M9M spectral sensitivity

[henri]

Advertisement (gone after registration)

Only Erwin Puts gives some informations on the spectral sensitivity of the Leica M9M:

"The spectral sensitivity of the Monochrom is not yet known, but engineers told me that there is a shift into the green part of the spectrum. That should be good news, because the best definition can be found in that green range. If you still know your silver-halide facts a green filter quite often improves the recording quality of the lens/film combination."

 

Very strange that Leica did not mention the spectral "personality" of the new M9M - neutral panchromatic or other? It would be very interesting, especially when using classical color filters, to know the sensors spectral sensitivity, also concerning near infrared.

[jaapv]

Leica did not mention this because Erwin is talking about a pre-production camera. When the camera comes to the market it will have different firmware. Leica cannot comment on unfinalized output.

Although it escapes me how firmware can influence spectral sensitivity.

[luigi bertolotti]

Leica did not mention this because Erwin is talking about a pre-production camera. When the camera comes to the market it will have different firmware. Leica cannot comment on unfinalized output.

Although it escapes me how firmware can influence spectral sensitivity.

 

... indeed... no doubt that are unrelated factors... of course firmware can make some contrast corrections... but I don't know how within a BW sensor bitmap the firmware can be so smart to recognize, for instance , greens that could be "corrected"...

[sandymc]

I'd very much think that there's a filter on the M Monochrom's sensor, or at least I hope so - I don't think that a "naked" CCD would be a very good mono sensor. For those interested, I've written more on the technical aspects here: ChromaSoft: The Leica M Monochrom sensor

 

Sandy

[henri]

Thanks Sandy, great article on the mono sensor!

I think it is rather strange that the leica engineers (or persons in charge of marketing) do not publish (or: even worse do not know) the cameras colour sensitivity curve. "Just black and white" is really not enough information!

[01af]

Although it escapes me how firmware can influence spectral sensitivity.

It cannot.

 

Maybe they were talking not about firmware but about the sensor's IR filter which most likely not only cuts out infra-red but also some of the red and orange light in order to shape the sensor's spectral response curve (silicon photo diodes naturally are way more sensitive to red than blue—much unlike silver halides where it is just the other way around). That's why a semi-conductor image sensor's surface always looks green. I guess they are still working on the exact shade of green ...

[jaapv]

Advertisement (gone after registration)

I was told the Bayer filter was replaced by a clear filter.

I'd very much think that there's a filter on the M Monochrom's sensor, or at least I hope so - I don't think that a "naked" CCD would be a very good mono sensor. For those interested, I've written more on the technical aspects here: ChromaSoft: The Leica M Monochrom sensor

 

Sandy

[sandymc]

I was told the Bayer filter was replaced by a clear filter.

 

Jaap,

 

Hmmmm......well, like I said, I hope not.

 

 

Regards,

 

Sandy

[01af]

I was told the Bayer filter was replaced by a clear filter.

Don't take statements like this too literal. The filter replacing the Bayer mosaic filter most definitely is not really clear. It just has no Bayer pattern. Umm, or maybe it is clear but then another filter is not.

[sblitz]

btw .... went out yesterday and took photos with my m9, hoya green filter x1, dng+jpeg fine, b&w standard ..... wonderful photos right out of capture one, put a few through silver efex pro 2 (whatever its called) ..... no reason as far as i can tell to buy Henri .... and if i really truly wanted b&w archival photos, m4 plus the bw film of choice for that day ...

[Guest Holy Moly]

Sandy,

 

thanks for the link to your explanation about monochrome sensors.

What I'm missing here in this forum is the question how the internal light meter cell might differ in precise reading for the monochrom sensor?

 

The pics I have seen aren't precise in terms of exposures (in my eyes) of course.....

[sandymc]

Sandy,

 

thanks for the link to your explanation about monochrome sensors.

What I'm missing here in this forum is the question how the internal light meter cell might differ in precise reading for the monochrom sensor?

 

The pics I have seen aren't precise in terms of exposures (in my eyes) of course.....

 

Well, there could be a mismatch between the spectral sensitivity of the light meter cell and the sensor, which could cause some exposures, probably with lots of one or the other color, to be off. For example, if the sensor was a "naked" CCD, it would probably be really sensitive to red light. If the light meter was the same as the M9's, the sensor could over expose a scene with lots of red in it.

 

But it would be a bit surprising if the responses were that different, unless Leica really do have a "naked sensor" in there. On a prototype camera, it would be more likely that the firmware just didn't have right calibration factor for the more sensitive mono sensor programmed in.

 

Sandy

[wok64]

I'd very much think that there's a filter on the M Monochrom's sensor, or at least I hope so - I don't think that a "naked" CCD would be a very good mono sensor. For those interested, I've written more on the technical aspects here: ChromaSoft: The Leica M Monochrom sensor

 

Sandy

 

Interesting to read but is it confirmed that each CCD monochrome sensor has to have this type of characteristic? Leica works in close colaboration with KODAK (or the current owner), so maybe they managed to develop a sensor with a more film-like characteristik?

[sandymc]

Interesting to read but is it confirmed that each CCD monochrome sensor has to have this type of characteristic? Leica works in close colaboration with KODAK (or the current owner), so maybe they managed to develop a sensor with a more film-like characteristik?

 

No, whatever Leica are using might be different, although that graph is for a monochrome Kodak sensor. Just not the exact Leica one - I don't think there's a data sheet for that. But I wouldn't really expect a big difference - sensitivity at that level is intrinsic to the semiconductor process.

 

Sandy

[lars_bergquist]

[ ... ] I wouldn't really expect a big difference - sensitivity at that level is intrinsic to the semiconductor process.

 

Sandy

 

I agree. Do you remember when SLR cameras were touted as having 'silicon blue' metering cells? This was because even a plain silicon photoreceptor cell is over-sensitive to red and infra-red. So in order to get a response that approximately matched that of film, there had to be a blue filter over the cell. Hence 'silicon blue cells'. The difference was not in the silicom, but in the filtering.

 

Any photoreceptor site on a doped-silicon-based sensor converts photons into electrons in the same manner. The differences beween different-brand sensors, and even between CCD and CMOS sensors, should not be there but in the filtering and the conversion electronics, because the basic mechanism of photon capture is the same. If there is anyone out there who's in the sensor business and can correct me, please do.

 

The old man from the Age of the Silver Halide Photoreceptor

[Manicouagan1]

Thanks Sandy for the perceptive article and to Erwin for his note. One comment that I would add is that the spectral response of the Monochrom M needs to not extend beyond the spectral range for which Leica's current lenses are designed to bring in focus simultaneously and synched with the rangefinder. I expect a lot of the use of the Monochrom M is going to be in areas lit by incandescent lights with lots of infrared, low light levels and wide apertures. That combination probably defines where the sensor’s infrared cut off needs to be. On the Ultraviolet end I would suggest matching something like Kodak Plus X, which cuts off light with a wavelength shorter than 400nm. I am not sure how much the focus shift is with the ultraviolet; my older lenses had IR focusing marks but not UV.

[Jamie Roberts]

Jaap,

 

Hmmmm......well, like I said, I hope not.

 

 

Regards,

 

Sandy

 

Me too--but this is what everyone so far has said.

[sandymc]

Me too--but this is what everyone so far has said.

 

Jamie,

 

Well, really, if Leica get the color sensitivity of yet another M-series camera wrong, then we need to send the entire R&D department for eye tests or something.

 

But no, they couldn't, really, not again!!

 

Could they?

 

Regards,

 

Sandy

[Mauribix]

The "green factor" (i.e. sensitiveness to green) of senso (even a B/W sensor) is something that it's needed to preserve resolution and DR.

 

Many may already know this:

"The main approach is to use a Bayer filter in front of the sensor. This kind of filter uses a regular spaced colour matrix. It has double of green sensels because our eyes are double sensitive to green light, this is an evolution result because our ancestors used to live in the tree crowns (in this chart from Wikipedia you can see that both the sensitivity maximum and the wider range stands in the green frequencies)."

 

but if something is missed, this seems to be a good reading too:

The sensor and the Bayer matrix

 

Cheers,

Maurizio

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

Erwin Puts said green sensitivity was a good thing, as lenses perform best at green. Fwiiw.

Btw, for what reason should the camera mimic the spectral response of film? Just because our eyes are habituated to the panchromatic convention?