So I'm looking at the histogram Michael shows, from an IR converted camera. And I'm thinking, if I took a shot with my R1, then used levels or something similar depending on your software, to modify (reduce) the green and blue channels so the histogram looked a lot like that sample, and then converted to B&W, would I be in the same place as Michael is? Or should I think about this more than 10 sec. and figure out what I'm overlookin? Or just post this and let someone quicker than me explain why that wouldn't be similar.

What you're missing is that almost no visible light is being used to create an IR image with the modified camera.

The light being filtered though the Red part of the bayer matrix is made of of Infrared, but also that being captured by the Blue and Green segments of the Bayer.

It's all Infrared. The filter is cutting off almost all visible light, so juggling the R G and B channels goesn't get you anything other than a tonal shift.

Michael

give it 20 sec.

With your idea you would still be working with light from the visible spectrum in your image. As the article mentioned, IR light is outside the visible spectrum and normal cameras have filters to block that out.

The IR modified camera blocks out most of the visible light and allows the IR light in.

IR light is light just beyond the red spectrum (thus the adverb "Infra") and so naturally most of the data is recorded in the red channel.


...and Michael types faster than I do.

There is an interesting psychological twist to IR photography. It is akin to giving a blind person electonic stimulation to the cortex of the brain from a "bionic eye". IR photography converts invisible (to the human eye) rays to visible. It raises the old debate of weather we each see the same colour or object. Perhaps photography is like politics: perception is reality.

I have seen some very accurate appearing "infrared" images made with the channel mixer. Most of the pre-recorded actions available online include boosting the red and lowering both green and blue, but some also feature a bit of gaussian blur on the green channel. I am wondering if anyone has experimented with idential images, one taken from an IR camera and one from a standard, and tried to match the standard to the IR using the channel mixer. I'd be very interested to see how close they can get.

They might look accurate, and for some purposes they might do, but they are not accurate. For example, some insects see in other wavelengths than us and you might want a bugs eye view of the world. Your solution would not show the patterns on that flowers that to us look dull. (Okay I can't recall if they see IR or UV. I fear it is UV!!!)

Leif

Insects have been known to be UV-sensitive, yes.

That's why you get these nice blue and UV-intense lamps to attract insects and "zap" them.

Is it really necessary to modify the camera? Maybe you could just use longer exposure times with an IR-filter on the lens?

The camera still needs to have its IR blocking filter removed.

Then you can add a visible light blocking filter (Infrared pass).

But, if you put the filter in front of the lens your exposures, even in bright sunlight, will become VERY long.

This is all discussed in the article.

Michael

Yes. But this guy (http://www.naturfotograf.com/UV_IR_rev07.html#top_page) seems to take pictures with unmodified cameras.

As he says on that page: "Recently, I acquired an old D1 and managed to get it modified by stripping out the entire AA-filter pack in front of the CCD. You should not attempt to do this modification yourself, because the camera needs to be disassembled entirely before the filters can be reached. However, try talking your local Nikon repair facility into doing this for you. After the filter pack is removed, you get a tremendous boost of IR sensitivity, at least 7 stops compared to the non-modified D1 when a Wratten 89B filter is attached to the lens in use."

He got a boost by modifying the camera, but it was still possible before. I'm going to pick up my first camera (Nikon D50) on tuesday, so I'd like to keep it bog standard, at least for a while. :^) But I would like to try IR too. I read somewhere else that you can make an IR-filter from an old five and a quarter inch floppy disk, so maybe I don't even have to buy an expensive filter to give it a try.


Thanks for an amazing site by the way, if you are the same Michael who write the articles. You kept me up all night reading when I should have been sleeping.

Lucidor:
You can take IR images with an unmodified camera by using a screw-on IR-pass filter (to filter out the visible light), and using a tripod becasuse the resulting exposures are very long (because most but not all of the IR is filtered out by the anti-IR filter in the camera, so only a very little gets through). That's how Bjorn Rorslett at naturfotograf.com does it (and how I currently do it). I haven't heard anything about being able to use floppy disks, and I've read a sizeable number of articles about IR photography on the web, so I would be extremely surprised if anything like that worked, especially with reasonable image quality (think about it - if you need to care about the quality of your lenses and filters because poor-quality ones give you crappy images, imagine how bad the image quality would be if you passed the light through something like a floppy disk that wasn't designed for photography to begin with...).

Lisa

Floppy disks have holes in the center. Just thought I'd throw that one out there...

That it would be poor quality images goes without saying. :^) But just imagine how cool and fun it would be if it did work. Clearly worth an experiment to see if pictures like these http://gallery.mercurycube.com/irphotos/ are fake or not.

Thanks for confirming my suspicion that you can use unmodified cameras. I've read a lot about the subject, but no-one said it straight out.

61Dynamic: You're not supposed to use the part of the disk with a hole in it. :^) You can see a picture of a floppy filter in the link above.

Hmm. Well, I'll concede it might be possible...
Could be fun to play with despite the poor image quality, and, if you find you're really enjoying it, *then* spring $100-$200 for a good IR-pass filter.

Lisa

I have a question regarding the problem of flare in an IR modified camera (modified to shoot only IR). How bad is the flare problem as opposed to the same camera shooting regular color (non IR) images? Is there any filtration on the infrared filter that eliminates most of the flare? Do you get that bright spot in the middle of the frame that I got using an unmodified 10D with just an IR filter over the lens? Thanks, Eleanor Brown

One drawback to the F717 is that the aperture is limited to wide open when in nightshot mode. I glanced at the photos shown in the above mentioned link and they appear to have limited depth of field. Also, I notice from two left-hand photos in the third row down (same picture, one IR one visible) that the exposure difference is 8 stops - 1/500 down to 1/15 and f/5.6 to f/2.