• Ultraviolet Photography
  •  

Accidental IR-filter, a bit like Aerochrome?

Filters
36 replies to this topic

#1 UlfW

    Ulf W

  • Members
  • 57 posts
  • Location: Sweden, Malmö

Posted 02 August 2017 - 19:55

This spring I was mainly collecting filter for IR-photography. (This was before I leaped over to the UV-side)
After reading about filter problems for simulation of Aerochrome, I became interested in comparing a cheap Orange filter against my B+W Yellow-Orange 040 (orange) and ordered one.

When the package arrived it also contained a second green filter in a slightly worn outer box.
I assumed it was a reject item that they sent as bonus.
This is the green filter I got:
http://www.ebay.com/...m-/121537504290

I never got around testing either filter and assumed that I never would have any use for the green one, until last week.
Then a friend helped me doing transmission measurements on most of my filters and the green filter gave this response:
Attached Image: Pixonyx Green 77mm copy.jpg
The transmission graph with the big IR-leakage sparked my interest and motivated me to take some test images with the green Pixonyx filter.

This is the result:

Attached Image: _MG_6381_v1 copy.jpg

Attached Image: _MG_6399_v2.jpg

I have not processed the images very much. Appart from a white balance against the clouds, I have adjusted the saturation and slightly tweaked the hue of foliage and sky.
Directly after white balance the foliage got a very saturated bright magenta tone and the sky has a pale slightly blue-green tone.

Could this filter be useful for simulating a Kodak IR-film?

Used camera and lenses:
Canon EOS 60D, full spectrum modified by replacing the sensor filter and dust shaker window with a sensor window from Astronomik.de
Canon 24-105mm/4.0L, Canon EF-S 10-22/3.5-4.5
Ulf Wilhelmson
Curious and trying to see the invisible.

#2 JMC

    Member

  • Members
  • 148 posts
  • Location: UK

Posted 02 August 2017 - 20:42

How cool. Just checked the eBay link though and they don't ship to the UK :(

#3 OlDoinyo

    Member

  • Members(+)
  • 477 posts
  • Location: North Carolina

Posted 03 August 2017 - 02:47

Aerochrome produces an IRG-->RGB cross-sampled image. One property of such an image is vegetation that appears reddish, because vegetation is a good reflector of infrared light and infrared light is mapped to red in an IRG image. There are other kinds of image which can also produce red vegetation, for reasons similar (but not identical) to those circumstances which produce it in an Aerochrome image. Such images do not reliably reproduce the other properties of an IRG image---so do we call them similar? The ultimate answer depends on how much one cares about these differences. It is possible to produce real IRG images with digital gear, if that is what is desired, but the process is more complex than simply slapping a filter over the lens.

#4 UlfW

    Ulf W

  • Members
  • 57 posts
  • Location: Sweden, Malmö

Posted 03 August 2017 - 06:43

JMC -
The UK is not on the exclusion list and they state that they ship worldwide.
UK is not in the list of selectable destinations, but they might still accept an order if you ask them.
If the answer is no, I might help you as I think I'll buy a second filter from them soon.
The shipping cost of an additional filter is only 0.3€.

However we do not know if the filter material always is of the same type for such a cheap filter. This is a risk.


OIDoinyo -
I do not think that it is possible to do any simulation prefect, only more or less accurate.
A map always differ a bit from the real world.

The results with this filter might still be interesting as it rejects blue and red mainly transmit IR and green from the source before it has contaminated the image data in the infrared-dominated areas.

I agree that it is a quick and dirty way to just slap a filter over the lens.
The result from the filter is absolutely not as from a real Aerochrome, but might still be useful for artistic pictures.
That is why the title was "a bit like Aerochrome".

Do the quest to simulate the Aerochrome with digital technology have a real scientific reason?
How close is good enough to be acceptable as a good simulation of the film?

Is grain, similar to the film needed?
Is a film-like intensity response needed or is a gamma response enough?

Edited by UlfW, 03 August 2017 - 10:51.

Ulf Wilhelmson
Curious and trying to see the invisible.

#5 JMC

    Member

  • Members
  • 148 posts
  • Location: UK

Posted 03 August 2017 - 10:12

Thanks UlfW I'll keep that in mind - I assumed it was not an option to buy from them direct as UK wasn't on the list for shipping. I also have a few old filters up in the loft, and I think it'll be worth me running them through the UV Vis spectrometer at work to see what they look like before going ahead and ordering more. While my spectrometer only goes up to 800nm it'll be interesting to see what they look like in the IR.

#6 enricosavazzi

    Member

  • Members
  • 380 posts
  • Location: Stockholm, Sweden

Posted 03 August 2017 - 12:03

Good find, I ordered this and a few others. At this price, we cannot go wrong.

eBay did accept shipment to Sweden, but gave me no shipment discount for an order of 5 filters (it just added up the individual shipment costs five times, about the same total shipment cost as the total items' price). I sent the seller an email via eBay about this.

Edit - The seller refunded 24.80 € excess shipment charges.

Edited by enricosavazzi, 03 August 2017 - 12:57.

-- Enrico Savazzi

#7 OlDoinyo

    Member

  • Members(+)
  • 477 posts
  • Location: North Carolina

Posted 03 August 2017 - 14:33

View PostUlfW, on 03 August 2017 - 06:43, said:

Do the quest to simulate the Aerochrome with digital technology have a real scientific reason?
How close is good enough to be acceptable as a good simulation of the film?
Is grain, similar to the film needed?
Is a film-like intensity response needed or is a gamma response enough?

Today, the primary technical applications for IRG imaging are for vegetation health monitoring and archaeological prospecting. Much of this work is done from satellites, many of which are capable of other kinds of multispectral imaging as well. The images these satellite cameras produce are not simulations; though they may not look exactly the way a film image would have, they encode the same kinds of information (and in fact, they do look somewhat similar.)

There are historical preservationists who try to reproduce the exact characteristics of a particular film emulsion by digital means. I have never attempted to do this, myself--it is very difficult to do convincingly and I would rather just try to make images that are pleasing.

It may amuse you to read this forum:

https://www.flickr.c...57601045395608/

Among the things described is a method to get reddish foliage by reprocessing an ordinary color image--no infrared needed at all!

#8 UlfW

    Ulf W

  • Members
  • 57 posts
  • Location: Sweden, Malmö

Posted 03 August 2017 - 21:00

Thank you for the link. The first look there was amusing.

I also would prefer to make images pleasing.
I hope I can do that with the filter I found, with some practice.

Of the IR-type images I have found so far, my two absolute favourites are

a image by Bjørn of a field with a narrow road in the middle. This, I think, is a real AIR.
And a second one of a spruce? in fog. Possible a digital dual exposure composition.

I think I have seen both on this forum, but cannot find them just now. Help please?

Both are very nice in their composition.
They also have a very controlled well balanced color scale without the typical highly saturated color often seen in bad AIR-simulations.

These images are for me some kind of reference images.
If I could come close to those qualities with my pictures I would be happy.
Ulf Wilhelmson
Curious and trying to see the invisible.

#9 Cadmium

    Member

  • Members(+)
  • 1,002 posts

Posted 04 August 2017 - 04:37

Interesting filter.
Typically, with EIR/Aerochrome, things that are red in visual look yellow. So test that out, see if red things end up being yellow.
For example, a good test is a non illuminated red plastic tail light of a car, those should look yellow. If they stay red in the photo, then...
Often red clothing, will be yellow... look for that.

See the yellow tail lights below?
Attached Image: DSC_5341_ER_LCH_1_1280.jpg

Yellow cars and sign in this pano were visually red.
Attached Image: Maple_Down_Town_Pana2ff_color_1920.jpg

Promaster P01 and Petri P01 are green filters, they both work the same, they may not be the same filter glass as the Pixonyx Green filter, but seem to have some similarity with how they function on a full spectrum camera.
This is a Promaster P01 shot, white balanced only. Note that with the P01 filter the red barn stays red, sorry, no yellow barn.
Attached Image: DSC_0068_P01_1280.jpg

Edited by Cadmium, 04 August 2017 - 05:25.


#10 Cadmium

    Member

  • Members(+)
  • 1,002 posts

Posted 04 August 2017 - 06:01

View PostUlfW, on 03 August 2017 - 21:00, said:

And a second one of a spruce? in fog. Possible a digital dual exposure composition.

I think I have seen both on this forum, but cannot find them just now. Help please?

Maybe this may be tree you are referring to? I think this is a fir tree though.
https://www.flickr.c...157609825290683

#11 UlfW

    Ulf W

  • Members
  • 57 posts
  • Location: Sweden, Malmö

Posted 04 August 2017 - 06:18

Noted.
OK, I see the yellow lights.

Pixonyx green do not make visually red objects yellow directly.
The image is just white balansed.
Attached Image: _MG_6453.jpg

Are those pictures real AIR or simulated?
Ulf Wilhelmson
Curious and trying to see the invisible.

#12 UlfW

    Ulf W

  • Members
  • 57 posts
  • Location: Sweden, Malmö

Posted 04 August 2017 - 06:33

This explains a lot:
http://www.candelapr...phy/aerochrome/
(Sorry for posting an external link, but I do not know how else to do this better)

To really simulate the AIR sensitivity pattern must be quite complex!
Ulf Wilhelmson
Curious and trying to see the invisible.

#13 UlfW

    Ulf W

  • Members
  • 57 posts
  • Location: Sweden, Malmö

Posted 04 August 2017 - 06:49

That is the tree I was referring to. A very nice image!

My botanical knowledge isn't big enough to tell the difference from that picture.
It looks like a christmas tree without ornaments.
Ulf Wilhelmson
Curious and trying to see the invisible.

#14 Cadmium

    Member

  • Members(+)
  • 1,002 posts

Posted 04 August 2017 - 08:52

My first two were processed, simulated from Wratten/Tiffen Yellow #12 (minus blue) filters.
Kodak EIR/Aerochrome film will also render yellow from visual red as do the simulations.
If you do a two shot composite (visual + IR) you will also get yellow from red.
Basic transposition from IRG-->RGB (as OlDoinyo mentioned). IR > Red, Red > Green, Green > Blue.
Yellow being part of green.
Many of the more 'normal' green filters (other than the P01 which is slightly non normal) will look like this below 060 green filter, I have tried them all except the one you are testing.
P01 is a little different than most Green filters, but here is an example of a more common green filter:

B+W 060 green (purple/pink foliage and green skies...and also no yellow tail lights of the horses):
Attached Image: DSC0275_4_green.jpg

Various green filters
Attached Image: DSC_1593a_4.jpg

Edited by Cadmium, 04 August 2017 - 08:55.


#15 JCDowdy

    JCDowdy

  • Members
  • 1,019 posts
  • Location: Arlington, TN

Posted 04 August 2017 - 14:08

View PostUlfW, on 03 August 2017 - 21:00, said:

I think I have seen both on this forum, but cannot find them just now. Help please?

Yes, they both are also on UVP. See: Digital EIR Experiment

#16 enricosavazzi

    Member

  • Members
  • 380 posts
  • Location: Stockholm, Sweden

Posted 04 August 2017 - 15:56

These filters are far more expensive, but it is interesting to know that they are also available:

http://midopt.com/fi...multi-bandpass/
-- Enrico Savazzi

#17 Cadmium

    Member

  • Members(+)
  • 1,002 posts

Posted 04 August 2017 - 20:43

Here is the graph of Schott VG green filters, the only one I have is the VG9, and it only starts to show some red (but not much) at 1mm thickness.
Attached Image: Schott_VG4_VG5_VG6_VG9_VG10_VG14_2mm.jpg

Attached Image: VG9_two.jpg


Here is a comparison of VG9 with P01.
Attached Image: green_4_1280.jpg

Visual (Schott BG38).
Attached Image: Visual_Schott_BG38_640w.jpg

#18 OlDoinyo

    Member

  • Members(+)
  • 477 posts
  • Location: North Carolina

Posted 05 August 2017 - 02:45

Some of the images with green skies and purplish foliage remind me not so much of IRG as they do of RBG (the Lomochrome look.)

#19 Cadmium

    Member

  • Members(+)
  • 1,002 posts

Posted 05 August 2017 - 05:20

VG9 1mm is about the same as VG6 2mm. So VG4, VG5, and VG10 might show a lot more red foliage than the VG9 1mm, however, the 'on request' makes those hard to acquire.
Comparing Pixonyx.Green with P01 will be interesting.

#20 Cadmium

    Member

  • Members(+)
  • 1,002 posts

Posted 11 August 2017 - 23:16

Here is Pixonyx Green, direct form Germany. This looks similar to your lavender foliage example in post #11. Tail light is red, not yellow.
I have not processed this at all other than white balance with CNX2 on gray. Whole frame Marquess was about that same.
No mater how I white balanced it, i could not get red foliage like is shown in post #1.
The main difference I note with this green filter is that it has blue skies instead of green skies.
Attached Image: Pixonyx_Green_1280.jpg