the point is that they already got the FF competition that you are talking about. the D3 is a pro FF body that is as fast as a 1D series and has the IQ of the 5D.
Not to put a too fine of a point to it, but back when Canon introduced the 5D they already voiced their commitment to full frame
Brian Worley, product manager of D-SLR cameras at Canon Europa, said that it's a long way before all EOS models will have a full size sensor, but Canon's aim is to equip all models, except the cheapest, with a 35mm sensor in the future.
Tentacle wrote:
Numbers for ISO 200 are wrong in the first place because that's not base ISO, regardless of when DR is highest. At base ISO you assume zero gain before the A/D conversion takes place. At ISO 200 there is some amplification taking place. How much? Hard to tell. Is the gain for the 40D sensor the same as the 1D3 sensor? We don't know.
slightly OT, but why is it (if you know) that Nikon can't make a camera with ISO 100, even in the D3?
blonde wrote:
the point is that they already got the FF competition that you are talking about. the D3 is a pro FF body that is as fast as a 1D series and has the IQ of the 5D.
If that's addressed to me, the answer is NO, there is NO FF "price point" competition. The D3 offers lower resolution than the 5d and costs almost 3x the price. The D3 offers only 57% of the 1Ds3 resolution and costs 62% of the price.
Where's the head to head competition? Doesn't exist....
my point was that now that Nikon has gotten into the FF, they should have no problem (and they will) put that sensor into a D300 body and than, the 5D is finished.
Tentacle wrote:
Numbers for ISO 200 are wrong in the first place because that's not base ISO, regardless of when DR is highest. At base ISO you assume zero gain before the A/D conversion takes place. At ISO 200 there is some amplification taking place. How much? Hard to tell. Is the gain for the 40D sensor the same as the 1D3 sensor? We don't know.
I presume you mean that the use of such numbers is wrong; the numbers are not wrong, while they are my own data they agree well with those of others (see eg Roger Clark's compendium of sensor data at clarkvision.com). If you want to do the comparison at ISO 100 that's fine with me; double the figure for electrons/raw level for both cameras, it's a wash for the relative comparison. There is a slight added feature that for both cameras the sensor saturates before the max raw level of 16383; the 1D3 saturates at 15280, the 40D at 13824. If one computes the full well capacity, it's 71K electrons for the 1D3, and 40K electrons for the 40D; scaling that to the 1D3 pixel size that's a little over 63K electrons, not a substantial difference. All this really means is that the native ISO on the 40D sensor is slightly higher than the 1D3 sensor. It actually is more relevant what is the DR at the ISO where it is maximized, and that is ISO 200, which is why I did the comparison there.
You seem to be under some sort of misapprehension as to what ISO means. It is a fixed international standard of signal amplification. Both the 40D and 1D3 adhere to the same standard (older generations of Canons measured at 1/3 stop more sensitive than the standard, but consistently for cameras of a given generation), and so have the same signal amplification taking place when set to the same ISO.
Since a given amount of light should result in a given exposure, at set ISO sensitivity, it's obvious that both 40D and 1D3 are equally sensitive.
No, it is not obvious. The 1D2N has vastly poorer sensitivity (in electrons per raw level per unit area) than the 20D, although they are from the same generation of technology. When I write "sensitivity" here, I mean light collecting efficiency, not ISO gain. Comparisons should of course always be done at the same ISO gain.
Besides, since you have divided the light "into smaller parcels" the error in measurement (the A/D conversion process) will become relatively bigger. Noise will become more apparent.
Only at the pixel level, which disadvantages the small pixel sensor by measuring its properties at a higher spatial frequency (smaller scale) than the large pixel camera. What I have been trying to show you (apparently without success) is that if you compare the two sensors at the same spatial frequency (same scale) there is zero difference.
Rather than presenting a rather garbled theory, show me the data that refutes my analysis.
Google for, and read, the Canon Full Frame White Paper. Page 5:
Yes, you and the Canon marketing department are equally misinformed about how sensors make images. You would both be right if one were only intent on putting images on computer screens and staring at small parts of them at 100% magnification. If instead one likes to view the whole image at a fixed size then my analysis is the proper way to do comparisons.
Edited by ejmartin on Mar 11, 2008 at 10:35 AM GMT
blonde wrote:
my point was that now that Nikon has gotten into the FF, they should have no problem (and they will) put that sensor into a D300 body and than, the 5D is finished.
... in 2010 perhaps and at a higher price point. By which time the 5D successor will have raised the bar and (hopefully) lowered the cost of joining the FF club even further.
The D3 is mor than 2.5 times the price of the D300. For what? a battery grip and FF sensor? (OK, so a bit more but you get my point) Makes me think that any "compact" FF body from Nikon will also have a hefty premium that Canon will easily be able to undercut.
slightly OT, but why is it (if you know) that Nikon can't make a camera with ISO 100, even in the D3?
I don't know for sure, but I can do an educated guess.
Can Nikon design / let Sony manufacture sensors with base ISO 100? Sure they can! But it's a matter of deciding what's more important to them. Semi conductors can be tuned in various ways, including the sensitivity to light.
The challenge is the signal amplification (gain) you need to apply at ISO values above base ISO. If base ISO is 100 then native ISO 3200 (not pushed ISO H) means you've had to double the original signal 5 times. Amplification by a total of 32x. That's quite a lot, considering you have to do that on a sensor, where heat generation and electrical noise is an issue.
Now, if you increase the base sensitivity a stop to ISO 200, by changing the type and/or concentration of dopants in your photosite semiconductor, you can keep the same 32x amplification and reach ISO 6400 without pushing. It saves a lot of headaches.
Nikon (or Sony?) decided to sacrifice true ISO 100, going for a pulled ISO 100 instead, in order to make it easier/less noisy to get to true ISO 6400 for the D3 and true ISO 3200 for the D300.
If ANY of these 5D succsssor predictions proves to be true, continuing Canon's "as little as we can get away with" approach to succession planning, I'm willing to end any future speculation right now by naming the D3 as my 5D successor, lens hit be damned.
jamesf99 wrote:
slightly OT, but why is it (if you know) that Nikon can't make a camera with ISO 100, even in the D3?
The higher base ISO of the D3 is an indication that it is natively more efficient at collecting photons, so that the longer exposure entailed by the use of ISO 100 would saturate the pixels too easily, resulting in a loss of headroom above middle gray and thus dynamic range in highlights.
The "native ISO" of the 1D3 and 40D are also higher than 100, BTW. One can see this in that ISO 100 exposures saturate at raw level 15280 and 13824, respectively, out of 16383. The ISO that would make use of the full set of available raw levels is a bit above 100 for both cameras.
To my mind, higher base ISO is a good thing -- it means that the sensor is more efficient at capturing photons, and so has higher S/N ratio at any given ISO.
blonde wrote:
my point was that now that Nikon has gotten into the FF, they should have no problem (and they will) put that sensor into a D300 body and than, the 5D is finished.
You may be right, but until they do, Canon will sell mediocre upgrades for as long as they can and my bet is that's at least another 12 months or more....
I want to make it clear that I'm not simply bashing Canon. I've used their products for over 30 years and have enjoyed them quite a bit. They used to be, as in the old days, competitive on pricing. They left that outmoded ideal behind long ago and now want to provide as little, for as much, as possible. What company doesn't though?
I continue to buy and use their products, but to be honest I'm sick of it. I want some value for my money, and Canon is no longer the company that can deliver it......
Every time I think of the 1Ds3 now, a camera that has been budgeted and funds allocated for 18 months +, I want to spit in a bucket if you know what I mean.
Again, no head to head, no excellent products. Period....
ejmartin wrote:
The higher base ISO of the D3 is an indication that it is natively more efficient at collecting photons, so that the longer exposure entailed by the use of ISO 100 would saturate the pixels too easily, resulting in a loss of headroom above middle gray and thus dynamic range in highlights.
The "native ISO" of the 1D3 and 40D are also higher than 100, BTW. One can see this in that ISO 100 exposures saturate at raw level 15280 and 13824, respectively, out of 16383. The ISO that would make use of the full set of available raw levels is a bit above 100 for both cameras.
To my mind, higher base ISO is a good thing -- it means that the sensor is more efficient at capturing photons, and so has higher S/N ratio at any given ISO....Show more →
Thanks. I guess I disagree though as I would much prefer a native ISO 50 option, let alone ISO 100. I've been in situation where I was fully stopped down, max shutter speed, and didn't have a ND filter with me. Even 1 more stop would have helped, but ISO 200? Nah.... I want some "slow time".
Remford wrote:
If ANY of these 5D succsssor predictions proves to be true, continuing Canon's "as little as we can get away with" approach to succession planning, I'm willing to end any future speculation right now by naming the D3 as my 5D successor, lens hit be damned.
Spoken with true disgust!
It happens to us all, and I reached that point quite a while ago, 1D3 fiasco aside....
Yes, you and the Canon marketing department are equally misinformed about how sensors make images. You would both be right if one were only intent on putting images on computer screens and staring at small parts of them at 100% magnification. If instead one likes to view the whole image at a fixed size then my analysis is the proper way to do comparisons.
I'll try to gracefully bail out of this part of the discussion. Dynamic range is the difference between the smallest accurately measurable signal and the highest measurable signal.
It's, in this case, the difference between the lowest measurable signal above noise floor and the signal at saturation. These numbers are fixed/constant. If you view a whole image or just 10x10 pixels it makes no difference. Each pixel will still saturate after the same amount of photons. It doesn't matter if I look at 1, 100 or 1600x1200 of them.
If you can't grasp this concept about DR, then this particular discussion is aborted.
Lawrence Lee wrote:
I agree with you that his guess seems like a good one. I don't think the "higher end model" will destroy the 1DS MK III because the 1DS camera have a higher resolution sensor. I can't think of any other features the 1DS have at the moment that would justify the $7000+ for it, but that is one thing that I can think of at the moment.
I agree re: the Ds lll. A lot of people with that camera want something that gets them close to medium format but is easier to use and have the lighter, smaller, more versatile benefits. I doubt the guys I know with a Ds lll would take a step back.
To my mind, higher base ISO is a good thing -- it means that the sensor is more efficient at capturing photons, and so has higher S/N ratio at any given ISO.
Correct for the first part, not necessarily for the second part. Making the photonic semiconductor more sensitive to photons means it is also more sensitive to phonons (read: heat) and other sources of signal interference/noise. It's not guaranteed that the actual S/N ratio is better with higher sensitivity.
Tentacle wrote:
I'll try to gracefully bail out of this part of the discussion. Dynamic range is the difference between the smallest accurately measurable signal and the highest measurable signal.
It's, in this case, the difference between the lowest measurable signal above noise floor and the signal at saturation. These numbers are fixed/constant. If you view a whole image or just 10x10 pixels it makes no difference. Each pixel will still saturate after the same amount of photons. It doesn't matter if I look at 1, 100 or 1600x1200 of them.
If you can't grasp this concept about DR, then this particular discussion is aborted....Show more →
OK, I buy that from my old audio days. Now, would there not be such a thing as a "useable" dynamic range? For example, the 40D, according to some measurements/tests (DPReview tests for example), claims a dynamic range greater than the 5D, yet from experience, I know that the 5D offers more latitude to pull up shadow exposure in post processing while maintaining lower noise. And both have similar performance in terms of highlight recovery. Note: I shoot RAW with both.
In other words, the measured range doesn't quite mesh with the actual observation. Perhaps the measuring technique isn't quite right.
Tom_W wrote:
[...] For example, the 40D, according to some measurements/tests (DPReview tests for example), claims a dynamic range greater than the 5D, yet from experience, I know that the 5D offers more latitude to pull up shadow exposure in post processing while maintaining lower noise. And both have similar performance in terms of highlight recovery. Note: I shoot RAW with both.
In other words, the measured range doesn't quite mesh with the actual observation. Perhaps the measuring technique isn't quite right.
The DPReview DR test on the 5D was flawed because it used JPG shots. The JPG still had recoverable image data in it which suppressed the measured DR. Other 5D tests have shown it has a better DR than reported in that DPR review, between 9 an 10 stops.
The problem with the dark side of the DR range is image quality. If you accept 75% accuracy you need to be just 4 times above noise floor. This is half a stop in deviation. Narrow it down to accepting only 0.1 f-stop in noise, which means better image quality, and DR shrinks. The quality of the RAW converter also makes a difference.
Tentacle wrote:
Correct for the first part, not necessarily for the second part. Making the photonic semiconductor more sensitive to photons means it is also more sensitive to phonons (read: heat) and other sources of signal interference/noise. It's not guaranteed that the actual S/N ratio is better with higher sensitivity.
That component of noise (thermal noise) is totally negligible for normal exposure times (less than a few seconds).
