From: David Dyer-Bennet on
Roger N. Clark (change username to rnclark) wrote:

> So for current cameras, 8x or 3 stops is about what we can
> hope for by removing the Bayer and IR filters in current
> digital cameras, and another stop increase with higher QE
> detectors.

But removing the IR filters probably remains undesirable (for a
low-visible-light camera) because lenses don't focus IR and the visible
spectrum close enough together to really want both at once; you really
have to choose one or the other.
From: David J. Littleboy on

"David Dyer-Bennet" <dd-b(a)dd-b.net> wrote:
> Roger N. Clark (change username to rnclark) wrote:
>
>> So for current cameras, 8x or 3 stops is about what we can
>> hope for by removing the Bayer and IR filters in current
>> digital cameras, and another stop increase with higher QE
>> detectors.
>
> But removing the IR filters probably remains undesirable (for a
> low-visible-light camera) because lenses don't focus IR and the visible
> spectrum close enough together to really want both at once; you really
> have to choose one or the other.

Not necessarily: For low-light work under incandescent or candlelight or the
like, the incident spectrum will be heavily tilted to red and IR, and I'd
think that red and IR wouldn't be as bad for focus discrepancy as IR would
be in landscape work. But even in landscape work, one is often stopped way
down, so it might not be that much of a problem.

But one is going to want to borrow some IR cut filters from the Leica M8
folks at least some of the time...

--
David J. Littleboy
davidjl(a)cheapshots.com
Tokyo, Japan


From: ASAAR on
On Mon, 16 Apr 2007 20:24:30 -0700, Roger N. Clark (change username
to rnclark) wrote:

>> At first glance I agree with your first glance opinion. :)
>> Charles Schuler posted an informative link in this thread a couple
>> of days ago that may clarify the amount of attenuation that takes
>> place if the Bayer Filter Transmission vs. Quantum Efficiency chart
>> (Figure 4) is a fair representation. It shows fairly broad filters
>> with a good deal of overlap, and quantum efficiencies at different
>> frequencies ranging from about 8% to 38%, maybe averaging a little
>> bit more than 20%.
>
> It is best to use real data when making such estimates,
> rather than cartoon diagrams (although the fsu.edu has great
> explanations--just not enough info is given on the diagram).
> This Kodak data sheet gives more quantitative info:
> http://www.kodak.com/ezpres/business/ccd/global/plugins/acrobat/en/datasheet/interline/KAI-11002LongSpec.pdf

Why so dismissive ("cartoon diagrams")? The illustrations were
actually done quite well and serve the intended purpose, which was
to explain and educate, not to display precise measurements.


> See pages 5, 17 (Figure 10), 18 (Figures 12).
> The green filter bandpass is about 75% transmission with a bandpass
> (called the Full Width at Half Maximum) FWHM = 0.07 microns.
> One needs to numerically integrate the area of that bandpass
> function times the IR filter spectrum times
> the solar spectrum (assuming the sun for the light source),
> times the quantum efficiency spectrum and compare that result
> with the same integration without the filters.
> Figure 1 here shows the solar spectrum through the earth's
> atmosphere:
> http://speclab.cr.usgs.gov/PAPERS.calibration.tutorial
> To first order the bandwidth of the solar spectrum * quantum
> efficiency ~ 0.35 microns, so if we assume the IR filter
> transmits ~80%, the increase by removing the filters would
> be: 0.35 / (0.07 * .75 *.8) ~ 8.3
>
> The factor of 50 I cited previously was from a thread where
> different cameras were being discussed and included different
> f/ratios between the lenses the systems used too, and compared
> to a back-side illuminated 95% QE detector, so that 50x
> included more factors than included here.

This is probably of more interest to David. Your reply, that is,
not Kodak's PDF file. Thanks for the link. The different filters
show quite different quantum efficiencies than was shown in FSU's
chart. Kodak's paper is an example where extensive data for their
KAI-11002 sensor serves a purpose. My reply was intended only to
provide David with the chart which might help him to decide if his
estimate of the degree of filter attenuations were "ball park"
correct.


> So for current cameras, 8x or 3 stops is about what we can
> hope for by removing the Bayer and IR filters in current
> digital cameras, and another stop increase with higher QE
> detectors.

A camera with a 3 stop improvement would have a small but
significant number of photographers salivating and reaching for
their credit cards and checkbooks. I'd hope that the use or removal
of the IR filter would not be decided by the manufacturer but would
be an easily changed option controlled by the photographer.

From: David Dyer-Bennet on
David J. Littleboy wrote:
> "David Dyer-Bennet" <dd-b(a)dd-b.net> wrote:
>> Roger N. Clark (change username to rnclark) wrote:
>>
>>> So for current cameras, 8x or 3 stops is about what we can
>>> hope for by removing the Bayer and IR filters in current
>>> digital cameras, and another stop increase with higher QE
>>> detectors.
>> But removing the IR filters probably remains undesirable (for a
>> low-visible-light camera) because lenses don't focus IR and the visible
>> spectrum close enough together to really want both at once; you really
>> have to choose one or the other.
>
> Not necessarily: For low-light work under incandescent or candlelight or the
> like, the incident spectrum will be heavily tilted to red and IR, and I'd
> think that red and IR wouldn't be as bad for focus discrepancy as IR would
> be in landscape work. But even in landscape work, one is often stopped way
> down, so it might not be that much of a problem.

Hmmm, yes, true, red will predominate pretty heavily, and IR will very
likely focus closer to red than to blue with pretty much any lens.

Since my original reason for asking for this camera was for low-light
work, being stopped way down is not on my radar :-).

> But one is going to want to borrow some IR cut filters from the Leica M8
> folks at least some of the time...

I might need a red-and-infrared-pass filter for optimum results.
From: David Dyer-Bennet on
ASAAR wrote:
> On Mon, 16 Apr 2007 20:24:30 -0700, Roger N. Clark (change username
> to rnclark) wrote:

>> So for current cameras, 8x or 3 stops is about what we can
>> hope for by removing the Bayer and IR filters in current
>> digital cameras, and another stop increase with higher QE
>> detectors.
>
> A camera with a 3 stop improvement would have a small but
> significant number of photographers salivating and reaching for
> their credit cards and checkbooks. I'd hope that the use or removal
> of the IR filter would not be decided by the manufacturer but would
> be an easily changed option controlled by the photographer.

Drooooool!

Yes, having control of the IR filter would be nice. While IR isn't my
main interest here, I *do* have some interest in IR and if I happened to
get good capabilities there at the same time, that would be very nice.