IR Pass Filter Wavelength Question

[xpatUSA]

Hello IR practitioners,

The Sigma SD10 has a easily-removed IR blocking filter (one screw and it's out). So, some IR filters are on their way. I bought a selection:

ICE 680nm
Marumi 700nm
Zykkor 720nm (tried to get a Hoya but the transaction went belly up)

I have no idea how sharp the cut-off slope is on any of these filters

Has anybody tried any of these and, if so, should I expect any significant differences between them, image-wise?

General comments are also welcome.

I am not expecting great success since the SD10 is not the finest camera for long exposures. On the other hand, the Texas sun can get quite strong at times i.e. all summer!

cheers,

[t6b9p]

Ted
one of the problems with "off" brands is not knowing the source of the glass or the quality control. If they specify they use Schott or Hoya glass then you can obtain the transmission plots from the manuf. site, but remember thickness will affect the plot. If they use "other" glass, I have seen several reported issues of incorrect labeling and poor quality control. I typically test the transmission properties of my filters to determine their "real" transmission properties as suppliers often just post the manuf. plots which may be of different thickness to the product they are selling. Some are shown here http://www.beyondvisible.com/BV3-filter.html

You picked an interesting selection of wavelengths as most people pick a larger spread. Yours are interesting because they cover the narrow transition range from R/IR to IR/weak R (an 830nm filter just transmits IR and no R). This range is best for colour IR and the popular "channel swaps" but can also be converted to B&W, however the B&W versions will not exhibit as high an IR contrast as using an 830nm.

In a Bayer sensor, the 680 can be more of a challenge with respect to the extra R transmission leading to blown R channel, whereas the 720 results in more evenly distributed signal in the RGB channels. In the brief time I owned a Sigma SD10 I didn't get the opportunity to see the effect of different IR filters and how the (non-Bayer) Foveon sensor responded. I would be interested to see/hear your conclusions on this.

I also seem to remember that the Sigma histogram vs the dcraw histogram indicates that there is some behind the scene "trickery" occurring. A pure IR image resulted in the Sigma histogram with data in the R and B channel only, while the G channel was empty. Now given the fact that the channels are stacked B on G on R and that IR penetrates deeper than R, and R more than G, and G more than B, how is this possible. It would be anticipated that in the worst case there should be a little blue signal, a little more green and the dominant channel would be red, and in fact that is exactly what occurs when the histogram is viewed using dcraw or X3F.

[xpatUSA]

Thanks Shane,

Yes, the three I chose were for just giving it a try with some chance of the scene showing up on the LCD. Of the three, I have seen examples for the Marumi 700nm that looked quite good. Might go higher later, if I get into the genre.

As to histograms, Dave Coffin is a law unto hisself, as you probably know. For example, his base illuminant is D65 even for ProPhoto! As to "trickery" the conversion matrix for Foveon sensor levels to XYZ is pretty fierce, e.g:



cheers,

[t6b9p]

I started with 695 and 830nm and fell in love with the great B&W contrast from the 830nm so I had a D200 converted. Now I am considering a D610IR830, so yes I am "hooked". Please keep me updated on your results/conclusions.

[xpatUSA]

I started with 695 and 830nm and fell in love with the great B&W contrast from the 830nm so I had a D200 converted. Now I am considering a D610IR830, so yes I am "hooked". Please keep me updated on your results/conclusions.

Hi Shane,

I made a start, see my thread here IR and RawDigger and also here http://www.dpreview.com/forums/thread/3611363

The SD10 does very well as far as not needing obscene exposure times and you may be interested in the use of RawDigger as the converter from X3F to TIFF (for IR shots).

I've ordered a hot-shoe mounting OVF (for obvious reasons) and took delivery of a sturdy Giottos tripod and ball-head. Should be fun . . .

An early conclusion is that I see quite a lot of visible content, even at 720nm, so definitely considering going up in wavelength once the novelty of what I have wears off

cheers,

[t6b9p]

thanks for the update - I'll check out the other posts.