A modest proposal: a 35mm CMOS sensor in a DB.

[ondebanks]

Obviously, a CMOS medium format back would be a game changer. But we keep hearing that the tech/costs of developing a MF sized CMOS sensor are too great.

So here's an idea, which I am just going to throw out there for discussion: how about a digital back with an existing CMOS 35mm full frame sensor? Say the 24MP Sony sensor (D3X/A900). Sure it's smaller than medium format, but think of what it opens up, on say a Hasselblad H4D or a Hy6 or an RZ67 - a really high performance sensor on a body with leaf shutter lenses and interchangeable finders.

Remember, we used be able to do this with our Mamiya 645s and Bronica ETRS's: 35mm backs for those "special emulsions" not available in 120/220.
And that was when 35mm format was 2.7x smaller in area than 645 film; now 35mm full frame is only 1.7x smaller than a P30+, H4D-40, or Aptus 8.

So think of a CMOS 35mm sensor as a "special emulsion" for high ISOs, long exposures, and faster frame-rates.

I can see less merit in the idea for the Mamiya/Phase 645 users - fixed prism, focal plane shutter - we can already use our lenses on a FF DSLR anyway.
But maybe it would suit someone who is heavily invested in the new LS lenses and the DF body.

Ray

[torger]

I don't think it would sell much.

I don't think that CMOS is very critical for MF either. What does it provide? High ISO, good live view and video, and a little bit better DR. For the typical MF studio photographer (which I guess make up the largest part of the MF market) none of this is necessary.

As a tech camera user I surely would love live view, but I don't think tech camera users is a very large part of the total MF market (am I right?).

Lower prices on existing backs is my number one request...

[ctz]

for product / still life, on a view camera:
this might be GOLD!

[MrSmith]

 What does it provide? High ISO, good live view and video, and a little bit better DR.

that's the holy grail right there. i would buy one tomorrow.
actually it would have to be 2 sony sensors joined in a 36x48 72mp configuration.

[gerald.d]

for product / still life, on a view camera:
this might be GOLD!

Adapters to put DSLR bodies on view cameras already exist though.

[ctz]

Adapters to put DSLR bodies on view cameras already exist though.

yes, and mainly because of the mirror box (of DSLR bodies), they're pretty much useless. (i.e. when applying bigger shift, as an actual product shooter do more often than not).
also, focusing a (let's say) 55mm on a MFDB is tricky, but with a viewcamera/DSLR adapter is virtually impossible.

[uaiomex]

I think everything will come up to price. Under 4k, I think. imo
Eduardo

[BernardLanguillier]

My guess is that the toppings of current DSLR sensors (micro-lenses, IR filters,...) are finely tuned for the expected position of the lenses and resulting incoming light angles.

I would expect complex color issues to show up if they were used in a different context without adjustments.

Cheers,
Bernard

[torger]

I guess color cast would be an issue too. If we look at the current MFDBs it seems that smaller pixels leads to more problems concerning wide angle color cast. It also depends on how the pixel is designed, if I would guess I think CMOS pixels would have more troubles with color cast than CCD, due to the smaller area used for light pickup, but I'm not sure.

It might be that current tech camera wide angle lens designs would be impossible with a CMOS sensor.

I would prefer if medium format would try to have moderately higher resolution than 135 format and larger pixels at the same time. Going for small pixels seems to kill the possibility of non-retrofocus / weak retrofocus lens designs which I think may be the "secret" of the good tech camera wide angle performance.

Unfortunately I don't think MFDB makers care that much about supporting tech cameras, the money is in SLR type of cameras. We'll see what happens.

[Sareesh Sudhakaran]

what's stopping someone from clubbing together an array of 35mm or APSC sensors in a 4x5 or 8x10 grid to take advantage of large format lenses and stitching?

[torger]

what's stopping someone from clubbing together an array of 35mm or APSC sensors in a 4x5 or 8x10 grid to take advantage of large format lenses and stitching?

I don't think there is any current technology to put sensors so close together that there are no gaps between them. If there was it would be a much cheaper way to make large sensors. Arrays of sensors are used in for example astronomy, but then there are gaps between them.

[Graham Mitchell]

So here's an idea, which I am just going to throw out there for discussion: how about a digital back with an existing CMOS 35mm full frame sensor? Say the 24MP Sony sensor (D3X/A900). Sure it's smaller than medium format, but think of what it opens up, on say a Hasselblad H4D or a Hy6 or an RZ67 - a really high performance sensor on a body with leaf shutter lenses and interchangeable finders.

The sensor size is one of the main benefits of medium format, imo, so your suggestion makes little sense. The lenses are big, expensive and heavy because they create large image circles which would be wasted here. The maximum apertures are not so fast because they don't need to be on a large sensor, but they wouldn't be so well matched to a 35mm sensor. Compared to a 35mm DSLR you would be giving away a lot of usability and features, just for an extra stop of sync speed. I don't see this working at all, sorry.

[Pingang]

Obviously, a CMOS medium format back would be a game changer. But we keep hearing that the tech/costs of developing a MF sized CMOS sensor are too great.

So here's an idea, which I am just going to throw out there for discussion: how about a digital back with an existing CMOS 35mm full frame sensor? Say the 24MP Sony sensor (D3X/A900). Sure it's smaller than medium format, but think of what it opens up, on say a Hasselblad H4D or a Hy6 or an RZ67 - a really high performance sensor on a body with leaf shutter lenses and interchangeable finders.

Remember, we used be able to do this with our Mamiya 645s and Bronica ETRS's: 35mm backs for those "special emulsions" not available in 120/220.
And that was when 35mm format was 2.7x smaller in area than 645 film; now 35mm full frame is only 1.7x smaller than a P30+, H4D-40, or Aptus 8.

So think of a CMOS 35mm sensor as a "special emulsion" for high ISOs, long exposures, and faster frame-rates.

I can see less merit in the idea for the Mamiya/Phase 645 users - fixed prism, focal plane shutter - we can already use our lenses on a FF DSLR anyway.
But maybe it would suit someone who is heavily invested in the new LS lenses and the DF body.

Ray

I doubt it. With digital backs, ISO setting, white balance, and focus check is what really needed.

Pingang
Shanghai

[ondebanks]

Nice to see all the responses and viewpoints so far.

The sensor size is one of the main benefits of medium format, imo, so your suggestion makes little sense. The lenses are big, expensive and heavy because they create large image circles which would be wasted here. The maximum apertures are not so fast because they don't need to be on a large sensor, but they wouldn't be so well matched to a 35mm sensor. Compared to a 35mm DSLR you would be giving away a lot of usability and features, just for an extra stop of sync speed. I don't see this working at all, sorry.

Graham, you're always a rational guy so I am quite surprised at your response - of all MF users, Hy6 owners would stand to gain virtually the most from this idea. Have you never used a DSLR and thought "damn, I'd love if this thing had an optional WLF/chimney finder!" (I miss the Nikon F series with their interchangeable finders...terrible shame that the Nikon D series never had them!) But your Hy6 has them. And you cannot use your fabulous Rollei lenses on any other smaller format camera, because no other camera can operate the electronic aperture mechanisms (the leaf shutter could be left open on a DSLR). This idea would allow your lenses (retaining their fast synch) to do "double duty" with high ISO, live view, unlimited long exposures...heck even video recording could be part of the back's abilities.

So to me, what you're really saying is not that "sensor size is one of the main benefits of medium format", but that in your view it is the only benefit of medium format. Take away the sensor size advantage, and there's nothing else about the cameras and lenses that make them special to you? As I said, I find that view really surprising.

Ray

[MrSmith]

So to me, what you're really saying is not that "sensor size is one of the main benefits of medium format", but that in your view it is the only benefit of medium format. Take away the sensor size advantage, and there's nothing else about the cameras and lenses that make them special to you? As I said, I find that view really surprising.

Ray

i have to agree with that statement, poor noise performance at mid-high iso, no usable live view, limited lens choice and cost. all minus points for MFD. sensor size is the only plus point for me.

[ctz]

i have to agree with that statement, poor noise performance at mid-high iso, no usable live view, limited lens choice and cost. all minus points for MFD. sensor size is the only plus point for me.

+1.

[uaiomex]

Hasselblad had the OP proposed back years ago. The original CFV back is the thing. At this stage of time, I think it is possible to create an "economical" 36X36 CMOS sensor. Though it sounds super good to me to get into DMF for cheap and use my beloved 500CM, it may not be popular with present owners that can afford the luxury of the bigger sensors. However, the proposal is a fine idea.
Eduardo

[ondebanks]

Hasselblad had the OP proposed back years ago. The original CFV back is the thing. At this stage of time, I think it is possible to create an "economical" 36X36 CMOS sensor. Though it sounds super good to me to get into DMF for cheap and use my beloved 500CM, it may not be popular with present owners that can afford the luxury of the bigger sensors. However, the proposal is a fine idea.
Eduardo

Eduardo, I'm glad you agree with the proposal.

But there never was a 35mm-format CFV back, and there never was a DSLR-originated CMOS sensor in any MF back. So when you say,  "Hasselblad had the OP proposed back years ago. The original CFV back is the thing.", do you just mean that since this was a highly cropped back, at 36X36 mm, it had a resemblence to my idea? That may be the case, but there the similarity ends. I'm talking about a whole different type of sensor, with whole different types of purposes.

Ray

[PdF]

Is this the solution ?

PdF

[uaiomex]

Right. There never was a 24X36 CFV but the heavily cropped 36X36 sensor designed especially for the V system is nothing more than a compromise between the square medium format camera camera and 35 implemented into the digital era. The sensors evolved and the 36X36 yield to the bigger sensors. I like your proposal of using a modern 24X36 cmos sensor for high ISO photography and Live View among other amenities, amen of the inexpensive cost. The 36mp Sony sensor would be great inside a CFV (for an example), but I'd like to go further and I would ask Sony (if I was Hasselblad) to cut a square 36X out of those wafers used for the Nikon D800. The advantages of a square sensor are several in my opinion. Few months ago, TOP blog published an article I wrote on behalf of square sensors. Please read it and let me know what you think. My idea was thought on existing full-frame systems but your proposal is for existing MF systems. If we combine your idea and mine, it sounds like Mozart to me!
Eduardo
http://theonlinephotographer.typepad.com/the_online_photographer/2012/01/squares.html



 

Eduardo, I'm glad you agree with the proposal.

But there never was a 35mm-format CFV back, and there never was a DSLR-originated CMOS sensor in any MF back. So when you say,  "Hasselblad had the OP proposed back years ago. The original CFV back is the thing.", do you just mean that since this was a highly cropped back, at 36X36 mm, it had a resemblence to my idea? That may be the case, but there the similarity ends. I'm talking about a whole different type of sensor, with whole different types of purposes.

Ray