Advices and opinions for choosing a deeper IR filter

[Andrea B.]

Thank you for the analysis on that RM90. I will retest that filter under more stable light.

---

 

Attached are some good test photos for the B+W093, Schott RG830, Schott RG850. I had been wanting to review my IR-pass filters and how they worked, so it is nice that this topic came up. (If only I could get the weather to be more cooperative!)

 

There is very little difference between these 3 IR-Pass filters except that, as mentioned above, the B+W093 has a coating which provides some protection and flare resistance.

 

The Schott RG850 had a fractionally longer exposure time of 1/25 sec. while the B+W 093 and Schott RG830 both had exposure times of 1/30 sec. (ISO was set to 200 and aperture to f/8.) It is difficult to interpret such a small difference in exposure times as anything particularly meaningful. After all, there's only 20 nm difference between the 850 and the 830/093 pair.

 

The raw colors are also similar for the three tested filters. For whatever that is worth!!  The raw colors are somewhere past magenta (300°) headed towards cerise (330°).  I've observed that when using IR-pass filters between 700 nm to about 780 nm, A lot of raw red is recorded. As the IR length increases, some blue is recorded along with the red.

 

GEAR:  Nikon D610 + UV-Nikkor + Sunlight

SCENE:  Labsphere Standards (99%, 75%, 50% reflective) and some Euonymous leaves.

CONVERSION:  Photo Ninja. White & black point adjustment. White balance on the Labsphere standards. Details +10. Plain color setting. Resized in Photo Mechanic and labeled in PS Elements. Dust blob clone-out in PS Elements. No sharpening except for Euonymous leaves in IR photos.

 

Visible reference

 

 

B+W 093 IR-pass

 

 

Schott RG830 IR-pass

 

 

 

Schott RG850 IR-pass

 

 

 

RAW COLOR PROCESSING: Photo Ninja. Demosaic only with no white balance adjustment. White & black point adjustment. Details +10. Plain color setting. Resized in Photo Mechanic. Average raw color of standards calculated in PS Elements. Labeled in PS Elements. Dust blob clone-out in PS Elements. No sharpening.

 

B+W 093 IR-pass

 

 

 

Schott RG830 IR-pass

 

 

Schott RG850 IR-pass

[Andrea B.]

.....continued.....

 

I was not able to make any 30 sec "force tests" using the Kolari Uvir Cut filter over each of the IR-pass filters because my old D610 has again been producing a light leak. Nothing leaky showed up in the short exposure time photos just posted immediately above. But at 30 sec, I could see some streakiness indicative of a light leak.

[Andy Perrin]

I guess that camera is ready for sainthood, Andrea. It’s finally hole-y enough. 

[FL_96]

Hello, 

 

Thank you all for your detailed replies and comparisons. The filters can sometimes behave quite unpredictably with unconverted camera, maybe because their cut-off are less sharp than those with lower wavelengths, added to the hot mirror blocking. 

 

Andrea, your comparisons are very helpful and detailed ! thank you a lot, it is really nice to see they how the various strong IR filters behave in real condition with daylight luminosity.

You also answered a secondary interrogation regarding the RG780 with it's potential for false color with residual tints, I have not included it previously as the R72 can produce similarly colored results in some circumstances. 

I tough the B+W093 had no coatings, that is nice to know :) 

 

Regarding the long exposure leakage, I can run a test on my side too to see if with a hot mirror converted cameras behave the same as with unconverted. 

 

My post with comparisons has been updated to include R72 shot of the same view, with same WB and post processing settings as the other two. 

 

Now the decision is between B+W093/RG830 and RG850 or Kolari850 (not sure if the glass is from the same origin but curves are really close). 

 

Best Regards,

 

 

[Andrea B.]

FL-96, I'm happy the tests were useful to you! And it was a good review for me. 

 

Those 3 filters -- B+W 093, Schott RG830, Schott RG850 are so close in the way that they photograph, I was thinking you might enjoy going out to the RM90 or the Schott RG1000 just to have something different. But -- those two really work better on a converted camera.

 

As for false color, the RG780 does not produce very much at all. My favorite false color IR filters range from about 650 nm to 720 nm. I'm going to find a link showing those and put it here:  LINK

You can find the IR false colors there for white balanced IR.

[FL_96]

Hello, 

 

Following Andrea's last reply I finally decided to acquire a copy of the Kolari KV1000 filter. From the curve they shared by mail discussion, transmission looks very similar to RG1000. The exposure time, being slower seems to be more impacted by the influence of cloud shadows.

 

The images produced are like 830/850nm, mostly monochromatic.

Anecdotally in special circumstances (wet and stormy weather), some very slight color differential between sky and ground might appear when saturation is pushed to the maximum (slider and curve in rawtherapee). This is also blended with corner colorshifts (related to the lens)

 

Here are some examples taken with the filter :)

Gear : Canon EOS 6D Full Spectrum + Nikon 50mm F/1.8D + KV1000

 

ISO 160, 1/25s, F/4, B&W

 

ISO 160, 1/30s, F/5.6 (~ +1 ev of correction in post process), B&W

 

Example with saturation in rawtherapee

ISO 160, 1/60s, F/4, WB at 4434K :

 

Best regards,

[colinbm]

Excellent, a bit of leveling on the last one might help ?

[FL_96]

Hello, 

 

Yes the last one needs a bit of rework, it is more of an example (although not the one with the most pronounced color differential).