.

we both know about the color spectrum but aren't too concerned if nm measurements aren't quite within what's in the charts as long as it's looking good.

At least that's how I feel about it.

I prefer seeing a great picture instead of no picture and reading about calculations and theories.

But that's just my opinion.

[dabateman]

The problem with being too scientific,  is that the cameras themselves aren't standard in how they apply a custom white balance and will even vary with models from the same manufacturer. 

So I would have to agree, that in the case of attempting to mimic the Aerochrome look, a picture is required,  as well as noting which camera and model that your specific filter combination worked.

 

The Kolari Chrome filter is a set standard glass formulation,  but it looks nothing like Aerochrome on various cameras.  Case in point the Fuji camera with that filter is orange and not red.

[nfoto]

UVP is assumed to be a forum for exchange of information and viewpoints. Science and aesthetics combine since we are involved in photography, which is a visual channel of communication.  It is not about being in total agreement in a discussion. What we cannot accept is personal attacks.

 

The curve plot JMC referred to can at best be characterized as sloppily drawn (by hand?).  Its contents should be considered unreliable until verified by other sources.

 

We have the application procedure for membership to make people understand this forum is special and they should behave accordingly.

 

My current state of long Covid prevents me from being active on UVP at present. I hope this will change for the better over time. Meanwhile, I ask members to think twice before posting contentious matter.

[JMC]

Birna, firstly our thoughts are with you, and wish you all the best for a full recovery.

 

I'd like to believe we are all adults and can have and share opinions on here. UVP offers the global 'invisible light' imaging community an incredible resource, bringing together experts and enthusiasts across the whole imaging field from technical to aesthetic. Whenever someone contacts me asking about my imaging research, the first this I say to them is 'Have you signed up to UVP?'. Personally, I come here mainly to learn and share what I have learned with my research, with the hope that it can be of use to others. I'm a scientist first before anything else, so of course I'm going to comment when I see something that doesn't look right to me, just as I expect it to be called out when needed (I'm sure some of us remember the Hoya R72 spectra issue where I shared dodgy data). Rest assured, I don't feel 'attacked' by anything that has been said, and look forward to many more interactions on the forum in the future.

[Nisei]

7 hours ago, Andy Perrin said:

Colin, he attacked Jonathan for a reasonable criticism. I literally just quoted the moderators own reasons for founding the forum back to them. I wish Andrea would come back though. 

Andy, it really wasn't my intention to "attack" Jonathan so I'm very sorry and my apologies to the both of you if it came across that way.

To me it was clear from the start that this topic was about trying to come up with a cheap solution for the IR Chrome filter without being too scientific about it (even though Marc didn't just grab some blue filters without doing research first).

So when Jonathan said the result couldn't be trusted without measuring I was like "well yeah but that's not the point here, he came up with a glass filter to get the IR Chrome effect".

Hence I felt the need to respond and explain not everyone's here for scientific data. I have no scientific background, I just came here because IR photography has been my hobby since I was 17 years old (mid 1980's). I like experimenting with different filters and I've learned a lot through this forum. I do like reading transmission charts and figuring out if something could work to get the result I'm after. I think I was the first here to use the Lee 115 filter and discovered it could give very interesting results. This wasn't based on scientific data but just trying things out. If a member with more knowledge than me can figure out WHY it's giving this result then that's great and gives us all more insight but for me it's the result that's most important. If someone wants to get to the bottom of it and do measurements then that person is free to do so and scan the filters.

I just hope we can all respect eachother's reason why we're part of this forum and use eachother's input to learn and inspire.

[Marc del Piro]

19 hours ago, JMC said:

Marc. Nice results and thanks for sharing.

 

However I must point something out. You shared a transmission graph for QB19. In that graph, in that steep rise in the IR, the curve actually goes back on itself at one point which is obviously wrong. It's things like that which make me distrust provided spectra. I'm afraid without getting them measured properly its hard to know what you're actually getting.

 

I appreciate your insights on it.

 

After reading all the discussion above, I am pretty sure JMC somehow was asking where did I get that chart from. The answer is that it comes from Nantong Foric Optical Glass. I won't discuss about its correctness, I already said I am no physicist. Since I have no spectometer, I cannot share the real curve of the QB19. If you have one, the filter is quite cheap, so I encourage you to buy it and share the results with us; I assure I'll be the first one interested to listen to your findings. Anyway, I think nobody shared IR Chrome spectrum before, so without it, it is kind of pointless to obtain an accurate curve of the QB19 or the QB19 + KG3, since the point here is to obtain same results as Kolari's filter. Hence, my work has been based on a rough scientific method:

1. All photos taken with the same camera and lenses.

2. Taking pictures without balancing exposure.

3. Taking pictures by balancing exposure by histogram.

4. Working on a selfmade colourchart by the method explained on the post.

5. Having fun in the process.

 

Histogram is a descriptive method that makes sure the results are the same and that there is no interaction of human perception. It seems the histograms presented in my former comment are quite similar. I never did the inference calculations to know if with the IC 95% can obtain p > 0.05% or < 0.05% to know if H0 can be accepted or not. So yeah, I don't have the numbers to support my hypothesis.

All I can tell you is that I got where I got by studying histograms, so I bought QB19 after QB2 and QB3 after comparing their histograms to IR Chrome's. So what I am getting is what the histogram is showing, despite of whatever the real transmission curve is like. But again, I'd love to see the comparision of IR Chrome and QB19 + KG3 curves. IR Chrome looks like a thick (2.5mm?) KG3 filter with a blue coating, so both filters are physically different to begin with, and I expect different curves no matter what (am I wrong?).

 

I have to agree with Nisei:

 

15 hours ago, Nisei said:

 

I think both Marc and I are here because we like the aesthetic part of IR photography.

we both know about the color spectrum but aren't too concerned if nm measurements aren't quite within what's in the charts as long as it's looking good.

 

 

If I can get the same final results (histogram) as IR Chrome, then it's fine for me. My point here was to imitate the aesthetic part, not the curve itself. I might have not exposed the post correctly to state my final intention. I'll note this for the next post.

 

I am going for a photo trip this week. So by next week I will be able to post more insights on the filter.

 

[Nisei]

Thanks again Marc

Just a few observations:

1: I really want to encourage you to buy a WB card (I'm using the grey card from the ColorChecker Passport 2) to get more consistent results

 

2: You're talking about KG3 but the only company producing that is Schott

B+W used to sell the 489 filter which was Schott KG3 but other than Steve (user Cadmium here on the forum) I don't know of any source that sells real Schott KG3 in a filter ring

Any Chinese company claiming to sell you KG3 is selling an equivalent (usually referred to as GRB3).

So are you talking about real KG3 or a filter you got from a Chinese source?

 

3: If I white balance Lee 729 in-camera and open the RAW file in Lightroom, the WB slider is maxed out to the right and out of Lightroom's range. The same happens when using an orange filter, only then it's maxed out to the left.

I need to use Adobe DNG editor to adjust the WB slider and create a custom profile to make it useable in Lightroom (this is common practice for anyone using Lightroom to process color IR)

How come you don't have this problem? Or is it because you didn't do proper in-camera white balancing?

[Marc del Piro]

18 hours ago, Nisei said:

Thanks again Marc

Just a few observations:

1: I really want to encourage you to buy a WB card (I'm using the grey card from the ColorChecker Passport 2) to get more consistent results

 

 

 

1. I will. It is necessary to be consistent on publications. Right now I usually white balance on the blue sky, or white tshirt. It works fine for me. IR is quite tricky, and really, balancing on blue sky gave me really good results in the past. Then, Fuji allows for manual balancing. I will make a post specifically on how to deal with the filter WB and postprocess next week. I am on a backcountry trip to test the filter on different situations. 

 

18 hours ago, Nisei said:

2: You're talking about KG3 but the only company producing that is Schott

B+W used to sell the 489 filter which was Schott KG3 but other than Steve (user Cadmium here on the forum) I don't know of any source that sells real Schott KG3 in a filter ring

Any Chinese company claiming to sell you KG3 is selling an equivalent (usually referred to as GRB3).

So are you talking about real KG3 or a filter you got from a Chinese source?

 

 

 

2. My fault. It is GRB3. The chinese imitation of KG3. Indeed I will change it for future posts. 

 

18 hours ago, Nisei said:

3: If I white balance Lee 729 in-camera and open the RAW file in Lightroom, the WB slider is maxed out to the right and out of Lightroom's range. The same happens when using an orange filter, only then it's maxed out to the left.

I need to use Adobe DNG editor to adjust the WB slider and create a custom profile to make it useable in Lightroom (this is common practice for anyone using Lightroom to process color IR)

How come you don't have this problem? Or is it because you didn't do proper in-camera white balancing?

 

3. That is because I am a sourcerer of light magic. Or because Fuji works this way. Could be the filter physics. Who knows. I guess we need to test more cameras and light situations. With the IR Chrome, the slider is set at 50000°K if I set the auto WB, but by setting it manually, it stays as told to the camera. Results are the same, as I could see with the IR Chrome many times before. Normally I use 10000°K in camera, plus color adjustments. I will discuss this one on a specific post together with the former question. 

[Nisei]

Thanks for the response and looking forward to your future posts.

In the meantime, enjoy your trip to the countryside and let's hope you have perfect weather

[Andrea B.]

Marc, thank you for your excellent presentation. We welcome more such topics. I enjoyed seeing those brilliant colors and learning which filters produced them.

 

I also encourage you to find an IR-reflective white balance item such as PTFE. The reason is simply so that color results can be reproduced. That way a reader would have a baseline idea of what a particular filter or filter stack can produce.

BUT, posting a photo white-balanced against an IR white standard does not mean, however, that you cannot immediately follow it up with an artistic/aesthetic version of the same photo!! UV/IR work thrives on individual interpretations of the false colors. Please continue that, everyone. 

 

And that comment brings me to the following ---

"Our primary goals for Ultravioletphotography.com are the publication of scientific, documentary reflected Ultraviolet photographs and the delivery of reliable, high-quality technical information on Ultraviolet photography and related topics."

Those GOALS were written in 2013. We were simply trying to get this forum going. I apologize for not updating those Goals to include that ANY version of UV or IR or any UV/IR combo is welcomed because we also value aesthetics in this unusual kind of photography. This is especially true for emulations such as are shown in this topic.

 

There is only one section where we ask for white balance and that is in the formal Botanical section. It is important there to show that the floral UV-signatures can be reproduced with any gear. By "floral UV-signature" I am referring mostly to the UV-absorbing/reflecting areas and not so much to the false color. Let's note that applying WB to a floral UV-signature photo usually brings out the lights/darks better than other renderings.

 

I hope this soothes any frayed nerves.

I'm now going to go amend those Goals. And I'm going to consider removing any snark above.

 

 

 

 

 

[Nisei]

11 hours ago, colinbm said:

This is all very difficult Andrea.
None of these are SOOC, but require processing.

So the achieved flavour is, camera model, filters no. & brand/batch & CWB & finally processing to personal taste !

Colin, the first pictures he posted are SOOC.
If anyone's gonna post filter combinations to get false color the only fair way is to post SOOC JPEG files.
That means do proper in-camera white balancing, mentioning the camera make and model and no post processing.
That's how I posted my first results with Lee 115 as well. As soon as an image needs heavy post processing to get "that certain look" then there are too many variables and people don't get a fair idea what to expect when using the filter(s).
 

[colinbm]

10 hours ago, Nisei said:

Colin, the first pictures he posted are SOOC.
If anyone's gonna post filter combinations to get false color the only fair way is to post SOOC JPEG files.
That means do proper in-camera white balancing, mentioning the camera make and model and no post processing.
That's how I posted my first results with Lee 115 as well. As soon as an image needs heavy post processing to get "that certain look" then there are too many variables and people don't get a fair idea what to expect when using the filter(s).
 

Nisei, Exactly my thoughts too.

[Marc del Piro]

Andrea, I appreciate your insights. Indeed, a grey card would be preferred since this would be a more standard approach. I really don't have anything in my hands right now I can really use for such a purpouse, so I try to give as many insights on how I manually WB. I am interested in getting one in a near future, I will upload new results by then.

 

Nisei and Colin, indeed. SOOC images are the best way to go. JPG will show how the camera process the raw data, and RAW SOOC (no post process) will reveal how different the images are from scratch. I guess both of them are necessary, as well as trying to obtain a desired result with LR and seeing how different the sliders look. IMHO all methods are valid, they just desrcribe different parts of the process.

[Nisei]

Member A Stranger In the Wind did some great tests with colored filters. A great example of how it should be done: click

[Marc del Piro]

Following are the tests I realized these last days. I could not use a colour/grey card, but I tried to get the tests done in a methodological way. I will try to produce some tests with a card in a future.

 

Below there is the comparision between IR Chrome and QB19 + GRB3 with a WB done on a white sheet of paper. The order is as follows.

 

1. IR Chrome (WB on a white sheet of paper).

2. QB19 + GRB3 1:1 (WB on a white sheet of paper).

3. IR Chrome histogram.

4. QB19 + GRB3 histogram.

 

 

The images show slight differences on my screen, detectable by human eye. The major difference lies on the amount of blue and red on the right side of the histogram. This is translated on a slightly redish overall tone on the QB19 + GRB3 picture. Shadows also look a bit warmer. The sky tone is different on both images, and it is represented by the blueish hue on the right side of the histogram of IR Chrome.

 

Following this test, I did a manual WB by trying to match histograms as much as possible on camera. Fuji allows to fine tune Reds and Blues by addition or substraction.

 

1. IR Chrome (manual WB 10,000ºK R6 B6)

2. QB19 + GRB3 1:1 (manual WB 10,000ºK R4 B6)

3. IR Chrome histogram.

4. QB19 + GRB3 histogram.

 

 

 

Hues here tend to equalize, there is some blue coming out on the right side of QB19 + GRB3 histogram. The shadows keep being warmer in QB19 + GRB3.

 

The following pictures are the RAW images without any kind of postprocessing. The temperature is set to 10,000ºK. RAW image won't carry any adjustment on colour or fine tuning of hues, so it is possible to see what the camera actually sees, other than the process of the camera itself. Since up to this point, the value of the tests were including camera's postprocess of the RAW image. Both RAW are from the pictures of this last example.

 

1. IR Chrome (manual WB 10,000ºK)

2. QB19 + GRB3 1:1 (manual WB 10,000ºK)

3. IR Chrome histogram.

4. QB19 + GRB3 histogram.

 

 

The difference is still evident. Note how IR Chrome has a bit more contrast on the grass with the white flowers and shadows are tend to the blue.

 

Due to the clear differences I was obtaining with the QB19 + GRB3 1:1mm ratio, I decided to try the 1.5:2mm ratio. The next set of images are the results of the tests conducted with a manual WB of 10,000ºK. The following are exports of RAW images with no color postprocessing, there is no intervention of camera's postprocessing.

 

1. IR Chrome (manual WB 10,000ºK)

2. QB19 + GRB3 1:1 (manual WB 10,000ºK) (-0.21 EV)

3. IR Chrome histogram.

4. QB19 + GRB3 histogram.

 

 

 

Finally, results tend to be quite similar one to each other. The thicker filters keep IR light penetrating the sensor too much. Still, the colour of the blue filter is a bit different, which means it will give rise to slight differences on the final image.

 

In conclusion, I think those examples show how similar both filters are and the big amount of possibilities the chinese filter combination offers. Sadly, the 1.5:2mm barely fit on a single ring. A 1.5:1.5mm test should be conducted to determine if the results are similar or not. 3mm can fit in a single ring, but won't allow the cap of the lenses to hold properly. In a futur I will do a 1.5:1mm test, if it gives any interesting result I will post it, otherways I won't bother.

 

I encourage you to try the combination with your cameras and post the results here. I am curious how it will work with your lenses and cameras.