POTN Office Pool - First "New" Monochrome DSLR

[FlashZebra]

A virtual office pool bet.

Which camera manufacturer will deliver the first consumer monochrome only DSLR. And yes, I know Kodak has already produced these (several years ago on discontinued cameras), but I am talking about a new offering available to any consumer for less than $3000.00 US.

A monochrome only camera would certainly have a grand upside for monochrome images, over the grayscales from interpolated Beyer RGB array sensors.

As the DSLR market matures, it is bound to happen eventually, but who will deliver it?

Canon had the astronomical version of the 20D, so there must be promise for niche cameras.

When might also be an interesting question? But lets stay with who for this thread.

Enjoy! Lon

[Tom W]

I vote "nobody".

[delinian]

Canon of course!

[solinger]

I vote "nobody".

I have to agree with Tom.

[Ronald S. Jr.]

I don't think either of the two "big dogs" will come out with this first. Maybe Sony will release another similar to the "Alpha", or maybe samsung or pentax will take the "crown". I just don't see N**** or Canon doing it anytime soon.

[FlashZebra]

I vote "nobody".
I can assure you that somebody will eventually. The extra fidelity of monochrome only will be a sufficent draw for somebody.

Enjoy! Lon

[Jon]

Why would they bother? The mass market just won't be there. They're giving us monochrome selectable as a picture style and people are still advocating doing it in PhotoShop. As it is, you can post-process to get the effects of various colour filters; if they make a strictly-monochrome camera, you'll need to go back to carrying Wratten filters in all your favourite colours.

[FlashZebra]

Why would they bother? The mass market just won't be there. They're giving us monochrome selectable as a picture style and people are still advocating doing it in PhotoShop. As it is, you can post-process to get the effects of various colour filters; if they make a strictly-monochrome camera, you'll need to go back to carrying Wratten filters in all your favourite colours.
Well, they would bother to garner the extra fidelity of inherent grayscale bit values at every pixel sensor location as opposed to interpolated Beyer RGB bit values converted to grayscale.

Humans are current carting huge cameras about in search of additional fidelity. Five additional filters would hardly defray this crew.

As for market share for such a device you may be right, but if Canon sees an upside for an astronomical version of the 20D, surely a grayscale only camera would garner as much (most likely more) market.

Possibly we need to stock up now on those Wratten filters, everyone is dumping them.

Enjoy! Lon

[Jon]

The 20Da didn't need a new sensor - just a different high-pass filter over it. So they were able to use the exact same manufacturing process for their sensors. A monochrome sensor would need a different process, as the cells wouldn't need separate colour filters. What I'd like to see is a 4 or 5 layer Foveon (or some equivalent arrangement; multiple sensors are fine, if that's what they want to do) with UV and IR cells as well that I could post-process for the desired multi-spectral effects. Can you imagine looking at a butterfly on a flower in IR and UV only?

[FlashZebra]

The 20Da didn't need a new sensor - just a different high-pass filter over it. So they were able to use the exact same manufacturing process for their sensors. A monochrome sensor would need a different process, as the cells wouldn't need separate colour filters. What I'd like to see is a 4 or 5 layer Foveon (or some equivalent arrangement; multiple sensors are fine, if that's what they want to do) with UV and IR cells as well that I could post-process for the desired multi-spectral effects. Can you imagine looking at a butterfly on a flower in IR and UV only?
Points well taken, but the market for a monochrome only camera has to be much larger than an astronomical only camera (20D derivate). Meaning some additional development capital could be accommodated.

Leaving the RGB filters off the CMOS or CCD, would seem to require some additional work, but most likely not a huge project.

In any case, I expect it to happen at some point; just not sure when; or by whom.

I am reasonably young and not senile yet (would one know that they are senile), I will dredge up this thread when it happens, if I am not banned first.

Enjoy! Lon

[steved110]

I think there would be a small niche market for such a camera - but not at $3000 - if it was priced near the rebel XT/ 350D it might have a chance. I have read a few reader's letters in UK mags asking this very question. Considering how easy it is to use photoshop to create black and white from a colour RAW or Jpeg it would seema bit limiting to me - bit like making a film camera that only does 35mm B&W....
I suspect that if anyone does it, at first it will be a 'specialist' compact like the ricoh type - poss by sony, samsung or olympus - rather than a proper DSLR - after all the SLR system is supoosed to be versatile, so why clip its wings?

[Tom W]

Points well taken, but the market for a monochrome only camera has to be much larger than an astronomical only camera (20D derivate).

I would think the opposite, though my distaste for monochrome might have some influence here. Although I'd buy neither, I would prefer the astrophotography setup to a monochrome-only setup. But that's me - I suppose that I can't speak for all.

[brucea]

Leica or Ricoh. Strictly niche and expensive. Rangefinder.

[Mayfly]

I don't see it happening only because anyone who would be interested in that would also be the type of person to be savy with photoshop which can do an even better job of producing a great monochrome.

[Woolburr]

Too small a niche market for any major player to go after. Leica or Ricoh might waste some time and energy on it. I haven't heard a lot of US photogs screaming for a monochrome camera. Guess it is a regional buzz. I find myself asking "why?". Are mug shot photos in high demand?

[dpastern]

Nah, won't happen. Total redesign of the pass filter would be needed, and most probably the sensor. Just costs too much money in R&D for too little return. Most photographers are happy to shoot in colour and convert to greyscale in Photoshop. Canon has better things to spend R&D money on I think.

Dave

[DocFrankenstein]

I'd venture to guess that a foveon type sensor will come around first.

[FlashZebra]

I don't see it happening only because anyone who would be interested in that would also be the type of person to be savy with photoshop which can do an even better job of producing a great monochrome.
Photoshop work will not harvest the upside of increased fidelity from a direct grayscale value from every pixel in the sensor array, as opposed to interpolated results from a Beyer RGB grid array, converted to grayscale.

The fact that a Beyer grid array devotes 1/2 of the pixels to green, 1/4 to red, and 1/4 to blue, but the resulting file interpolates an RGB value for every pixel seems to be getting lost here. Two of the three color values at every sensor grid position is interpolated. This resulting file, with the inherent downside of interpolation, is then converted to grayscale.

With a monochrome sensor, every pixel position produces a grayscale value that is not built on and underlying RGB interpolation.

This is not just a grayscale tone issue addressable in Photoshop, but a fundamental fidelity issue harvested at the sensor pixel level and only available with a monochrome sensor.

Enjoy! Lon

[FlashZebra]

Nah, won't happen. Total redesign of the pass filter would be needed, and most probably the sensor. Just costs too much money in R&D for too little return. Most photographers are happy to shoot in colour and convert to greyscale in Photoshop. Canon has better things to spend R&D money on I think.

Dave
Unlikely the high pass/ anti-alias filter would be much of a problem as it would still need to pass the same quality of light to a sensor of existing resolution; the need for removing the individual RGB filters at each pixel array position is the big issue.

But, if they can put a complicated RGB filter grid down, they should be able to put down a filter grid where every pixel is the same.

Production numbers would be the issue. But, if you were the only camera maker with this sort of product, there might just be enough interest.

Enjoy! Lon

[FlashZebra]

I think there would be a small niche after all the SLR system is supoosed to be versatile, so why clip its wings?
It certainly would be a niche product, but if niche is never a possibility, then explain the astronomical version of the 20D?

I also personally know 4 different photographers in my camera club that have infrared converted cameras (that is a niche). This is more than own a Nikon D2X or Canon 1Ds combined in the club of devoted enthusiasts (of 65 members).

Enjoy! Lon

[Longwatcher]

Coming up with a digital panchromatic camera (AKA monochrome) is easy technology, just remove the bayer filter from the sensor and then add a pass/block filter to keep the sensor within the range of B+W film.

Oh wait there are multiple kinds of B+W film, so we need a removable filter in front of the sensor to simulate the different types of film or we have to permantly live with the look of the sensor as is.

Secondly some software would have to be developed to handle the files, but it should be easier to develop then color bayer interpolation software is. The best part is the raw files would probably be smaller on a pixel for pixel basis.

But most importantly we need a sensor with enough dynamic range to be at least the equivilant of B+W film instead of color slide film. That we have a technology problem with, not unsolvable, but is it worth the cost.

Given that we can interpolate the bayer pattern sensor into a fairly close (for humans) approximation of B+W that reduces the demand significantly.
Given that producing a sensor with a greater well-depth (for greater dynamic range) will cost more, that reduces the demand.
Given that the reduced demand will also increase cost further reducing demand.
Given that if heading towards a foveon sensor would produce even closer interpolated B+W results it might be practical to wait for that development.
and finally given that the design teams are working on things that are higher up on the way to make money and we haven't quite hit the physics limits yet.
I would assess that we are unlikely to see a panchromatic only digital camera within 3-5 years, but the odds improve greatly after that point if demand remains high. I would actually guess that a MF format camera will have the capability first, followed about 1-2 years later by a 35mm or APS-C format DSLR.

Just my opinion,

[Balliolman]

I do not see a monochrome dslr having mass market appeal. I guess many people would choose colour over black and white anyway. The likelihood is a digital stereo camera will be mass marketed beforehand ...

[FlashZebra]

The best part is the raw files would probably be smaller on a pixel for pixel basis.
Well, if compression is not considered, and on a "pixel by pixel" consideration, and the bit depth of the B&W sensor would be the same as existing technology, the resulting RAW files would be the exact same size as a Beyer RGB sensor.

Of couse after opening the RAW file from the Beyer sensor it would be three times larger as an opened RAW Grayscale file (for the same bit depth due the required RGB interpolation).

Enjoy! Lon

[Longwatcher]

Well, if compression is not considered, and on a "pixel by pixel" consideration, and the bit depth of the B&W sensor would be the same as existing technology, the resulting RAW files would be the exact same size as a Beyer RGB sensor.

Of couse after opening the RAW file from the Beyer sensor it would be three times larger as an opened RAW Grayscale file (for the same bit depth due the required RGB interpolation).

Enjoy! Lon

Badly worded on my part, as you are correct. I was thinking of file compression but forgot to mention it.

[JMAS]

If IR with regular speeds has some market, then B&W could have too.

For what is worth, the B&W always needed much post processing, even in the film days, and I really don't see how the direct B&W could be better than converting using channels mixer and then using all the dodging and burning one needs.

IR is not obtainable (at least it is not pratical) without a transformed camera and the big companies never expressed interess on it. Why would they with B&W?

[SWPhotoImaging]

I'd certainly like a monochrome sensor DSLR, if it had a wide dynamic range for luminance. It is one area where digital still lags way behind film, IMO.

It would definately be more useful than an IR camera.

[René Damkot]

If it's ever released, my guess is Epson or Leica (RF)

[cjm]

My vote is ADOBE PHOTOSHOP. Oh wait they have that featuer in Channel Mixer so what am I talking about... oh well.

My vote really is Canon D60 since that seems to be the first modified SLR camera.