Jump to content
UltravioletPhotography

Digital EIR experiment


Recommended Posts

Hi,

 

I hope it falls under Infra-red category :)

 

I believe some of you know the Kodak Aerochrome EIR film, which was developed for military use and records IR and visible too. I think it became really famous recently when Richard Mosse used it to his project: The Enclave. If you haven't seen the stunning images he captured with it, then it worth to Google it :)

 

I find a quite good how-to on the internet and gave it a try. I tried many filters but so far a sepia filter works for me the best. For the photos I've used the Sony E 18-55 kitlens as it blocks UV very well as I don't have a UV block filter (yet). All shots are handheld, f8.0 (hence a slight hot spot here and there) and shutter speed around 1/50-1/100.

 

As of now I am trying to unify my workflow to be able to get roughly the same colours, but it seems even if I use exactly the same settings the colors are different :)

 

You can see that toward the sides there is a huge CA. It's most likely because the focus shift in VIS and IR, so a lens with almost no focus shift is required for the best results.

 

 

deir2.jpg

deir3.jpg

deir4.jpg

Link to comment

Making emulated 'infrared colour' pictures is easy, making them appear as the beautiful (ancient film from the late '60 and 'early -70s, E-4 process and long since discontinued) Infrared Ektachrome IE is pretty tricky indeed. One easily winds up with something replicating the latest Kodak EIR film, which was designed for AR-5 chemistry but in a pinch could be done in E-6 (of course, meaning 99% of the users put it through the non-optimal E-6 and were thanked by pictures having extremely garish colours and reds that appeared to be applied by a spray can).

 

I only have had some success with my quaint Fuji S3 Pro UV/IR LE using an orange (O-56 equivalent filter) towards getting a more believable IE-type rendition.

Link to comment
  • 1 month later...

We now have a group on Flickr where some information on the making of digital IRG-->RGB images is shared, as well as some of the images themselves:

 

https://www.flickr.com/groups/2678889@N25/

 

It must be pointed out that there are many ways of creating infrared images where vegetation happens to turn out red, but few of these produce authentic IRG images (i.e. what the old film produced.) I do not like the term "digital EIR" that some use to label such images; after all, we do not call our conventional images "digital Kodachrome" or "digital Velvia" or whatever; and examination of many actual color infrared film images has convinced me that there is no single "EIR look," anyway; different photographers have their own favorite filters, exposure preferences, and workup protocols, and the results vary quite widely. For me, the endeavor is about keeping a kind of imaging alive, rather than imitating a particular historic product.

Link to comment

Hi Clark

I don't quite get this......."For me, the endeavor is about keeping a kind of imaging alive, rather than imitating a particular historic product."

It seems contradictory ?

Col

Link to comment

Colin-

 

Perhaps an analogy will help clarify. For years, the main medium of b&w IR photography was Kodak HIE, which as we all know, has a very particular look which goes beyond its being merely IR-sensitive. Some traditionalists (call them film chauvinists, if you will) still assert that a proper b&w IR photograph should look like an HIE shot, and that failure to do so is a fault. Most digital IR shots have very different texture and tonality from HIE photos, although some digital photographers have tried, through various post-processing ploys, to emulate the tonality, graininess, halation, etc. which were the film's hallmarks. Other camps were equally devoted to Efke IR820 Aura or to a Konica emulsion, each of which had their own looks; but all of them had in common that they were means of imaging infrared light, as is digital infrared. Digital b&w infrared carries the basic tradition forward, even if its results do not inherently resemble those of any particular one of the old emulsions.

 

With IRG imaging, it is less obvious, because so few IRG emulsions were ever marketed, and those only by a single firm. However, the principle, for me, still applies, and the mission has more urgency. Modern consumer gear does not casually produce this type of image either out of the box or with a simple filter switch. High-priced cameras on satellites still do take IRG images, but such kit is far out of the reach of almost all of us. Thus, a type of image that photographers could casually produce in the film era threatens to disappear unless someone actively develops the strategies for producing them with consumer gear, and this last is what I have been trying to encourage. Once the basics are perfected, there is plenty of time for anyone to decide whether they wish to emulate a particular historic emulsion; but it won't happen unless the basics are gotten right.

 

JC-

 

I have played with Foveon color IR myself (and posted two such shots recently.) Vegetation does appear red, but these are not IRG images, as there is IR in all three channels. Technically the mapping is [RI][GI'][bI"]-->RGB, where I, I' and I" are differing infrared fractions. There is no cross-sampling in the visible channels, so the resemblance to IRG images is, in my opinion, superficial.

Link to comment

I thought the IRG PP scheme was IR mapped to Red, Red mapped to Green and Green mapped to Blue.

My understanding is that two photos must be taken, an IR filter photo and a spatially indexed normal Vis photo.

The IR photo must be converted to B&W to combine all false IR color channels into one.

The Blue channel of the matching Vis spectrum photo is discarded.

I do not understand what you are saying by "....I, I' and I" are differing infrared fractions" but perhaps I am confused?

According to the spousal control unit, this is a common operating condition. :D

Link to comment

Thanks John for bringing my link forward, much appreciated.

Thanks Clark for the explanation, I don't think my, Sigma Foveon Rich Full Spectrum with CPL, will meet the criteria for your IRG group.

Cheers

Col

Link to comment

JC-

 

I, I', and I" are in principle different because the top sensor layer is on the surface, whereas for the other strata infrared photons must traverse either one or two overlying layers before being recorded. Think of this as either zero, one, or two layers of a filter of admittedly unknown transmission curve. We know that the layers cannot be perfectly transparent to IR, because if they were, they would also be incapable of recording it. Actually, it is more complicated than that, because the camera's firmware must do some math even to extract a simple RGB image from the data coming off the sensor; the top layer receives all the incoming photons of whatever wavelength while the lower two layers are progressively shielded from shorter wavelengths.

 

The two-exposure method you cite is indeed one way to create a digital IRG image, and the easiest to understand conceptually. For converted cameras with a conventional sensor, a single-exposure method is also being explored. This is harder to explain and the math to allow this was worked out by a Canadian photographer, J.W. Wong (Infrachrome) on Flickr. The single-exposure method is not as radiometrically precise as the two-exposure method, but decent-looking results have sometimes been obtained and the choice of subject matter is less limited than that for either sequential or concurrent two-exposure methods.

 

Colin-

 

If you are so inclined, you actually could use your camera to create an IRG image with the sequential two-exposure method. Hot mirror for the first exposure, black-IR filter for the second, and don't change lens settings in between (use manual mode.) The rest is not difficult. I have done this once with my SD14. (B+W 486 and 093 filters.) You might find it instructive to compare the results to your no-filter images.

Link to comment

Clark,

 

Thanks for answering my question. I have seen some of Wong's(Infrachrome) stuff before. He has a great diagram of the two photo method I had bookmarked. His other single shot ~EIR PP technique you mention is one I had not seen but now makes sense to me, THANKS!

 

By the way, his posts on the Restoration / Resurrection of a Kodak DCS 460CIR Color Infrared Camera make interesting and entertaining reading. It is a sad sometimes how rapidly evolving technology so quickly renders such unique instruments obsolete.

 

- JD (not JC) ;)

Link to comment

My approach to false-colour IR in the digital domain always has been with a Fuji S3 Pro UV/IR LE camera ("broad spectrum") and an orange filter (O-56). This produces excellent emulated IE in single captures in a fairly straight forward processing workflow. Do note I'm speaking of the old IE-2443 for AR-4 process which had beautifully details in red, not the garish sprayed-on reds like the newer EIR film. The Zoom-Nikkor 35-135 mm f/3.5-4.5 lens is synergetic with the Fuji and the optic usually attached to it. Over the years, several of my best photos have been taken with this combination.

 

UVP_IE_Fuji_Z1109172280.jpg

Link to comment

If you are so inclined, you actually could use your camera to create an IRG image with the sequential two-exposure method. Hot mirror for the first exposure, black-IR filter for the second, and don't change lens settings in between (use manual mode.)

 

We do these 2-exposure or 3-exposure multispectral shoots quite a lot. Make 3 frames: one UV, one IR and one Visible. Then mix and match the RGB channel assignments. Make bee vision stacks, EIR stacks (omitting the UV of course), UV/Vis/IR stacks or whatever stacks.

 

I haven't posted much of my stacked stuff using either IR or UV. Hope to get around to it someday. And I do hope to see lots more such stacks from the members!!

 

There is some wonderful software for automating such combinations written by Ben Lincoln. Here is the link from one of our Stickies in the Lists section. I'd bet he has an EIR/IRG selection in there.

http://www.beneathth...ace_Breaks.html

Added: no EIR a la Wong. But IRG with 2-exposure stack could be done.

Link to comment

I have a copy of Ben Lincoln's software, but I have not succeeded in figuring out how it works. In principle, such an engine could perform many tasks, but a user-friendly interface is sadly lacking.

 

I do my multi-exposure images with simple Photoshop operations. My forays into single-exposure IRG have been with with a converted Sony A900 and (mostly) an Alpha 525LP dichroic filter; workup currently consists of initial transformation in Adobe Pixelbender via a variation of Gilles Motut's IRGv4 plugin (https://drive.google...dit?usp=sharing) followed by standard Photoshop workup. Output is currently limited to 8-bit depth and no more than 4095 pixels on a side.

 

A British photographer is working on a program called Wavelength Pro (http://neuraloutlet..../wavelengthpro/) which shows promise. I await his first beta, as he promises relief from the above limitations.

 

I am not having any luck trying to cross-display external images on this site, so please excuse my use of links here. One of my most successful two-exposure IRG images has been extensively reblogged via Tumblr and has become my most viewed image on Flickr (which I did not expect--I am not even an IR specialist!) It is also amusing to read comments from some viewers who obviously do not know what sort of image this is--the guesses can be quite funny.

 

https://www.flickr.c...ool-2678889@N25

 

My favorite single-exposure effort so far comes from the Badlands:

 

https://www.flickr.c...nyo/15063396002

 

Not perfect: the focus is a wee bit short of infinity, and there is a problem with radial tint gradient, due to careless channel leveling and/or incorrect gamma parameters. But I like the range of reds, from the magenta of the grass to the orange-red of the junipers to the purple of the scrub brush; I think I got some things right here. The slightly off-white color of the center stripe on the road (derived from yellow-orange paint) also looks dead-on to me.

Link to comment

Here are the image from ficklr. Can everyone see them?

  • On other website, copy image location.
  • In post editor here, click the picture icon.
  • Paste the image location string into the box.

 

We are not too fond of such links however and prefer the upload and attachment method. If the other website or the other account goes away, then the linked photo disappears and the post is dead. Dead posts will be removed and all the technical information will be lost. Also linking like this increases the bandwidth of the other website, so they may not like it either.

 

3030629077_68b687b0a0_b.jpg

 

 

15063396002_e89fe46336_b.jpg

Link to comment
  • 7 months later...

Postscript: Yes, you can do two-exposure IRG with a Foveon camera:

 

"The Detritus of Delusion". Sigma SD14 with internal hot mirror out; 10-20 zoom lens at 17mm; Rocolax hot mirror used for visible exposure; Haida 850 filter used for IR exposure. Sanpete County, Utah.

 

post-66-0-55484200-1435023167.jpg

 

I even managed to get some differential red coloration in the vegetation (radiometric significance, anyone?)

The hot mirror gives a nasty color hot spot at this FOV due to dichroic angle effect. The lens also hotspots a bit in the IR. The two effects partially (but not completely) mask each other in this composite. The quasi-anaglyphic sky gives a further touch of funky surreality.

Link to comment
This is an interesting surreal image! I don't have any 3D glasses on hand, but the sky does look like it might leap out were they to be donned.
Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...