Canon Testing a 75+ Megapixel EOS-1 Body? [CR1]

[Marsu42]

Japanese companies are blame/credit centric. Before the start of any project, a chain of blame (in the case of project failure) and credit (in case all works out well) has to be established. This is very time consuming.

Thanks a lot, I'll store this information in my permanent memory, it's always great to have people with professional experience disclose these things!

[bvukich]

The engineering of production equipment, and product testing equipment most often is a larger task that the original product development. Leading edge semi products are the worst, especially considering "leading Edge" is a fleeting concept. Add site development, personnel training, and local "government influence" two years to create a high tech fab facility is the best you could expect. (Mind you, these facilities are now constructed in what were/are 3rd world countries).

Japanese companies are blame/credit centric. Before the start of any project, a chain of blame (in the case of project failure) and credit (in case all works out well) has to be established. This is very time consuming.

You're assuming they'd be standing up a new fab to create this chip, or at least a new process, both would almost certainly be incorrect.

They almost certainly will be using their existing fab, and existing 500nm process.

Granted, this still isn't something you can stand up in an afternoon, but there's NO WAY it would take years.

[kaihp]

The engineering of production equipment, and product testing equipment most often is a larger task that the original product development. Leading edge semi products are the worst, especially considering "leading Edge" is a fleeting concept. Add site development, personnel training, and local "government influence" two years to create a high tech fab facility is the best you could expect. (Mind you, these facilities are now constructed in what were/are 3rd world countries).

Japanese companies are blame/credit centric. Before the start of any project, a chain of blame (in the case of project failure) and credit (in case all works out well) has to be established. This is very time consuming.

You're assuming they'd be standing up a new fab to create this chip, or at least a new process, both would almost certainly be incorrect.

They almost certainly will be using their existing fab, and existing 500nm process.

Granted, this still isn't something you can stand up in an afternoon, but there's NO WAY it would take years.

flanderscamera has said his opinion is based on professional experience in manufacturing.
What are your counter-opinions based on?

Sure, they could retro-fit an old fab with a newer process, but this is rarely done (the value of keeping the old fab running and the cost building a new one is higher than stopping the fab and retrofitting costs).

flanderscamera: I would have thought that Canon would put a new fab inside Japan, but I'm not exactly familiar with their fabs. Sounds to me you are suggesting to put a fab in countries like Indonesia, Philippines, or Vietnam.

[jrista]

Japanese companies are blame/credit centric. Before the start of any project, a chain of blame (in the case of project failure) and credit (in case all works out well) has to be established. This is very time consuming.

Ah, yes...Japanese business politics. I am curious how long that would actually take, though...the Japanese can also be exceptionally efficient when they want to be.

[bvukich]

The engineering of production equipment, and product testing equipment most often is a larger task that the original product development. Leading edge semi products are the worst, especially considering "leading Edge" is a fleeting concept. Add site development, personnel training, and local "government influence" two years to create a high tech fab facility is the best you could expect. (Mind you, these facilities are now constructed in what were/are 3rd world countries).

Japanese companies are blame/credit centric. Before the start of any project, a chain of blame (in the case of project failure) and credit (in case all works out well) has to be established. This is very time consuming.

You're assuming they'd be standing up a new fab to create this chip, or at least a new process, both would almost certainly be incorrect.

They almost certainly will be using their existing fab, and existing 500nm process.

Granted, this still isn't something you can stand up in an afternoon, but there's NO WAY it would take years.

flanderscamera has said his opinion is based on professional experience in manufacturing.
What are your counter-opinions based on?

Sure, they could retro-fit an old fab with a newer process, but this is rarely done (the value of keeping the old fab running and the cost building a new one is higher than stopping the fab and retrofitting costs).

flanderscamera: I would have thought that Canon would put a new fab inside Japan, but I'm not exactly familiar with their fabs. Sounds to me you are suggesting to put a fab in countries like Indonesia, Philippines, or Vietnam.

They've been using the same process in the same fabs for a decade for their DSLR sensors. Even at 75MP, 500nm is plenty fine enough for any feature, unless they move amps & ADCs on die (which would be awesome btw), or did fancy stuff like on die binning, in either of those cases I'm not even remotely qualified to make a statement of whether 500nm would be sufficient.  So my position is based on historical precedence, and simple logic.  If flanderscamera has evidence to the contrary I'm sure everyone would love to hear it, myself included.

Migrating to a new process, or even just to a larger wafer size in the same process, is incredibly expensive.  And a whole new fab with modern tooling is in the neighborhood of $1-2bn.  I would love to hear they are making that investment, and would love to be proven wrong.  And if that is the case, then flanderscamera's timeline would be just about right.  I just haven't seen anything to support that, so I'm skeptical.

[jrista]

The engineering of production equipment, and product testing equipment most often is a larger task that the original product development. Leading edge semi products are the worst, especially considering "leading Edge" is a fleeting concept. Add site development, personnel training, and local "government influence" two years to create a high tech fab facility is the best you could expect. (Mind you, these facilities are now constructed in what were/are 3rd world countries).

Japanese companies are blame/credit centric. Before the start of any project, a chain of blame (in the case of project failure) and credit (in case all works out well) has to be established. This is very time consuming.

You're assuming they'd be standing up a new fab to create this chip, or at least a new process, both would almost certainly be incorrect.

They almost certainly will be using their existing fab, and existing 500nm process.

Granted, this still isn't something you can stand up in an afternoon, but there's NO WAY it would take years.

flanderscamera has said his opinion is based on professional experience in manufacturing.
What are your counter-opinions based on?

Sure, they could retro-fit an old fab with a newer process, but this is rarely done (the value of keeping the old fab running and the cost building a new one is higher than stopping the fab and retrofitting costs).

flanderscamera: I would have thought that Canon would put a new fab inside Japan, but I'm not exactly familiar with their fabs. Sounds to me you are suggesting to put a fab in countries like Indonesia, Philippines, or Vietnam.

They've been using the same process in the same fabs for a decade for their DSLR sensors. Even at 75MP, 500nm is plenty fine enough for any feature, unless they move amps & ADCs on die (which would be awesome btw), or did fancy stuff like on die binning, in either of those cases I'm not even remotely qualified to make a statement of whether 500nm would be sufficient.  So my position is based on historical precedence, and simple logic.  If flanderscamera has evidence to the contrary I'm sure everyone would love to hear it, myself included.

Migrating to a new process, or even just to a larger wafer size in the same process, is incredibly expensive.  And a whole new fab with modern tooling is in the neighborhood of $1-2bn.  I would love to hear they are making that investment, and would love to be proven wrong.  And if that is the case, then flanderscamera's timeline would be just about right.  I just haven't seen anything to support that, so I'm skeptical.

Canon does have on-die amplifiers. As far as I know, they have per-pixel amps as well as an additional downstream amp. They also have on-die CDS. The old APS-C 18mp sensor seems to already be indicating that full well capacity is getting low enough that its problematic to IQ (and that is at 4.3 microns...a 75mp FF would have 3.4 micron pixels, and an even lower FWC...meaning even more noise.) With a 500nm process on an FSI design, the actual photodiode area at 3.4 microns would be around 2.3 microns. When small form factor pixels in phones and P&S cameras started reaching pixels that size a few years ago, they were already using smaller processes between 250nm and 180nm, and they were already starting to look to BSI to improve IQ.

I am not sure the 500nm process still really has much, if any, life left in it. Canon is really riding up against the walls of physics at this point, and smaller pixels will get progressively noisier without either a shift to a smaller process, or a move to BSI (or something else as radical...supercooling to improve Q.E. of their photodiodes in combination with lightpipes, etc.)

I am not necessarily saying you are wrong...I too am waiting to hear Canon has done something about their fabs. You are, actually, probably quite right. I guess that is rather depressing, however...having used the 7D for about 19 months now, I can't imagine what 3.4 micron pixels on a 500nm FIS design would be like...but I can't imagine them being better, or even as good.

[tpatana]

Time to put in an order for more external hard drives.

Here's how my PC root looked couple days ago when I had the card reader connected too:

[bvukich]

what do you mean JRISTA that smaller pixels will be noisier?

The snr on an individual pixel will get worse as it's scaled down, the snr of the sensor as a whole will (should) improve though.

[Aglet]

There's no perfect system for everything, which is why I use 3 of the majors (I'm a PentNikCan shooter).  It's taken some time to learn them all and get used to their individual pros and cons but it's been enlightening, worthwhile, and enjoyable for me.

Where I have to give Canon top kudos is their user manuals.
At least all the english ones I've read.

They're much better written, IMO, than the Nikon and Pentax manuals I've read within the last few years.  (I haven't had a recent Panasonic or Sony to compare.)
If a new user takes the time to completely read a Canon manual, cover-to-cover, they'd learn a LOT, even if they aren't using a Canon camera.

For this reason, I can still recommend Canon to a new user, they'd likely find it a less frustrating and more intuitive introduction to DSLR shooting.  Someone with more experience would do well, possibly better, with one of the alternatives, depending on what their intent was for such gear.

This kind of thing can result in more sales. and, once in, likely repeat customers.
Mfrs should really give more thought to how well their manuals are written - that's if the buying demographic still READS paper manuals...  All too often I see questions on various forums that are well documented in the basic user manuals.

[jrista]

what do you mean JRISTA that smaller pixels will be noisier?

The snr on an individual pixel will get worse as it's scaled down, the snr of the sensor as a whole will (should) improve though.

noise is lower from a smaller pixel

Read noise may be lower (depends on a number of factors)...total noise is not, simply because the SNR is lower (i.e. ~20k e- for the 7D, vs. ~90k e- for the 1D X...both are 18mp, but the amount of noise in a similarly framed photo from the 1D X will be significantly lower than that of a 7D. In a focal length limited scenario, the 7D will offer higher resolution despite the increase in noise.)

[neuroanatomist]

right answer will always be, smaller pixels, lower noise
NOISE
simple as that

Is it that simple?  What about the signal?

[jrista]

right answer will always be, smaller pixels, lower noise
NOISE
simple as that

That is read noise.

What about photon shot noise, which is relative to the signal? Are we just going to ignore that, now?

[jrista]

right answer will always be, smaller pixels, lower noise
NOISE
simple as that

Is it that simple?  What about the signal?

now it is motor cycles again
please learn to discuss the subject, noise
small pixel always gives lower noise than bigger, the signal has not been up to discussion
You and Jrista has some problem to know what we discuss or not
Must it be like this?
and are you going to make you funny about me again?
head room, signal/noise , fluorite glass ,  etc etc  what next?

Ok, lets get to the root of this here.


1. Assume we have a hypothetical sensor that introduces ZERO read noise whatsoever...no dark current noise, no high frequency noise...no read noise of any kind from any electronic source in the camera, on the sensor die or anywhere else...just for discussions sake.

2. That sensor captures an image projected by a lens in dim light, at an ISO setting of 1600.

3. Is that image noise free, or is there an intrinsic component of noise that is a very part of the image itself?


This is a test. Your answer does matter. You will be judged upon it. Go!

[neuroanatomist]

please learn to discuss the subject

Sage advice....try following it, instead of DRagging every thread down into the dolDRums.

[jrista]

right answer will always be, smaller pixels, lower noise
NOISE
simple as that

Is it that simple?  What about the signal?

now it is motor cycles again
please learn to discuss the subject, noise
small pixel always gives lower noise than bigger, the signal has not been up to discussion
You and Jrista has some problem to know what we discuss or not
Must it be like this?
and are you going to make you funny about me again?
head room, signal/noise , fluorite glass ,  etc etc  what next?

Ok, lets get to the root of this here.


1. Assume we have a hypothetical sensor that introduces ZERO read noise whatsoever...no dark current noise, no high frequency noise...no read noise of any kind from any electronic source in the camera, on the sensor die or anywhere else...just for discussions sake.

2. That sensor captures an image projected by a lens in dim light, at an ISO setting of 1600.

3. Is that image noise free, or is there an intrinsic component of noise that is a very part of the image itself?


This is a test. Your answer does matter. You will be judged upon it. Go!

test? what kind of tests?

I got guests here in my summer house, what kind of answer do you want?
A  pragmatic answer will be-YOU don't se any difference from middle grey up to high light, (I can call it infinity)
But you are going to se better results in the shadows - then there are a huge among of other answers
BUT first my guests who are staying here  2  days more.

and PS You will be judged upon it. Go!
what in h... is that?

And there you go, peoples! No answer, more obfuscation, and some beating around the bush about "guests".

I think we can safely come to the conclusion that Mikael does not understand the concept of intrinsic noise in an image signal.