Canon has held patents in this area for many years now, and we are likely over a year from the next 1 release. That is a lot of development time. When they do intro this new tech, yes, it will require a major systems overhaul. As for new technology, FF was introduced in the the 1 series, 1.3 has not been duplicated outside of the 1 series, and I see no reason for this to necessrily be altered in the future.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
The end of Bayer?
- Thread starter Mauka
- Start date
PalmsWestPhoto
Forum Pro
The sensor for the next 1 series is already done. They may still tweak it but they do not start developing a sensor that late
The patents that canon has are highly unlikely to be used in still photography as they pretty much wouldnt work too well for it.
--
Michael Salzlechner
http://www.PalmsWestPhoto.com
The patents that canon has are highly unlikely to be used in still photography as they pretty much wouldnt work too well for it.
--
Michael Salzlechner
http://www.PalmsWestPhoto.com
Derekasaurus Rex
Senior Member
I did not mention Moore's Law. Perhaps that's because I designed chips for a living and most likely know a lot more about the subject than you do.
That doesn't give me prescience, but it doesn't take a crystal ball of predict that 16.7MP is not the highest resolution that Bayer sensors will ever achieve. 22MP is a certainty and I can easily see resolution going higher into the 20's.
Henrik Andersson
Senior Member
True. It's a slower cycle then that. But sensor technology still improves by a generation. Even tough sensor sizes doesn't shrink with better performance per generation as logic does.
Did he claim Moore's law applied ?
We have 16MP at 35mmFF and 12MP at 1/1.5x crop now. 12MP at 35mmFF at 2700 Euro ...
Why would a 1DsmkIII with 22MP 35mmFF at PMA 2006 be impossible ? Enlighten us.
I've heard this from you before but WHEN will we see the light ?
3 month, 6month (PMA), 12 month (Photokina), 24 month, ... ? How will the 5D replacement look then ? or the 20D replacement (or replacement replacement) ?
I hope you are right about the Canon/Nikon trasher
Canon need competition.
--
Henrik
Mike Fried
Senior Member
Look at it this way:
SD9/10: 3.5M sensor locations, 10.5M sensors -> 3.5M pixels
EOS 5D: 12.8M sensor locations, 12.8M sensors -> 12.8M pixels
One camera stores 10.5M units of data, the other stores 12.8M. One camera has smaller, less sensitive sensors that show noise at ISO 400, the other has cleaner images at ISO 3200.
If the Foveon sensors were made larger than 3.5MP, perhaps we'd have something more to talk about, but as it is, you can take a 12.8MP sensor and scale it down by a factor of 4 and have sharper photographs at ISO 6400 than you can on the Sigma. Or you can have 4x the resolution. The color resolution isn't as important. Humans are more sensitive to contrast than to color.
-Mike
http://demosaic.blogspot.com
Laurence Matson
Forum Pro
There is more than enough evidence around that many imagers are reaching the physical limits of the pixel pitch. That is where Moore's law does not apply. You cannot just keep cramming pixels, which have a size limited by light waves, unto a limited space. Folks can tell me all about their being sensor designers; if they don't realize this, then I wish them lots of luck.
A Sony engineer told me after the released the 828 that they had gone past the physical limits of the state of the art on that imager, and the results were cräp in his own words. That pixel pitch was about 2.2 um. Imager designers say that the limit taking light into consideration and allowing room to actually gather it is around 1.7 um. Do the math and put that pixel on a decent sized imager that will actually work with the current crop of lenses and not just with the Zeiss 21 mm.
Then start to work on the noise.
I am not saying it is impossible. The statement "I see no reason to stop there" implies a continual, linear development track. The results from the current 5D would certainly open the eyes of some engineers and cause them to say the opposite: "I see a reason to stop there". This is not to say that things can change. However, the current state of the art is reaching its limits imo, and light limitations are not that far away now.
We'll see.
6month (PMA), 12 month (Photokina), 24 month, ... ? How
I agree.will the 5D replacement look then ? or the 20D replacement (or
replacement replacement) ?
I hope you are right about the Canon/Nikon trasher
Canon need competition.
--
Laurence
There is a tide in the affairs of men,
Which, taken at the flood, leads on to fortune;
Omitted, all the voyage of their life
Is bound in shallows and in miseries.
http://www.pbase.com/lmatson/root
http://www.pbase.com/sigmasd9/root
http://www.pbase.com/cameras/sigma/sd10
http://www.pbase.com/cameras/sigma/sd9
http://www.beachbriss.com (eternal test site)
Henrik Andersson
Senior Member
We are not even close to real problems there yet.
The step from Canons old process technology like 1Ds, 10D, 300D vs the new technology in 1DmkII, 1DsmkII, 20D, 350D and 5D show that there is no practical evidence of a slow down yet.
Even Olympus small 8MP sensor is doing well for its small size. The problem is when it's magnified alot.
As a side note: Even Canons tele zooms can out resolve 80MP Bayer today. The ultra-wides are little different though ... both on electrical sensors and chemical film :-(
That is a totaly different technology. It's not even remotely relevant here.
And not all engineers are that well informed.
I don't agree. One of the reasons they stopped at 12MP for the 5D was probably becouse of all the 'per pixel' comparing ppl.
For me they could have used more sensels. I'm not fooled by a few big 'noise free' sensels vs many 'noisier' sensels.
It's far away, just look at the best benchmark chemical philm and other equipment for current 35mmFF lenses. They clearly outresolve all current DSLRs.
How can these films capture detail that is not availible for even future sensors becouse of physical light limitations ?
:-(
I'm off to bed now.
--
Henrik
PalmsWestPhoto
Forum Pro
Laurence
obviously there is a limit but if you take the 20D sensor as a starting point and scale that up to FF you already end up at 21MP. That is using the same old sensor technology and getting the same good high ISO response and DR as the 20D. Obviously in the meantime canon made their sensors better as well
We could easily see a 24MP sensor in a FF canon body with current high ISO response.
This would offer plenty of resolution for anyone. Actually too much for most people allready. The current 12 and 16MP bayer sensors allready do plenty anyway
As far as resolution limits go though the D2X has 12MP on a 1.5 so that could scale easily to 27MP and the D2X resolution is very good. The high ISO is not that great but if you dont need that it is still a good option.
Obviously we dont want to get down to small sensor pitch but we dont have to really. Anything around the 20MP mark will pretty much be enough for 35mm cameras. If you need more get a digital MF camera.
--
Michael Salzlechner
http://www.PalmsWestPhoto.com
obviously there is a limit but if you take the 20D sensor as a starting point and scale that up to FF you already end up at 21MP. That is using the same old sensor technology and getting the same good high ISO response and DR as the 20D. Obviously in the meantime canon made their sensors better as well
We could easily see a 24MP sensor in a FF canon body with current high ISO response.
This would offer plenty of resolution for anyone. Actually too much for most people allready. The current 12 and 16MP bayer sensors allready do plenty anyway
As far as resolution limits go though the D2X has 12MP on a 1.5 so that could scale easily to 27MP and the D2X resolution is very good. The high ISO is not that great but if you dont need that it is still a good option.
Obviously we dont want to get down to small sensor pitch but we dont have to really. Anything around the 20MP mark will pretty much be enough for 35mm cameras. If you need more get a digital MF camera.
--
Michael Salzlechner
http://www.PalmsWestPhoto.com
Leon Wittwer
Forum Pro
The problem with a new sensor technology is that it has to arrive fully competitive in cost and performance with existing sensors that already do the job to most peoples satisfaction. In fact, it likely has to be significantly better otherwise no one will buy it as a replacement for the old system. To ask for a new sensor method to do this is very tough considering the current performance of today's Bayer sensors. That was Sigma's problem with the Fovean sensor. If it has been marketed 2-3 years earlier when it was potentially better than the Bayer sensors, it might have flown. As it happened, the Sigma system, which attracted some by its interesting concept, did not do that much better a job at producing images as compared to the existing Bayer systems when it was introduced. Its market share has remained small.
Leon
http://homepage.mac.com/leonwittwer/landscapes.htm
--
Leon
http://homepage.mac.com/leonwittwer/landscapes.htm
Derekasaurus Rex
Senior Member
No, it doesn't. That's your assumption. Only a fool would think any technology could improve linearly indefinitely.
The 1Ds2 has already shown us that you can go beyond the 5D. And current technology is still far from optimal, with lots of room for improvement in just about every way. These improvements can't increase resolution forever, but they can easily take megapixels into the 20's, and some would argue beyond.
It this a fact? I had not heard this yet. What are you basing this on?
Kent D
Well-known member
The limit of resolution will be determined not by how small we can manufacture sensors but by the amount of light which hits each sensor site. We all know that as the pixels get smaller the light hitting each diminishes. The inherent signal to noise will be limited by counting theory. This is a basic mathematical truth which is true of both Bayer and RGB sensors such as the Foveon.
The inherent problem with Bayer is that it throws away 2 out of 3 photons because they are the wrong collor for that sensor site. Thus the Bayer approach is theoreticaly going to be limited more in terms of signal to noise for a given maximum resolution. Conversely an RGB Sensor, in theory can use every photon.
--
Kent Dooley
The inherent problem with Bayer is that it throws away 2 out of 3 photons because they are the wrong collor for that sensor site. Thus the Bayer approach is theoreticaly going to be limited more in terms of signal to noise for a given maximum resolution. Conversely an RGB Sensor, in theory can use every photon.
--
Kent Dooley
Fair enough, but Canon is a forward thinking company. They are quite aware of how far they can push Bayer, and have been for some time. They also know it's shortcomings (they get an earful on these boards). They are also aware that the megapixel race is hard stop once started. Knowing that they will abandon Bayer, it is a matter of when. The problem you have posed is real, but smart companies deal with such problems by tandem development teams that share info. To think that canon will take all of it's engineers off of Bayer products and set them on this new track would be unlikely. When the new is competitive, then the time has come.
Mike Fulton
Senior Member
Uh, there's no basis I can see for this idea. Canon's done nothing more than continue to improve their bayer-sensor technology, while the only existing single photosite sensor maker, Foveon, hasn't announced anything new in 3 years.
I thought the Foveon sensors sounded like a great breakthrough when they first announced them, but the bottom line is that they've essentially gone nowhere since then, while the rest of the world has continued to move forward.
Get excited when FOVEON announces something.
Mike Fulton
Senior Member
I suspect you'd find that the Foveon sensors do essentially the same thing. I don't see how you'd explain the comparatively poor noise performance if they were really that much more light sensitive.
Mike Fulton
Senior Member
When has Canon said they were working on non-bayer image sensors?
Derekasaurus Rex
Senior Member
Canon has a patent on an RGB sensor technology like Foveon's, but they've done even less with the technology than Foveon, at least commercially. I'm sure they've built prototypes, but there's nothing to indicate that they aren't going to milk Bayer for generations to come.
Keyboard shortcuts
- f
- Forum