This topic is in the 2009 archive. You can read it and like its posts here. It is read only.
Archive 20095D2 low ISO poor v 1Ds3
p.23 #15D2 low ISO poor v 1Ds3
I'm still to be convinced that at low ISO's the 1DsIII is superior to the 5DII - the test samples seen for low ISO shadow noise look nearly identical so is these some other area where the 1DsIII image quality is better?
Feb 20, 2009 at 04:41 AM
p.23 #25D2 low ISO poor v 1Ds3
This thread is a timely reminder to the many that have wasted much valuable time here (time they will never get back that could have spent taken photos rather than staring at ISO noise) to take a deep breath and few more photos and spend a lot less time worrying about things that, in the end, won't make a rubbish photo better or an amazing photo worse. 
Feb 20, 2009 at 05:00 AM
p.23 #35D2 low ISO poor v 1Ds3
shirozina wrote:
I see you have edited your post to retract the statement that you though the only qualm was the lack of significant AF improvement. Maybe there is some transatlantic misunderstanding but 'overall' means a situation when you take all factors into account and average them out - not that every factor has to be inline. Thus even if the AF was not a significant improvement the rest of the updates are enough to rate the camera as a worthwhile update. People will always whine and whinge when they don't get what they want and it's not ignorant to point that...Show more →
First I'd like to inform you that I did not edit my post and my issue has always been with AF only. I guess you missed what I said. There is no improvement when it comes to outer AF points. NONE. While the center point has improved, though only slightly to my eye, it doesn't do me a bit of good. Others may be content but when looking at the camera as a successor it falls way short in the AF department...there is no arguing that. If you and others are fine with that (stagnancy) then so be it, but it sure as hell doesn't make it right. A realistic expectation is a decent upgrade not a "tweak" specifically when speaking of a camera that took so long to update.
I see you have edited your post to retract the statement that you though the only qualm was the lack of significant AF improvement. Maybe there is some transatlantic misunderstanding but 'overall' means a situation when you take all factors into account and average them out - not that every factor has to be inline. Thus even if the AF was not a significant improvement the rest of the updates are enough to rate the camera as a worthwhile update. People will always whine and whinge when they don't get what they want and it's not ignorant to point that...Show more →
First I'd like to inform you that I did not edit my post and my issue has always been with AF only. I guess you missed what I said. There is no improvement when it comes to outer AF points. NONE. While the center point has improved, though only slightly to my eye, it doesn't do me a bit of good. Others may be content but when looking at the camera as a successor it falls way short in the AF department...there is no arguing that. If you and others are fine with that (stagnancy) then so be it, but it sure as hell doesn't make it right. A realistic expectation is a decent upgrade not a "tweak" specifically when speaking of a camera that took so long to update.
Feb 20, 2009 at 09:49 AM
p.23 #45D2 low ISO poor v 1Ds3
Canon did what they did, no one is forcing anyone to buy one
David
David
Feb 20, 2009 at 09:59 AM
p.23 #55D2 low ISO poor v 1Ds3
Sorry I misread what you said - you said you only qualm was the AF in a subsequent post. AF is not a big deal for me and If it was a priority I would choose a 1 series. I guess if they had upped the AF and only tweaked the sensor up a few megapixels I would have a whine and winge
Feb 20, 2009 at 10:02 AM
p.23 #65D2 low ISO poor v 1Ds3
Is there currently space in the lineup for an 'Eos 3' digital?
That is, a weathersealed 5D2 with 45 point AF?
Should be perhaps, not sure there is, treading on too many toes, spoiling sales of the other cameras in the range. It will probably come though
David
That is, a weathersealed 5D2 with 45 point AF?
Should be perhaps, not sure there is, treading on too many toes, spoiling sales of the other cameras in the range. It will probably come though
David
Feb 20, 2009 at 10:06 AM
p.23 #75D2 low ISO poor v 1Ds3
dhphoto wrote:
Is there currently space in the lineup for an 'Eos 3' digital?
That is, a weathersealed 5D2 with 45 point AF?
Should be perhaps, not sure there is, treading on too many toes, spoiling sales of the other cameras in the range. It will probably come though
David
People have been throwing out that hope for the last eight years, and Canon has been throwing it back. Maybe the D700 will finally make the point, but if it does, we won't see it for at least two more years, and it won't be any sort of "5D."
Is there currently space in the lineup for an 'Eos 3' digital?
That is, a weathersealed 5D2 with 45 point AF?
Should be perhaps, not sure there is, treading on too many toes, spoiling sales of the other cameras in the range. It will probably come though
David
People have been throwing out that hope for the last eight years, and Canon has been throwing it back. Maybe the D700 will finally make the point, but if it does, we won't see it for at least two more years, and it won't be any sort of "5D."
Feb 20, 2009 at 10:13 AM
p.23 #85D2 low ISO poor v 1Ds3
btw, I emailed Canon about this and exchanged a few emails back and forth with someone and he in the end told me to just send the camera in as it 'sounds like an isolated problem.'
I'm pretty sure it was a canned response, so apparently they don't know anything about it.
I do see a jump in chroma noise going from ISO100 to 200, immediately there are red and green noise specs in the same part of the photo in the same setting, only difference being the ISO speed. Very strange, but I have to wonder, what are other cameras like in this same situation besides the 1ds3 and 5d2? What does a 40d look like going from iso100 to 200 at the pixel level?
I'm pretty sure it was a canned response, so apparently they don't know anything about it.
I do see a jump in chroma noise going from ISO100 to 200, immediately there are red and green noise specs in the same part of the photo in the same setting, only difference being the ISO speed. Very strange, but I have to wonder, what are other cameras like in this same situation besides the 1ds3 and 5d2? What does a 40d look like going from iso100 to 200 at the pixel level?
Feb 20, 2009 at 12:46 PM
p.23 #95D2 low ISO poor v 1Ds3
Diglloyd has done tests comparing the D3, D3x, 5d II and 1ds Mark III to show noise while pushing 1-2 stops.
To my eyes, the results are pretty clear, the 5D II show more chroma noise. The Mark III is definately better but still has pattern noise at low ISO. The D3x is absolutely stunning and the D3 appears to be better then both the Mark III and 5D.
To my eyes, the results are pretty clear, the 5D II show more chroma noise. The Mark III is definately better but still has pattern noise at low ISO. The D3x is absolutely stunning and the D3 appears to be better then both the Mark III and 5D.
Feb 22, 2009 at 11:46 PM
p.23 #105D2 low ISO poor v 1Ds3
iso 100, Oly 28 f3.5 wide open:
http://cyberphotographer.com/5d2/oly28f3.5/aostacoffee.jpg
unscaled crop:
http://cyberphotographer.com/5d2/oly28f3.5/aostacoffee_crop.jpg
http://cyberphotographer.com/5d2/oly28f3.5/aostacoffee.jpg
unscaled crop:
http://cyberphotographer.com/5d2/oly28f3.5/aostacoffee_crop.jpg
Mar 06, 2009 at 09:44 PM
p.23 #115D2 low ISO poor v 1Ds3
Wow. Pretty loud.
FWIW, I friggin love the window/snowy river pic
FWIW, I friggin love the window/snowy river pic
Mar 06, 2009 at 10:01 PM
p.23 #125D2 low ISO poor v 1Ds3
I had the same results. It's not so much the noise, but the regular pattern to it, the vertical banding as if you are shooting through fine linen.
I think it might be the highlight tone priority, but I haven't done a controlled shoot to determine. I wrote to Canon about it. They said to send it in, but I'm not doing it till I replicate the circumstances better.
ISO 200 may be a common ISO, as that's the lowest Highlight tone priority will go.
My guess that's a firmware upgrade away from getting fixed, even though that was supposed to be the case with firmware 1.07.
My problem looked exactly like that upper corner.
Part of the reason I like the 5DII is highlight protection, but maybe I'll have to do without a bit.
It can be spotted in the lower left of this photo, the vertical banding:
http://gallery.me.com/ord4/100370/NockamixonStroll2-Edit/web.jpg
http://gallery.me.com/ord4/100370/NockamixonStroll2-Edit-2/web.jpg
Edited on Mar 06, 2009 at 10:18 PM · View previous versions
I think it might be the highlight tone priority, but I haven't done a controlled shoot to determine. I wrote to Canon about it. They said to send it in, but I'm not doing it till I replicate the circumstances better.
ISO 200 may be a common ISO, as that's the lowest Highlight tone priority will go.
My guess that's a firmware upgrade away from getting fixed, even though that was supposed to be the case with firmware 1.07.
My problem looked exactly like that upper corner.
Part of the reason I like the 5DII is highlight protection, but maybe I'll have to do without a bit.
It can be spotted in the lower left of this photo, the vertical banding:
http://gallery.me.com/ord4/100370/NockamixonStroll2-Edit/web.jpg
http://gallery.me.com/ord4/100370/NockamixonStroll2-Edit-2/web.jpg
Edited on Mar 06, 2009 at 10:18 PM · View previous versions
Mar 06, 2009 at 10:05 PM
p.23 #135D2 low ISO poor v 1Ds3
erichard wrote:
I had the same results. It's not so much the noise, but the regular pattern to it, the vertical banding as if you are shooting through fine linen.
I think it might be the highlight tone priority, but I haven't done a controlled shoot to determine. I wrote to Canon about it. They said to send it in, but I'm not doing it till I replicate the circumstances better.
ISO 200 may be a common ISO, as that's the lowest Highlight tone priority will go.
My guess that's a firmware upgrade away from getting fixed, even though that was supposed to be the...Show more →
Beautiful shot! Yes, I can see the vertical pattern at bottom left. My coffee cup picture had HTP switched off, so it's not that. I wish I could go back to firmware 1.0.6 to see if it's any better.
I had the same results. It's not so much the noise, but the regular pattern to it, the vertical banding as if you are shooting through fine linen.
I think it might be the highlight tone priority, but I haven't done a controlled shoot to determine. I wrote to Canon about it. They said to send it in, but I'm not doing it till I replicate the circumstances better.
ISO 200 may be a common ISO, as that's the lowest Highlight tone priority will go.
My guess that's a firmware upgrade away from getting fixed, even though that was supposed to be the...Show more →
Beautiful shot! Yes, I can see the vertical pattern at bottom left. My coffee cup picture had HTP switched off, so it's not that. I wish I could go back to firmware 1.0.6 to see if it's any better.
Mar 06, 2009 at 10:17 PM
p.23 #145D2 low ISO poor v 1Ds3
The base ISO is not very good on the 5Dii I agree, it is because of fewer readouts than the 1dsIII I believe; ie signal noise on the analogue readout before it goes through the A/D convertor. I don't think firmware can fix it.
I hear the Alpha A900 has an amazing base ISO, if that is what one shoots at and isn't tied to particular lenses it would be the ideal tool for the job.
Here is a similar shot to yours, Leica 35/2 summicron at f/3.5. ISO100 & 1/30s (sorry bout motion blur it was handheld). Zeroed on the left, as I would process it on the right.
.<br />
<br />
Also I wouldn't typically take a shot like this, since moving to digital I deliberately try to expose everything in the frame as close to the right as possible. If I needed the foreground and couldn't compose the window out of the frame I would use fill flash and then darken as necessary in post.<br />
<br />
One more point then I'll stop editing this post. Predictable noise like this is non-destructively removable in post, especially with profiled noise calibration in programs like neat image. It doesn't flatten the image much and if you're shooting at ISO100 you likely want the cleanest look possible anyway, or it was really bright and the shadow noise won't be an issue. If you like some grain but not the blotchy kind then you can leave the fine grain and only remove the blotches in the advanced options of neat image, gives my over-pushed images a nice filmy look with the fine grain left behind. The banding kinda sucks though, very hard to remove.</span><!--%startbox%--><!--%fmLike%--><!--%endbox%--><!--%divend%--></div>
</div>
<div class=)
Mar 06, 2009 at 11:26 PM
I hear the Alpha A900 has an amazing base ISO, if that is what one shoots at and isn't tied to particular lenses it would be the ideal tool for the job.
Here is a similar shot to yours, Leica 35/2 summicron at f/3.5. ISO100 & 1/30s (sorry bout motion blur it was handheld). Zeroed on the left, as I would process it on the right.
Mar 06, 2009 at 11:26 PMp.23 #155D2 low ISO poor v 1Ds3
5DII, ISO200, f/11, .8 sec, EF70-200mm at 200mm.
Full original image - no PP except for a bit of shadow recovery by lowering the blacks setting in ACR to 2 since shadows were slightly underexposed:
http://www.gdanmitchell.com/images/NoiseDemo/NoiseFull20090228.jpg
Section at 100% magnification (actual pixels) with no PP except as above:
http://www.gdanmitchell.com/images/NoiseDemo/NoiseCrop20090228.jpg
Seems pretty decent to me. Maybe some cameras are affected by a problem that doesn't affect all?
Dan
(And, yes, this isn't a great photo - I was just looking for the first thing I could find that went well into the dark tones.)
Full original image - no PP except for a bit of shadow recovery by lowering the blacks setting in ACR to 2 since shadows were slightly underexposed:
http://www.gdanmitchell.com/images/NoiseDemo/NoiseFull20090228.jpg
Section at 100% magnification (actual pixels) with no PP except as above:
http://www.gdanmitchell.com/images/NoiseDemo/NoiseCrop20090228.jpg
Seems pretty decent to me. Maybe some cameras are affected by a problem that doesn't affect all?
Dan
(And, yes, this isn't a great photo - I was just looking for the first thing I could find that went well into the dark tones.)
Mar 07, 2009 at 01:02 AM
p.23 #165D2 low ISO poor v 1Ds3
danmitchell wrote:
Seems pretty decent to me. Maybe some cameras are affected by a problem that doesn't affect all?
Looks good to me as well... It could be the different lighting/exposure or it could be that some cameras show worse low iso behaviour than others. Or both...
Based on my own experiences I think some cameras are worse than others.
BTW What FW do you have installed?
Seems pretty decent to me. Maybe some cameras are affected by a problem that doesn't affect all?
Looks good to me as well... It could be the different lighting/exposure or it could be that some cameras show worse low iso behaviour than others. Or both...
Based on my own experiences I think some cameras are worse than others.
BTW What FW do you have installed?
Mar 07, 2009 at 01:21 AM
p.23 #175D2 low ISO poor v 1Ds3
That pattern noise is also found on the 5D and for that matter the 1DsII when you push the shadows. Unless we see some more side by side pics of different cameras it's not very informative IMO. My guess is that the highligh priority option just pushes the shadows and compresses the highlight data to give a smoother roll off without obvious clipping. The woodland shot doesn't have the huge contrast range that the other 2 have which are heavily backlit so it's not suprising there is no noise.
Mar 07, 2009 at 05:19 AM
p.23 #185D2 low ISO poor v 1Ds3
Sirozina, I'm afraid you are wrong about the contrast levels in the "woodland shot." I had to lift the levels of the darkest shadows in post for this shot because increasing the exposure in camera would have blown the bright greens. In fact the original shot exceeds the dynamic range of the histogram display.
(I made three bracketed exposures of this scene for that reason, and would have combined them had this shot turned out to be worth it. For the record, this example is a single unsharpened RAW conversion as described in my earlier post.)
I'm very familiar with the need to increase "brightness" in shadow areas. I just printed 3 photographs from this particular shooting session, and all three contained extreme dynamic range - going from deep shadow (such as the black areas you can find in this image) to areas lit by full sun.
If you push shadows you will increase noise - that is just the way digital capture works. As I wrote once before in this long thread, there are ways to deal with this, some very old and some new to digital.
When I saw brainiac's most recent post with the really terrible chroma noise on the interior of the cup, what I saw was completely out of sync with what I'm getting from my 5DII. (And I'm a careful and knowledgeable shooter who prints.) I can only think of a few explanations:
1. There is something very different about how his camera works in comparison to mine.
2. The cup was in much deeper shadow than it appears in the final photograph, and the noise is a by-product of significant pushing of the shadows. (I might have used a tripod and exposure bracketed or thought about some artificial lighting.)
Take care,
Dan
(I made three bracketed exposures of this scene for that reason, and would have combined them had this shot turned out to be worth it. For the record, this example is a single unsharpened RAW conversion as described in my earlier post.)
I'm very familiar with the need to increase "brightness" in shadow areas. I just printed 3 photographs from this particular shooting session, and all three contained extreme dynamic range - going from deep shadow (such as the black areas you can find in this image) to areas lit by full sun.
If you push shadows you will increase noise - that is just the way digital capture works. As I wrote once before in this long thread, there are ways to deal with this, some very old and some new to digital.
When I saw brainiac's most recent post with the really terrible chroma noise on the interior of the cup, what I saw was completely out of sync with what I'm getting from my 5DII. (And I'm a careful and knowledgeable shooter who prints.) I can only think of a few explanations:
1. There is something very different about how his camera works in comparison to mine.
2. The cup was in much deeper shadow than it appears in the final photograph, and the noise is a by-product of significant pushing of the shadows. (I might have used a tripod and exposure bracketed or thought about some artificial lighting.)
Take care,
Dan
Mar 07, 2009 at 09:10 AM
p.23 #195D2 low ISO poor v 1Ds3
http://jlb.smugmug.com/photos/485898166_zw3T3-M.jpg
http://jlb.smugmug.com/photos/485898846_uDqqm-M.jpg
these pictures were shot in near darkness in a jazzclub in nyc...ISO 3200 with no noise reduction other than that which is applied by ACR automatically...in all seriousness,how much cleaner can we get?...this is of course the 5d mkII
http://jlb.smugmug.com/photos/485898846_uDqqm-M.jpg
these pictures were shot in near darkness in a jazzclub in nyc...ISO 3200 with no noise reduction other than that which is applied by ACR automatically...in all seriousness,how much cleaner can we get?...this is of course the 5d mkII
Mar 07, 2009 at 12:10 PM
p.23 #205D2 low ISO poor v 1Ds3
trumpet_guy wrote:
That ISO 100 performance is not thrilling in the shadows.
Would you post a crop of the shadows that has been downsized
the appropriate percentage to match the pixel density of the
original 5D? [Sharpen minimally after downsizing.] Pretty please.
http://cyberphotographer.com/5d2/aostacoffee_13Mpcrop.jpg
Dan is right: the coffee cup has had considerable pushing done to the shadows in order to cope with the wide dynamic range of the scene. However, my complaint is that when I used to do this with my 1Ds3, I did not get the criss cross pattern, but a more random kind of noise that was much more fixable and less intrusive.
That ISO 100 performance is not thrilling in the shadows.
Would you post a crop of the shadows that has been downsized
the appropriate percentage to match the pixel density of the
original 5D? [Sharpen minimally after downsizing.] Pretty please.
http://cyberphotographer.com/5d2/aostacoffee_13Mpcrop.jpg
Dan is right: the coffee cup has had considerable pushing done to the shadows in order to cope with the wide dynamic range of the scene. However, my complaint is that when I used to do this with my 1Ds3, I did not get the criss cross pattern, but a more random kind of noise that was much more fixable and less intrusive.
Mar 07, 2009 at 12:17 PM