Do Epson, Nikon and Sony have also a magenta problem?

[Vivek Iyer]

Advertisement (gone after registration)

Rex,

 

You refer to this thread and I quote what I posted there (response from Kodak):

Film Lives! ● View topic - Leica M8. UV friendly?

 

Here is the response I received from Kodak:

 

From: R F McKeever

 

Dear Sir,

 

As far as the nattive capability of the imager, it is spelled out in the specification. There is an IR Absorbtive cover glass on the part, and there is virtually no transmission above 780 nm. The color dyes open back up above that level. These are color parts, and you do not want extraneous transmission above that level corrupting your color data.

 

As far as the performance of the specific imlementation, you would need to ask Leica.

 

Sincerely yours,

Bob McKeever

Business Operations

Kodak Image Sensor Solutions

 

However, if you look at Kodak's own data..

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[rvaubel]

Rex,

 

You refer to this thread and I quote what I posted there (response from Kodak):

Film Lives! ● View topic - Leica M8. UV friendly?

 

 

 

However, if you look at Kodak's own data..

 

Vivek

 

There is a huge disconnect between the chart your are referencing and the statement that Mr McKeever is making considering the atenuation of the IR.

 

The chart makes it seem that the IR is essentially attenuated at 700 nm. That would be a good thing. I don't think we would be having the discussion and controversay of the last two weeks if the Kodak filter actually cut the IR frequencies above 700nm

 

But Bob McKeever said in your quote that attenuation of the IR was not fully effective until 780 nm. That is pretty far into the Infrared, isn't it?

 

Am I missing something here?

 

Rex

[Vivek Iyer]

Rex,

 

I am also at a loss to interpret the Kodak data. If I go with only the data ( I suspect Leica did that), then everything is peachy with regards to IR suppression. As is clear now, that isn't the case in reality at all.

 

As I mentioned above, Nikon and Epson also do not say anything about IR response.

Luckily, there is one informed reviewer of Nikon cameras (http://www.naturfotograf.com).

 

Bjoern Roerslett is equipped to check for UV and IR responses of a sensor.

 

Only a few months ago, I had a chance to check out an Epson R-D1 and found out that it has excellent UV response as well.

 

I really hope that Leica folks will do a better job of checking the camera behavior (not that difficult to do!) before sending it out of their factories.

[rvaubel]

Rex,

 

I am also at a loss to interpret the Kodak data. If I go with only the data ( I suspect Leica did that), then everything is peachy with regards to IR suppression. As is clear now, that isn't the case in reality at all.

 

As I mentioned above, Nikon and Epson also do not say anything about IR response.

Luckily, there is one informed reviewer of Nikon cameras (http://www.naturfotograf.com).

 

Bjoern Roerslett is equipped to check for UV and IR responses of a sensor.

 

Only a few months ago, I had a chance to check out an Epson R-D1 and found out that it has excellent UV response as well.

 

I really hope that Leica folks will do a better job of checking the camera behavior (not that difficult to do!) before sending it out of their factories.

 

Vivek

 

Look, I don't want to be pessimistic about Leica's ability to check something so basic as the frequency response of the sensor/absorbtive filter combination nor do I doubt Kodaks ability in this regard also.. But lets face it, somethings screwed up. I think it's "back to basics" with regards to this purple/black issue.

 

Rex

[scott kirkpatrick]

Kodak's data shows the increased IR sensitivity of the M8's chip (KAF10500) compared with the chip used in the DMR (KAF10100). For the M8 chip, transmission from 710 to 750 nm is spec'ed at < 10% and 750 to 1000 nm at <5% while the filter used in the previous chip had transmission from 700 to 750 nm <5% and on average, 700-1000 nm less than 0.5%. While the near IR is suppressed in both chips, it sounds like the M8 sees much more than the DMR saw.

 

scott

[rvaubel]

Kodak's data shows the increased IR sensitivity of the M8's chip (KAF10500) compared with the chip used in the DMR (KAF10100). For the M8 chip, transmission from 710 to 750 nm is spec'ed at < 10% and 750 to 1000 nm at <5% while the filter used in the previous chip had transmission from 700 to 750 nm <5% and on average, 700-1000 nm less than 0.5%. While the near IR is suppressed in both chips, it sounds like the M8 sees much more than the DMR saw.

 

scott

 

It sure does see a lot more IR. The explanation that I was assuming was the reason was that the IR blocking filter was a dichromic, reflective one and it had to be less than optimal because of angles of incident, blah, blah, blah. I don't have an M8 (on the waiting list) so I could never peer in and see.. But if it is absorptive, why not make it thicker than .5mm ? I don't know if I buy Leica's explanation about "astigmatism" being introduced at the edge of the field. What the hell does "astigmatism" have to do with it. OK, maybe lateral color, but I get the feeling that the people that write Leica's press releases don't know lateral color from astigmatism.

 

anyway, if a thicker filter causes problems, what about a combination of an absorbtive .5mm filter and a moderate reflective interference blocking filter?

 

My take on all of this is that it is solveable at the engineering level. Costly, but solveable.

 

Rex

[Vivek Iyer]

Advertisement (gone after registration)

I

My take on all of this is that it is solveable at the engineering level. Costly, but solveable.

 

Rex

 

No difference of opinion there, Rex.

[scott kirkpatrick]

It sure does see a lot more IR. ...snip ...

 

I don't know if I buy Leica's explanation about "astigmatism" being introduced at the edge of the field. What the hell does "astigmatism" have to do with it. OK, maybe lateral color, but I get the feeling that the people that write Leica's press releases don't know lateral color from astigmatism.

 

anyway, if a thicker filter causes problems, what about a combination of an absorbtive .5mm filter and a moderate reflective interference blocking filter?

 

I'd really appreciate a short white paper from the engineering people, not the marketing echelon, on what choices guided the change in IR filtering from the 10100 to the 10500. The only statement that was circulated (in an awkward translation sent to Sean Reid) jumps to the assumption that an absorptive filter was necessary, and that its front surface needed to be anti-reflective, thus could not be used for an IR-cut layer (without actually saying all that). And then loses credibility because of the apparently garbled discussion of the side effects that Leica believes to have eliminated.

 

The RD-1 also chose to use absorptive IR filtration (see the picture a few posts above), so there must be strong reasons for this, and late Nikon products apparently use a mixture of both techniques, as you suggest. Perhaps Leica restricted their choices to materials readily available without incurring extra engineering expense. Since the DMR is demonstrated to have the capability of removing the red-vignetting of a coded lens resulting from using a dichroic IR cut layer at the cover glass, that can't have been the dominant issue.

 

I don't think we will ever get a public statement of how the unhandled IR sensitivity was ignored until so late, but that's an internal matter.

 

scott

[rvaubel]

Scot. vivek

 

The sad thing is since a sensor/filter solution was possible, Leica has squandered maybe the only opportunity to develop a large sensor-short backfocus camera. I'll live with the filter on lens/color profiling kludges because I love the idea of large sensor, compact minimalistic camera, but elegant it ain't.

 

Oh well, maybe next time

 

Rex

[scott kirkpatrick]

Scot. vivek

 

The sad thing is since a sensor/filter solution was possible, Leica has squandered maybe the only opportunity to develop a large sensor-short backfocus camera. I'll live with the filter on lens/color profiling kludges because I love the idea of large sensor, compact minimalistic camera, but elegant it ain't.

 

Oh well, maybe next time

 

Rex

 

[wild speculation]

I think they constrained themselves by only using materials that existed at the time the KAF10500 specifications were set, for obvious reasons of cost and schedule. Wouldn't you want to bring the M8 to market and start a new revenue stream before the alligator-skin titanium M7 buyer stream ran dry? So they saw it as a choice between the existing DMR filter, with a front surface IR-cut filter and a back surface AR coating, and an absorptive filter, with two sides AR coated, backed up if necessary by an IR-cut filter in front of the lens, where its front surface cannot cause any ghosting or flare. They didn't explore finding a new sort of hybrid filter, or AR + IR coating, because these take time to develop. Even though Canon and Nikon appear to have such solutions in their proprietary imaging chips. It was scary enough choosing a thinned-down IR absorptive filter.

[/speculation]

 

I really would appreciate Leica sharing an engineering analysis of the choices they made. The marketing folks can put whatever Panglossian spin they want on it, but it would be great to know what they were thinking and what are the greater problems which have been avoided.

 

scott