• Ultraviolet Photography
  •  

Infrared Infographic

Infrared
23 replies to this topic

#1 SteveCampbell

    Member

  • Members
  • 141 posts
  • Location: Australia

Posted 08 July 2017 - 03:59

Since there's no "Infrared Essays & Tutorials" section, I figured this was the closest place to post this!

I created an infrared infographic not too long ago to introduce people to infrared photography. Perhaps it's of some interest to people here? Perhaps in the future I may create one for ultraviolet photography too...

Original full-size source image and context

Posted Image
STEVE CAMPBELL
DeviantArt | 500px | Tumblr

#2 JMC

    Member

  • Members+G
  • 1,193 posts
  • Location: London, UK

Posted 08 July 2017 - 09:31

I think it looks great Steve - really well laid out and easy to follow.
Jonathan M. Crowther

http://jmcscientificconsulting.com

#3 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,271 posts
  • Location: USA

Posted 08 July 2017 - 15:58

Hello Steve and welcome to UVP. It is great to have some Infrared photographers on board. You have a nice IR portfolio.
Your graphic is quite well done and entertaining!! Love it!!

With reference to your graphic, perhaps these comments will be useful? Let me know.
  • Newer digital cameras have tended to have much stronger internal UV/IR-blocking filters, so IR exposures may be longer and noisier unless the internal filter is removed.
  • When both IR and UV light are permitted to reach a sensor, the Infrared light will "win" because there is so much more IR in sunlight than there is UV. So we really cannot make a IR+UV photo. Only the IR is recorded - no, mostly only the IR is recorded. Any obvious UV-absorbing areas will appear washed out or non-existent. So the very bottom right-hand on your graphic is technically true -- both IR and UV reach the sensor when using a typical dual-bandpass filter like U-360 or similar -- but the UV in such small amounts as to be almost useless.
  • If you reference Photoshop®, then I would suggest making sure you use this form with the capital 'P' and Registrtion ® mark to avoid having Adobe come after you. And they will if your graphic is ever published!!!!!! Fortunately, the fix is really easy. B)
  • The false colour effects you reference are very very dependent on two things: (1) the particular IR-pass filter being used and (2) the white balance setting applied either in-camera or during raw file conversion. We really cannot make any general comments about IR false colour without reference to the filter in use. IR-pass longpass filters which cut in at a level beyond 800 nm or so only produce monochrome files, usually converted to just B&W. IR-pass longpass filters which cut in at a lower visible level between 600 - 700 nm can give a variety of false colours depending on conversion. You could get a blue/yellow or blue/gold effect. You could get a pale green/blue effect. The IR false colours you use in your graphic are generally associated with the "typical" IR-pass filter like a Hoya R72 and associated with a photoshop channel-swapping conversion. So I suppose it might be useful to your viewers to know that different results might be obtained with different filters and conversion techniques? ;)

Andrea G. Blum
Often found hanging out with flowers & bees.

#4 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 08 July 2017 - 16:56

View PostAndrea B., on 08 July 2017 - 15:58, said:

When both IR and UV light are permitted to reach a sensor, the Infrared light will "win" because there is so much more IR in sunlight than there is UV. So we really cannot make a IR+UV photo. Only the IR is recorded - no, mostly only the IR is recorded. Any obvious UV-absorbing areas will appear washed out or non-existent. So the very bottom right-hand on your graphic is technically true -- both IR and UV reach the sensor when using a typical dual-bandpass filter like U-360 or similar -- but the UV in such small amounts as to be almost useless.

UG1 and U-360 (and quite a few other filters) depend on the UV as well as the IR, they are quite often used for IR photos with blue/lavender skies right out of the camera.
This would not be possible without the UV transmission of those filters.
These are UV + IR photos.

Examples, and comparison graph.

Hoya U-360 2mm:
Attached Image: Hoya_U360_test_compare_1280.jpg

Schott RG9 2mm:
Attached Image: Schott_RG9_test_compare_1280.jpg

Schott RG850:
Attached Image: Schott_RG850_test_compare_1280.jpg

Attached Image: U360_2mm_RG9_2mm_RG850_2mm_compare.jpg

#5 Andy Perrin

    Member

  • Members+G
  • 3,665 posts
  • Location: United States

Posted 08 July 2017 - 21:45

Yep, those are true UV+IR photos. Do you see anything interesting with that combo, Steve? I have never tried it. It looks pretty similar to ordinary IR in the examples you posted?

#6 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 08 July 2017 - 22:34

Only one of those is a dual band UV + IR photo. The U-360 shot.
It doesn't look the same as 'ordinary' IR only photos, which have brown or black skies.
The thing that makes the U-360 dual band shot 'interesting' compared to IR only (longpass filters above 780nm to 850nm) and false color IR shots (longpass filters below 780nm to 850nm), is that it is an IR photo with a blue sky, with no channel swapping.
Nothing new, just showing how UV can work mixed with IR.
And that is an example of only one type of dual band filter, there are quite a few, BG3 being another often used dual band IR filter that has UV mixed in.

Below is an interesting example, using the dual band Hoya U-360 with and without the UV transmission subtracted from the mix.

Hoya U-360 2mm:
Attached Image: Hoya_U360_2mm_1280.jpg

Hoya U-360 2mm + Schott GG400 stacked (this removes UV from the U-360 transmission, leaving only the Red/IR part of the photo):
Attached Image: Hoya_U360_2mm_x_Schott_GG400_1280.jpg

Edited by Cadmium, 08 July 2017 - 22:53.


#7 Andy Perrin

    Member

  • Members+G
  • 3,665 posts
  • Location: United States

Posted 09 July 2017 - 01:14

I guess I've never cared all that much about the sky color (since one can just channel swap or even just replace the color if so minded), but I wanted to know whether UV/IR shows anything new about the scene itself that you wouldn't see using either a UV or an IR photo alone? Have you ever seen anything?

#8 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,271 posts
  • Location: USA

Posted 09 July 2017 - 01:32

I'm not convinced completely by the preceding examples. While there might be some false colour differences those are somewhat irrelevant. What needs to be shown are B&W versions in order to compare tonalities. But then I am somewhat of a purist. Mind you, I'm not claiming no UV gets through, but one look at the trees and grass in the U-360 frames (first two here) shows IR photos. (We need also have a UV-only to compare to.)

Here are some luminosity maps using (128,128,128). The Red+IR is very much like the UV+IR. Interestingly, it is the porta-potty which shows a rather dramatic difference between the IR Only frame and the other two, at least on one side.
Attached Image: u360_LUM01.jpg
Attached Image: u360gg400_LUM01.jpg
Attached Image: rg850_LUM01.jpg
Andrea G. Blum
Often found hanging out with flowers & bees.

#9 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 09 July 2017 - 04:12

The U-360 cuts off UV at 400nm. Remove the UV and you are left with the 700nm+ basically.
Because the IR is so strong, especially with foliage, and foliage is so dark in UV, you will get basically the same IR with no UV mixed in with the foliage (as seen with my last two shots).
The UV shows up in the sky, because skies have a lot of UV in them, but are black in IR.
Where are you getting your "IR Only" image from?
Just because the gray tone is about the same between your IR + UV and Red + IR doesn't mean the color is the same.
My test proves where the blue is coming from.

Would you then also argue that there is no UV in a dual band UG5 filter or a UG5 stack?
:-)

Edited by Cadmium, 09 July 2017 - 04:21.


#10 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 09 July 2017 - 04:43

I don't know where you are getting your B&W images.
If I take the two shots above, and desaturate them, they look much different from each other than the gray versions you show (see below).
So you are saying that the U-360 shots don't have UV influencing the color of the sky? Where do you think that blue is coming from?
I cut off the UV, and the blue is gone, and the image is reduced to a fairly typical false color IR shot.
People have been using UG1 and U-360 for eons now for this exact kind of thing, and it will not work without a full spectrum camera that is sensitive to UV.

U-360 2mm desaturated.
Attached Image: Hoya_U360_2mm_DS_1280.jpg

Hoya U-360 2mm + Schott GG400 stacked desaturated:
Attached Image: Hoya_U360_2mm_x_Schott_GG400_DS_1280.jpg

Edited by Cadmium, 09 July 2017 - 04:45.


#11 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 09 July 2017 - 05:09

Andrea, OK, I figured out where you got the IR only shot, from the RG850 shot.
The porta-potty looks different mostly because of the time of day, angle of sun. That shot was taken in the morning, the more recent U-360 +/- UV tests were shot in the afternoon.
Take a look at the other morning shots, you will see what I mean.
The reason there is a difference between the RG850 shot and the U-360 shots should be obvious from the graph I posted.
The RG850 filter cuts off the 700nm range and below, so there is no hint of false color, all the camera is seeing is monochrome IR.
Something else to keep in mind with the U-360/UG1 filter/curve is the proportion of the UV to 700nm+ range IR, which is greatly attenuated compared to that of other IR longpass filters,
the RG850 filter has a much stronger IR transmission for example.
This relatively strong UV transmission to IR transmission ratio ads to the ability of the UV to influence the color of the sky.
Other false color range IR longpass filter shots will also have lighter desaturated skies than does an 850nm longpass filter shot which typically have very dark or black skies.
For example, here is a Schott RG665 (different scene), desaturated.
Attached Image: RG665_ds_800.jpg

Edited by Cadmium, 09 July 2017 - 05:51.


#12 SteveCampbell

    Member

  • Members
  • 141 posts
  • Location: Australia

Posted 09 July 2017 - 07:06

You you all kindly for the comments!

Andrea:
- Many thanks!
- I briefly pointed that out unmodified cameras only pass small amounts of IR, but perhaps I could make it clearer for future versions
- I suppose for a U-360 the majority of the IR will register in the red channels since the U-360 only passes ~700-800nm, which registers primarily in the red channel, while a blue-channel IR signal is usually caused by >800nm due to Bayer array blue pixel filters having low transmission below ~800nm. I'd imagine therefore that the majority of the blue-channel contribution would be from UV, which registers in both red+blue with blue>red. I would expect the U-360 would cause a quite different color distribution, but not as evident a difference in luminosity.
- Will fix ®!
- I should definitely mention on the infographic that the images represented 700nm+ IR via R72-type filters. I suppose I just assumed this appearance to be "standard" IR as it represented the typical IR range (>=700nm), after white-balancing. I would imagine Bayer array RGB color filters differ between manufacturers, so perhaps RGB distributions of infrared signals vary as well. I've always considered anything passing any more than ~5% of light <700nm to be red+infrared. I'll try to clarify this in future versions of the infograpic.

I think Cadmium alluded to some aspects of the above in his posts, so apologies for any repetition!
STEVE CAMPBELL
DeviantArt | 500px | Tumblr

#13 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 09 July 2017 - 08:03

U-360 and UG1 definitely transmit IR all the way up to 1100nm. Although it may look very minimal above 800nm, that curve is not so different than the curve of RG9, given the proportion to the 700-800 range of the U glass transmission.
Exposure time is longer than RG9. Exposure is even slightly longer when the UV is subtracted/blocked with the GG400 filter.
Photos with UG1 and U-360, as well as many other dual band filters that have their feet in both UV and IR are definitely a mix of UV and IR.
Attached Image: UG1_U360_1mm_2mm.jpg

Edited by Cadmium, 09 July 2017 - 08:04.


#14 SteveCampbell

    Member

  • Members
  • 141 posts
  • Location: Australia

Posted 09 July 2017 - 10:23

The point is certainly well taken that the U-360 transmits rather more heavily >800nm than I appreciated, although the UG1 looks pretty good if you take into account that the Y-axis appears log-scaled
STEVE CAMPBELL
DeviantArt | 500px | Tumblr

#15 bobfriedman

    Member

  • Members
  • 396 posts
  • Location: Massachusetts

Posted 09 July 2017 - 10:52

form my past.. UG-1

Nikon D200 ,Nikkor 20mm f/3.5 Ai-S
1/30s f/8.0 at 20.0mm iso250
Posted Image


and BG-3

Nikon D200 ,Nikkor 20mm f/3.5 Nikkor-UD Auto
1/200s f/8.0 at 20.0mm iso125
Posted Image

Edited by bobfriedman, 09 July 2017 - 10:53.


#16 bobfriedman

    Member

  • Members
  • 396 posts
  • Location: Massachusetts

Posted 09 July 2017 - 10:56

BG-3 Spectrum

Posted Image

#17 Andy Perrin

    Member

  • Members+G
  • 3,665 posts
  • Location: United States

Posted 09 July 2017 - 13:14

View PostSteveCampbell, on 09 July 2017 - 10:23, said:

The point is certainly well taken that the U-360 transmits rather more heavily >800nm than I appreciated, although the UG1 looks pretty good if you take into account that the Y-axis appears log-scaled

Diabatic graphs are scaled on the y-axis according to y_new = 1-log10(log10(1/y)), which results in a middle portion from 0.1 to 0.9 that is nearly linear, and log-scaling for <0.1 and >0.9 (roughly).

Edited by Andy Perrin, 09 July 2017 - 13:15.


#18 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,271 posts
  • Location: USA

Posted 09 July 2017 - 13:29

Well, I've said it before and I'll say it again, do not be misled by false colour. We just do not know the specific properties of Bayer dyes. (At least so far on UVP, nobody has ever come up with any kind of research paper or documentation of Bayer dyes.) And given that the derived false colour in any UV or IR photo is dependent on so very many factors (conversion app; white balance tools; quality of light which is in turn dependent on time of day, geographic location, altitude; camera software and so on and so on), one needs to be careful about what conclusions are drawn.

What we can look at in any UV or IR or mixed photograph is areas which are UV-absorbing or UV-reflecting or somewhere in the middle. Ditto for IR-absorbing or IR-reflecting. Even this depends somewhat on the curves applied during conversion and the contrast choices made during editing.

Yes the UV+IR photo above differs from the Red+IR photo above -- but not by much at all. They are both basically IR photos when areas of IR-absorbtion and IR-reflection are examined. Bob, your UG1 and BG3 photos are basically IR photos also.

(1) If you want to really see the differences between the photos, a trip through Raw Digger is needed so you can look at the Raw Composite and the RGB histograms of the raw data.
(2) To better illustrate how the various dual band-pass filters work, you should include some kind of light and dark UV or IR reflective target.

*****

Added later: Well....here I am with a bag full of filters and some reflective standards, so I'd better go shoot them to illustrate this and see what conclusions we can all make. It will take me a bit of time to get them posted. There are already a few such shots posted somewhere here in some of the posts showing filter test results. But I've never posted an entire catalog of them. <She leaves now to go gather up the gear and shoot.>
Andrea G. Blum
Often found hanging out with flowers & bees.

#19 Cadmium

    Member

  • Members+G
  • 4,053 posts

Posted 09 July 2017 - 17:03

Andrea, You can call it false color if you want to, so many things are, however, the blue sky in the U-360/UG1 images is a result of the camera seeing UV through the filter.
If you use a camera that has the UV blocked, the U-360 image changes to a 700nm+ IR photo. Put the UV back in the mix and you get the blue sky.
How you argue that the UV is not part of the mix is astonishing to me.
I will leave it to you to prove that point.

#20 Bill De Jager

    Member

  • Members
  • 281 posts
  • Location: Northern California, USA

Posted 09 July 2017 - 18:29

Steve, welcome here. I like the graphic very much as it present the basic ideas in very comprehensible terms.

My only comment is that the term infrared has different meanings here vs. with a lay audience, the latter including people looking into infrared photography for the first time (and who I suspect are the target audience of your graphic). We understand that by infrared we normally mean the near-infrared which is quite different from the longer-wave infrared that makes up everyday typical thermal radiation. However, most people have no concept of that and think of infrared as being just one "thing" that's equal to radiant heat. That's a misconception which is reinforced in almost every article that mentions infrared outside of a specialized photographic context. Readers may even have seen photos or video taken in the thermal infrared band, which have nothing to do with the descriptions in your graphic.

Therefore, laypeople are likely by default to think of infrared photography as being thermal photography rather than photography of just a different band of reflected light. So it might be helpful to include a brief mention at the onset that the infrared being discussed in the graphic is a particular sort that's close to visible light and which (at normal temperatures) is reflected off objects rather than being emitted by them based on temperature.

Edited by Bill De Jager, 09 July 2017 - 18:30.