Canon EOS 3D X [CR1]

[wickidwombat]

With the specs listed, I really wonder if Canon started this project years ago with the development of the 7D...

Canon APS-C sensor: 22.2mm x 14.8mm
35mm Full Frame: 36mm x 24mm

Just going by the sensor dimensions, the 18 megapixel sensor on the 7D is equal to a 47.3 MP sensor when expanded to full frame 35mm.  The leaked numbers of a 46MP sensor are just too close for coincidence.

My curiosity is piqued.

if it also gets the low iso noise of the 18mp sensor there are going to be ALOT of tears

[dilbert]

With the specs listed, I really wonder if Canon started this project years ago with the development of the 7D...

Canon APS-C sensor: 22.2mm x 14.8mm
35mm Full Frame: 36mm x 24mm

Just going by the sensor dimensions, the 18 megapixel sensor on the 7D is equal to a 47.3 MP sensor when expanded to full frame 35mm.  The leaked numbers of a 46MP sensor are just too close for coincidence.

My curiosity is piqued.

A number of us have been saying things like this for some time.

[justsomedude]

if it also gets the low iso noise of the 18mp sensor there are going to be ALOT of tears

I probably should've been more clear in my first post; I don't think this body will be remotely close to the 7D.  In fact, I don't think it's going to be in the same universe.

I just think that Canon had this sensor/body in the works for quite some time.  The 7D was years in development itself, and just based on the MP count, it looks like that research has progressed to bring us the 3DX.

I'm glad Canon cut the ISO at 6400, and has a native ISO 50 with an expandable 25.  This is clearly a studio unit (perhaps some solid landscape applications as well) that will have some killer low ISO performance.

I just hope it's real!

[nicku]

Canon is obliged to meet the challenge imposed by Nikon D800. I believe this camera will appear. The price tag will be around ( in my opinion) $4499 - 4990.

Regarding the lens potential on a such big MP sensor.... 7D has the same amount of pixels /square cm, and the quality lens meet those demands.

[Ew]

A body like this would definitely yield a new UWA, and perhaps a TSE update ?

[funkboy]

All investigations that I have seen suggests that current 14-bit cameras tend to be noise-limited (photon/electronics noise is large enough that the quantization noise (error) is dwarfed), and that the 15th and 15th bit of 16 bit raw files contain only random noise. Random noise needs not be recorded, it can be recreated in your computer at any time at much lower cost.

Agreed.  Medium format cameras have 16-bit ADCs & RAW because (historically) their pixels are so much larger & thus capture enough photons to warrant the additional bit depth.

BUT, an additional two bits on DSLRs would lessen the need for exposure to the right.

[Aglet]

There is much debate about the "16-bit myth." From what I've read on the matter, I don't believe that a 16-bit camera would produce any tangible improvement over a 14-bit camera, because those extra bits are not actually doing anything useful, just quantizing noise.

Certainly the case on Canon's bodies to date.  Only about 12b of good data there.  Maybe 12.5 to 13 on the 1Dx

[Aglet]

BUT, an additional two bits on DSLRs would lessen the need for exposure to the right.

not unless the noise levels went down by an equivalent 2 bits or more
unlikely

[wockawocka]

You had me until:

'Another feature mentioned about the camera is better heat dissipation of the sensor compared to other EOS cameras. Apparently, this results in industry leading low ISO performance.'
 

[Aglet]

I'm glad Canon cut the ISO at 6400, and has a native ISO 50 with an expandable 25.  This is clearly a studio unit (perhaps some solid landscape applications as well) that will have some killer low ISO performance.

I just hope it's real!

I hope it's REAL too!
But Canon has to make some serious improvements in their hardware to reduce the (patterned) read noise that is the bane of their current technology.
Without a significant reduction in read-noise, the lower ISO settings will be useful for wider apertures, or slower shutter speeds, but without a commensurate improvement in DR or need for extra digitizing bits.

DR should increase by about the same amount as ISO decreases (stop-per-stop) but if you look at Canon's measured DR curves, they're pretty flat from ISO 800 down.  E.G. The 5D3's DR is only moving about 1 stop despite 4 stops worth of ISO change.

Recent SoNikon's DR curves behave as they should, DR increasing equivalent to the decrease in ISO.
If Canon can fix this, for a new high-MP body, they'll please a lot of their current customers who've been hoping for such an improvement for years!

And I also hope that, if they DO accomplish this, we aren't hosed 10x as much $ as such DR technology can be had from competitor's products for as little as $500 in a consumer crop body.

[funkboy]

BUT, an additional two bits on DSLRs would lessen the need for exposure to the right.

not unless the noise levels went down by an equivalent 2 bits or more
unlikely

There are a few main arguments for doing ETTR:

• Maximize the dynamic range of the exposure (i.e. use the full range of the camera's DR)
• Maximize the SNR by capturing as many photons as possible
• Minimize the effects of quantization by trying to get as much of the image as possible into the exposure stops with the most tonal values available

16-bit ADC & RAW will increase the number of tonal values available in each EV (the third point above) by a factor of four vs. 14-bit.

[pakosouthpark]

i think this will be the inside of the 3Dx

[Marsu42]

• Maximize the SNR by capturing as many photons as possible

... unless you have to raise iso to get an ok shutter speed when overexposing, so often anti-ettr by underexposing and/or lower iso might give you a higher snr :-p

[NotABunny]

This will obviously be called 0DX 

[tg]

I think the lack of people talking about bit depth is surprising, I think it should be the main thing to consider, more so than mp. I'd much rather have 16bit version of 5dmkiii than a 14bit higher mp camera. It's the main aspect I'm considering investing in medium format.

I think it's not talked about much because it just won't happen for another generation of canon pro bodies,probably... but like you, I would be more than happy to be wrong!

You might want to have 16 bits, and you might feel better if you had. But chances are that your images  would not be any better.

All investigations that I have seen suggests that current 14-bit cameras tend to be noise-limited (photon/electronics noise is large enough that the quantization noise (error) is dwarfed), and that the 15th and 15th bit of 16 bit raw files contain only random noise. Random noise needs not be recorded, it can be recreated in your computer at any time at much lower cost.

I would much rather have good image quality, or at least, some measurable and relevant indicator of image quality (such as dynamic range at base ISO), rather than some random PR-driven spec that few people understand, such as the number of bits in an A/D-converter.

-h

My point, as I hoped others would notice: there is a real, tangible, visible difference in MF files vs dslr files when it comes to colour. (think transitions/gradations/subtleties in skin tone) It's not necessarily 'more' ,like mega pixels, but rather higher sensitivities... and yes, of course there's factors that allow this, like larger pixels and great lenses, but that's what I'm saying I'd love to have in my dslr... larger pixels (the 1Dx has large pixels) with higher colour sensitivity. And I'd be fine if that meant a loss in hi iso and speed.

And the idea that few people understand all the tech behind it, I agree. I am one of them, trying to learn more. But I'm definitely not convinced that the progression to 16 bit is a bad idea. It needs to be implemented properly, and I wouldn't be surprised if it became something we all understood a bit better in the coming year or two...