Reading mode: Light Dark

Interesting 1Ds vs 1D pixel rate dilemma

Go to last post
C

Christopher Fritz

Forum Enthusiast
Messages
400
Reaction score
0
Location
TX, US
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS sensor estimated @ 3fps), there is an interesting pixel rate conflict.

It was stated in an interview with Canon 1D engineers that the reason they utilized a CCD image sensor is that they were not able to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high speed readout capabilities (and hopefully lower noise).

2. Given that the frame rate is the only estimated fact from above, the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
 
Chris said:
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS
sensor estimated @ 3fps), there is an interesting pixel rate
conflict.

It was stated in an interview with Canon 1D engineers that the
reason they utilized a CCD image sensor is that they were not able
to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or
the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate
from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high
speed readout capabilities (and hopefully lower noise).
2. Given that the frame rate is the only estimated fact from above,
the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
Click to expand...
 
What I read on the specs said it would be an 11.1Mpixel CMOS camera. Only the 3fps was speculation, unless the info on the canon web site was not correct.

Chris
 
that Canon reads the chip from two or more points ... as is done with the current 1D ... essentially two chips on one silicon die. If you have four 'feeds' ... you could, in theory, quadruple the read rate ... at a much greater expense, but suitable for a top-shelf camera.
Ken
Chris said:
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS
sensor estimated @ 3fps), there is an interesting pixel rate
conflict.

It was stated in an interview with Canon 1D engineers that the
reason they utilized a CCD image sensor is that they were not able
to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or
the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate
from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high
speed readout capabilities (and hopefully lower noise).
2. Given that the frame rate is the only estimated fact from above,
the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
Click to expand...
 
This is an interesting thought, but has inherent problems as seen in the banding of the 1D.

If two or more readouts are utilized to increase the pixel rate, there are two different amplifiers with two different characteristics. It becomes very difficult to balance each channel. I have seen this problem with CCD multi-readout sensors before and I expect that is what causes banding problems with the current 1D camera.

Since CMOS sensors are different, perhaps it can utilize one readout for the Green channel and the other one for the Red and Blue channel. Then, any readout amplifier mismatches would only be a white balance error and not line noise.

But this is pure speculation,
Chris
Ken Phillips said:
Chris said:
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS
sensor estimated @ 3fps), there is an interesting pixel rate
conflict.

It was stated in an interview with Canon 1D engineers that the
reason they utilized a CCD image sensor is that they were not able
to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or
the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate
from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high
speed readout capabilities (and hopefully lower noise).
2. Given that the frame rate is the only estimated fact from above,
the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
Click to expand...
Click to expand...
--
Chris
 
Maybe they have solved the problem to use TWO CMOS chip in one DSLR. So the D120/1Ds is acutally using two D60 CMOS sensor on a better body. The data rate problem can be solved as we have 2x19 mbps now.

Andy
Ken Phillips said:
Chris said:
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS
sensor estimated @ 3fps), there is an interesting pixel rate
conflict.

It was stated in an interview with Canon 1D engineers that the
reason they utilized a CCD image sensor is that they were not able
to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or
the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate
from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high
speed readout capabilities (and hopefully lower noise).
2. Given that the frame rate is the only estimated fact from above,
the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
Click to expand...
Click to expand...
--
Regards,

Andy
 
Chris said:
1. The 1Ds CMOS sensor uses a "new generation" sensor with a high speed readout capabilities (and hopefully lower noise).
Click to expand...
Or, just add more readout channels. Some CCD sensors (SITe for example) have four readout channels and the camera manufacturer can use some or all if desired. In this price range the cost of the extra electronics is manageable.
Christopher Fritz said:
2. Given that the frame rate is the only estimated fact from above, the 1Ds frame rate is only 2 frames per second.
Click to expand...
Look at the potential customer for the 1D. Are they going to pay $5000+ for a camera with less than a 5 frame/sec continuous shooting rate? No way.

The point of the 1D is not resolution. If Canon can add resolution without harming anything else that's fine, but if they give up on the other features of the camera it just becomes a D60 - at maybe 1/3 the price. Canon is trying to defend a $5000 price-point, and there’s no way to do that by staking your fame on resolution.

Another problem with a 11 MP or 12 MP sensor is the size of the buffer store. A 12 MP sensor produces an 18 MB RAW file at 12 bits/pixel. A 16 frame buffer is nearly 300 MB of RAM! Moreover, each frame takes 3+ seconds to flush with the fastest cards available today.

Features I’d expect to see in the 1D’s eventual replacement are full 35mm frame sensor, a longer buffer store, better on-camera image review, Lithium-Ion batteries, better low-noise performance, and a 6 MP or maybe 8 MP sensor.

A “Studio 1D” is another matter – all-world frame rates aren’t as important, and a computer is either tethered or easily at hand for frame storage. But I consider a studio camera such a different thing as to not be a direct comparison with an action/sports camera like the 1D.

PS. It is my understanding that Canon uses Interline-transfer CCDs to achieve the 8 frame/sec rate. It’s possible that this isn’t available in the densest sensors (highest pixel counts) due to the already too-small pixels.
 
James Van Artsdalen said:
Or, just add more readout channels. Some CCD sensors (SITe for
example) have four readout channels and the camera manufacturer can
use some or all if desired. In this price range the cost of the
extra electronics is manageable.
Click to expand...
If it were that easy with no other problems, why does the 1D have a CCD? It is difficult to balance different channels. The dual channel readout, in my opinion, is why the 1D has banding problems.
James Van Artsdalen said:
A “Studio 1D” is another matter – all-world frame rates aren’t as
important, and a computer is either tethered or easily at hand for
frame storage. But I consider a studio camera such a different
thing as to not be a direct comparison with an action/sports camera
like the 1D.
Click to expand...
I agree. I believe that it will be targeted at a different market, the studio. I also believe it will be either 2 fps or a new technology CMOS imager with potentially different characteristics than the D60's sensor. It is their on-chip noise reduction circuitry that works so well and has to be redesigned for a higher speed readout.

--
Chris
 
Is it possible that the 1Ds does 8fms/sec on a lower resolution, and maybe 2 or 3 at the highest resolution. So you could use it for Studio AND for sports/action???
Chris said:
James Van Artsdalen said:
Or, just add more readout channels. Some CCD sensors (SITe for
example) have four readout channels and the camera manufacturer can
use some or all if desired. In this price range the cost of the
extra electronics is manageable.
Click to expand...
If it were that easy with no other problems, why does the 1D have a
CCD? It is difficult to balance different channels. The dual
channel readout, in my opinion, is why the 1D has banding problems.
Christopher Fritz said:
A “Studio 1D” is another matter – all-world frame rates aren’t as
important, and a computer is either tethered or easily at hand for
frame storage. But I consider a studio camera such a different
thing as to not be a direct comparison with an action/sports camera
like the 1D.
Click to expand...
I agree. I believe that it will be targeted at a different market,
the studio. I also believe it will be either 2 fps or a new
technology CMOS imager with potentially different characteristics
than the D60's sensor. It is their on-chip noise reduction
circuitry that works so well and has to be redesigned for a higher
speed readout.

--
Chris
Click to expand...
--
Regards
Tom
 
Not likely, I say look at the Nikon model, I would assume that the Canon model wold be similar. The 1D is a much better D1h and the 1Ds will be a much better D1x. I would be surprised if the 1Ds shot frames faster than 3-4FPS regardless of mode chosen.
Thomas Streng said:
Chris said:
James Van Artsdalen said:
Or, just add more readout channels. Some CCD sensors (SITe for
example) have four readout channels and the camera manufacturer can
use some or all if desired. In this price range the cost of the
extra electronics is manageable.
Click to expand...
If it were that easy with no other problems, why does the 1D have a
CCD? It is difficult to balance different channels. The dual
channel readout, in my opinion, is why the 1D has banding problems.
[email protected] said:
A “Studio 1D” is another matter – all-world frame rates aren’t as
important, and a computer is either tethered or easily at hand for
frame storage. But I consider a studio camera such a different
thing as to not be a direct comparison with an action/sports camera
like the 1D.
Click to expand...
I agree. I believe that it will be targeted at a different market,
the studio. I also believe it will be either 2 fps or a new
technology CMOS imager with potentially different characteristics
than the D60's sensor. It is their on-chip noise reduction
circuitry that works so well and has to be redesigned for a higher
speed readout.

--
Chris
Click to expand...
--
Regards
Tom
Click to expand...
--
Valliesto
'A hero is no braver than an ordinary man, but he is brave five
minutes longer.'
  • R.W. Emerson
 
Thomas Streng said:
Is it possible that the 1Ds does 8fms/sec on a lower resolution, and maybe 2 or 3 at the highest resolution.
Click to expand...
Maybe. The CCDs used for astronomy support a readout method known as "binning" where the charge from several "wells" are sent to the A/D at the same time - the values are added. The problem here is that you can't add adjacent "wells" in the sensor since they are detecting different colors. But there may be a way around this.
 
Maybe they'll KEEP the 1D as the fast camera much the same as Nikon have the D1H as their fast low res camera and Oly had the E100RS as the low res high speed version of the UZI and Canon have added a 3FPS Studio camera?..

If anyone else has mentioned this already I apologise.. but if they HAVE done this - Kudos to them as it won't devaluate the existing 1D and there's always be a market for it in sports

Just a thought

--
Olympus C2100UZI +B300, E10, Canon D60.

My Ugly mug and submitted Photos at -------->
http://www.photosig.com/viewuser.php?id=27855

 
Would be a neat trick if they can pull that off. However, how would they do that? I can't imagine a way right now. But hey, they have an entire R&D department for that, so if it's possible they will find a way.
James Van Artsdalen said:
Thomas Streng said:
Is it possible that the 1Ds does 8fms/sec on a lower resolution, and maybe 2 or 3 at the highest resolution.
Click to expand...
Maybe. The CCDs used for astronomy support a readout method known
as "binning" where the charge from several "wells" are sent to the
A/D at the same time - the values are added. The problem here is
that you can't add adjacent "wells" in the sensor since they are
detecting different colors. But there may be a way around this.
Click to expand...
--
In the beginning there was nothing, and then even that exploded.
 
Chris said:
If it were that easy with no other problems, why does the 1D have a CCD?
Click to expand...
Quality? CCDs are still believed to be less noisy, albeit more expensive. Maybe now CMOS is getting competitive, at least in 12 bit A/D uncooled uses.
Christopher Fritz said:
It is difficult to balance different channels. The dual channel readout, in my opinion, is why the 1D has banding problems.
Click to expand...
Another reason some sensors have multiple readouts is so that the camera manufacturer can test each readout circuit and use the one that happened to be least noisy.
Christopher Fritz said:
I also believe it will be either 2 fps or a new technology CMOS imager with potentially different characteristics than the D60's sensor.
Click to expand...
I still come back to the idea that a D60 sells for ~ $2000, a D60 with twice the resolution wouldn't sell for even $4000 (especially if Nikon releases their prosumer 11 MP at $3000), and that I think Canon would want to try to defend their $5500 price-point for the 1D. Had the press release called it at D60s we'd all know what it was, but calling it a 1Ds makes it a big mystery to me.

Here's an idea: Dell sells RAM in notebooks around $100 per 100MB. Those SoDIMMs might fit into a camera body. Two of those could carry 1 GB. Frame rate problem "solved" if you can wait several minutes for the CF to catch up!

PS. 11 MP in a 35mm frame is going to result is wells that are only half the size of the 1D.
 
If I recall, for a CMOS sensor, each pixel is individually addressable, while on a CCD sensor, the sensor is read a row at a time. Basically, when you read a CCD sensor, each column has a queue and you reaad one value off the queue at a time. If you read across all the columns at the same time, you are reading one row of data at a time. You have to read each row like this in order if you want to get the entire image off the CCD.

I believe with a CMOS sensor, you can read all the pixels at the same time.

So, I think the issue is not the pixel rate but more of how quickly can each pixel in a CMOS sensor be read and then cleared. Given how CMOS sensors are read, and if there is a limitation of being able to read and clear each pixel only 3 times per second, then its logical that the D30, D60, and 1Ds each have the same frame rate of 3fps.

Joo
Chris said:
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS
sensor estimated @ 3fps), there is an interesting pixel rate
conflict.

It was stated in an interview with Canon 1D engineers that the
reason they utilized a CCD image sensor is that they were not able
to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or
the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate
from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high
speed readout capabilities (and hopefully lower noise).
2. Given that the frame rate is the only estimated fact from above,
the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
Click to expand...
--
  • Canon EOS D-30 & PowerShot S100
  • Maybe one day I'll take a decent picture. In the meantime, I'll blame the equipment. :)
 
are left-side/right-side ... and sometimes you can notice this. But it should have nothing to do with the banding, which is essentially nonexistent in my camera. (s/n 13,xxx).
Ken
Chris said:
If two or more readouts are utilized to increase the pixel rate,
there are two different amplifiers with two different
characteristics. It becomes very difficult to balance each
channel. I have seen this problem with CCD multi-readout sensors
before and I expect that is what causes banding problems with the
current 1D camera.

Since CMOS sensors are different, perhaps it can utilize one
readout for the Green channel and the other one for the Red and
Blue channel. Then, any readout amplifier mismatches would only be
a white balance error and not line noise.

But this is pure speculation,
Chris
Ken Phillips said:
Chris said:
Based on some assumed specs of the new 1Ds camera (11.1Mpixels CMOS
sensor estimated @ 3fps), there is an interesting pixel rate
conflict.

It was stated in an interview with Canon 1D engineers that the
reason they utilized a CCD image sensor is that they were not able
to obtain 8fps with their current CMOS sensor technology.

Also, if we list the current Canon camera pixel rates:

Camera Sensor eff pixels fps pixel rate
--------- ------ ---------- ---- ----------
1D CCD 4.15Mpixels 8fps 33.2Mpixel/sec
D60 CMOS 6.3Mpixels 3fps 18.9Mpixel/sec
1Ds CMOS 11.1Mpixels 3fps 33.3Mpixel/sec

Given the current state of CMOS sensors at the time of the D60 or
the 1D, they were not capable of achieving > 19Mpixel/sec pixel rate
from their CMOS technology. This leads to one of two conclusions:

1. The 1Ds CMOS sensor uses a "new generation" sensor with a high
speed readout capabilities (and hopefully lower noise).
2. Given that the frame rate is the only estimated fact from above,
the 1Ds frame rate is only 2 frames per second.

Just some thoughts,
Chris
Click to expand...
Click to expand...
--
Chris
Click to expand...
 
Actually, this is very possible. In current CMOS sensors I know about, and most Sony CCD's, it is possible to read out every other, or every 4th etc. pixel. That is how a 5Mpixel prosumer CCD digital still camera can support video capture modes.

The only problem is in the artifacts generated by subsampling the pixels without proper optical low pass filtering. But, I have seen some very good processing to eliminate these artifacts.

It would be a interesting feature,
Chris
Peter (B) said:
Would be a neat trick if they can pull that off. However, how would
they do that? I can't imagine a way right now. But hey, they have
an entire R&D department for that, so if it's possible they will
find a way.
Click to expand...
 
It just might be that the "S" in 1DS refers to "studio". That would explain the pixel rate business.

Regards - Seth
 
Maybe that's what the "s" stands for - Studio.
James Van Artsdalen said:
A “Studio 1D” is another matter – all-world frame rates aren’t as
important, and a computer is either tethered or easily at hand for
frame storage. But I consider a studio camera such a different
thing as to not be a direct comparison with an action/sports camera
like the 1D.
Click to expand...
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top