M8 Sensor Sensitivity by Wavelength

[AnselWannaB]

Advertisement (gone after registration)

What are the wavlengths and their relative sensitivities for the M8's Red, Green and Blue photosites? I'm talking about the sensor itself and before it gets altered for White Balance and such. If not for each photosite, for the sensor in general.

 

I've looked around and haven't seen it anywhere.

 

Thanks,

Mark

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[scott kirkpatrick]

It's spelled out in the Kodak KAF10500 sensor data sheet, which can be found at

 

KODAK Image Sensor Solutions - KAF-10500

 

scott

[AnselWannaB]

Thanks, I'm surprised that we are having the magenta problem if you look at the graph that has the sensativity for the RGB by wavelength. It seems the red one leaks out past 700 a bit, but not much. I'm surprised that it is enough to make some blacks go magenta.

 

Mark

[robsteve]

I think it was noted somewhere else in this forum that Leica specified a different cover glass in the sensor manufacture. The Kodak specs are for their standard production sensor, not for what they produced for Leica. The cover glass was made by Kyocera, so if you did a search on this forum for "Kyocera", you may come up with the relevant post.

[ho_co]

... I'm surprised that we are having the magenta problem...

Mark--

I think that's the irony: Leica didn't expect the magenta problem either!

 

Actually, if you compare the sensor info for the M8 with that of the DMR, the M8 sensor gets twice as much IR as the DMR. As far as the sensor is concerned, all those longer wavelengths are seen simply as red light, since the photosites are filtered as you said to accept only R, B, and G.

 

So as far as the sensor is concerned, all light of longer than visible wavelengths just adds more red because it goes into the same bucket as visible red.

 

You're right, the amount of IR that gets through is very little, but it's shifted by the sensor into the red channel, and our eyes interpret it primarily as magenta.

 

Bob's conjecture in the next post is also of interest; but the IR seems to have different effects based on what it's reflected from: We see it as magenta when it's coming from blacks, as yellow when it's coming from greens--though that's less obtrusive than the funny-looking blacks.

 

I guess we'll just have to wait for the next mutation to our eye receptors--maybe then we'll be able to see as the M8 does and happily able to toss the 486 filters.

 

--HC

[gravastar]

It could also be that the blue bayer filter also lets through some IR which would not have been a problem with the standard cover glass.

 

 

Bob,

[Vivek Iyer]

Advertisement (gone after registration)

I think it was noted somewhere else in this forum that Leica specified a different cover glass in the sensor manufacture. The Kodak specs are for their standard production sensor, not for what they produced for Leica. The cover glass was made by Kyocera, so if you did a search on this forum for "Kyocera", you may come up with the relevant post.

 

 

AFAIK, there is no such claim (that the sensor or the cover glass in M8 is any different) from either Leica or Kodak.

 

The Kodak specifications given above are the ones for the M8 sensor.

[ho_co]

I think it was noted somewhere else in this forum that Leica specified a different cover glass in the sensor manufacture. The Kodak specs are for their standard production sensor, not for what they produced for Leica. The cover glass was made by Kyocera, so if you did a search on this forum for "Kyocera", you may come up with the relevant post.

Robert S, Vivek is correct, and slipped in while I was still writing this:

 

Leica worked with Kodak in designing this sensor and its cover glass: the KAF-10500 is the final Kodak/Leica product built specifically for the M8. The Kodak specs above are for the M8 sensor.

 

The problem is that it allows 5%-10% transmission of IR, where Kodak's more usual sensor design with thicker cover glass in the DMR allows only 2%-4% transmission of IR. (Info also from the Kodak sensors pages, though at the moment I've forgotten the model number for comparison of the DMR sensor.)

 

Respectfully

--HC

[AnselWannaB]

On pg14 of the linked .pdf above, it just doesn't look like there is that much IR sensitivity. I guess what would be more interesting is the transmission curves for different cover glasses.

 

So you can just buy these sensors from Kodak? I know it is very, very far from an actual camera, but I would have thought there was some exclusivity for it. I wonder what the list price is .

 

I've worked a bit with IR absorption/reflectivity and its effect on solar heat gain and imaging technology. It gets really trippy since you want to talk about "color" but it is beyond what we see, so while the physics are the same you can get tripped up talking about what "color" something is in the IR.

 

An objects color is a combination of its reflectance/absorption/scattering properites, the illumination source, and the observer.

 

Mark

[scott kirkpatrick]

The problem is that fundamentally, silicon just loves IR, and it is excessively sensitive to red, so red has to be decreased and IR eliminated to come up with RGB sensitivities that correspond to what the human eye sees. (Note that this reverses 150 years' experience with film, which loves UV and doesn't care about IR, and takes great effort to even respond to red.)

 

To see the bare sensor response, look at some of the other Kodak chips currently described at variations of that URL. The KAF10100 is the DMR chip. The KAF39xxx is the one used in the biggest Phase One back, and is described without any IR filtration, since the camera manufacturer adds that in their backs. Also, the DMR uses a dichroic IR-cut filter over the CCD chip, and that probably is Leica-installed, not covered in the Kodak specs. It has a sharp cutoff at about 600 nm, I believe, but I haven't seen its specs.

 

The point made by several above about how IR isn't just red, but changes other colors that it adds to (like yellowing the greens in foliage) is very true. Take a look at Stephen Johnson's "Digital Photography." He does landscapes and got started with a digital scanning back about 15 years ago. The scanner has no IR filter, and Johnson describes dealing with IR, using IR to achieve desired effects (he likes high key pictures), and all that.

 

scott

[leicamr]

I have read several threads in the forum and note that some of you M8 users are using B+W IR filters. Do you consider these to be as good as the Leica IR filters? I am not sure if Leica currently supply IR filters.

 

I have just ordered my M8 so I am keen to know.

 

Thanks

 

Mark