alanwarp wrote:
Yes, as Thomas Knoll (from Adobe) has stated:
"However, it appears the 50D review was done using 4.6 RC, which had incomplete 50D support. Results would have likely have been much better had 4.6 final been used."
"Would have been likely" quoted from one of the makers of... Convincing? Maybe... But the conversions of 50D RAW I did with the beta/RC and official releases prove me otherwise. They look the same to me... Did you take a look at the comparison, or did you try it out yourself?
Daan B wrote:
"Would have been likely" quoted from one of the makers of... Convincing? Maybe... But the conversions of 50D RAW I did with the beta/RC and official releases prove me otherwise. They look the same to me... Did you take a look at the comparison, or did you try it out yourself?
I found better noise performance from ACR 4.6 final vs ACR 4.6 RC in my experience, but the usual non-scientific testing caveats apply. My main point here is not that this might save the 50D from it's lame review, or that the noise performance is much better in 4.6 final, but that why the heck are they reviewing a camera with non-final release software
They have already found some differences, Phil says:
"Richard and Lars are now working to replace all of the ACR images in the review with re-converted images using ACR 4.6 'Final'. Our observations so far are that there is less moire in the resolution chart conversion and that the pinking of negatively adjusted images has been resolved. Otherwise the results are virtually identical (but as I noted above we will still replace all of them)."
So OK, otherwise virtually identical, but he's saying they aren't identical? I'm just annoyed that they would pull an amateur-esque reviewing move, they are a business owned by amazon.com, not some guy working out of his basement.
alanwarp wrote:
I found better noise performance from ACR 4.6 final vs ACR 4.6 RC in my tests, but the usual non-scientific testing caveats apply. My main point here is not that this might save the 50D from it's lame review, or that the noise performance is much better in 4.6 final, but that why the heck are they reviewing a camera with non-final release software
They have already found some differences, Phil says:
"Richard and Lars are now working to replace all of the ACR images in the review with re-converted images using ACR 4.6 'Final'. Our observations so far are that there is less moire in the resolution chart conversion and that the pinking of negatively adjusted images has been resolved. Otherwise the results are virtually identical (but as I noted above we will still replace all of them)."
So OK, otherwise virtually identical, but he's saying they aren't identical? I'm just annoyed that they would pull an amateur-esque reviewing move, they are a business owned by amazon.com, not some guy working out of his basement.
I agree... it is rather amateuresque to use the beta/RC version of ACR instead of the official release. We will see if they have to correct some of their findings... My point was merely that I didn't see any difference between ACR4.6 beta/RC and the final release regarding noise myself. YMMV
Daan B wrote:
I agree... it is rather amateuresque to use the beta/RC version of ACR instead of the official release. We will see if they have to correct some of their findings... My point was merely that I didn't see any difference between ACR4.6 beta/RC and the final release regarding noise myself. YMMV
Indeed..Didn't look at the noise that much, I was actually more impressed with the ACR 4.6 final when used with the latest Adobe Beta DNG profiles, I think Adobe Standard Beta 2 was looking good.
I really question the review and reviewing process. Starting weeks ago with the miserable samples for the 50D posted in the DPreview samples area. They were a poor representation of what the camera can do.
Also, I'm not convinced that Adobe has nailed their RAW support. Even in the 5.1 version of ACR that I'm using.
DPP seems to produce results that are qualitatively different than ACR. Should the camera be tested against DPP ... which is better? Which should be used in the review?
Finally, I am really getting tired of Canon's incompetent, negligent communication of camera and lens performance data. Certainly, Canon could provide data on DR, noise, RAW headroom and other key specs that would help people make better decisions. Why should we be dependent on DPreview for accurate data on the DR performance of a Canon DSLR?
I hope someone at Canon takes note of the DPreview review. If it is accurate, than fine. But, if it is inaccurate than perhaps Canon should consider offering DNG as a standard, optimized output format for their bodies. As users, we deserve to have the ability to quickly integrate new bodies into our Adobe based workflows without this wasted time and effort.
I think the shortcoming of the camera wasn't the noise, but rather the resolution improvement over the 40d. I've used DPP to convert and really, the cameras look very, very similar. Not enough to justify paying a premium for one now.
I played with a 50D and a 5D in comparison with the 40D I own, and the 50D, on "normal shots", meaning not ultra-high ISO, not wide-open lenses, made better pics than my 40D, to the extent that I could recognise which was which any time on A4 prints from RAW, DPP and no PP. There is/was a positive difference, but is it enough to get me to upgrade? I am not sure. In any instance, 5D was clearly better yet, so I put in a pre-order for 5D2.
ACR is not really handling 50D files well at all just like it took 6 months for them to get the 40D right. DPR even used a beta of the ACR software. If you want to see what the camera can do you use DPP and then make your decision. I am not defending the 50D but in my tests the statements made by DPR are pure bunk. "Just" barely highly recommended and when you look at the flawed images against say the Pentax k20 they are much better. That camera is highly recommended by DPR with no reservations. I don't sell this stuff but I think it is only fair to give each and ever camera an impartial review and I think this one stinks. How can they not have even quality control in their review to not even know what raw software they are using. We are not testing the Mars probe here.
These tests showing increased noise with the 50D (in comparison to 40D) definitely bum me out. I was waiting for the 50D as an upgrade for my 20D which I love. Also, the limited dynamic range hasn't been discussed much here, but it's also an issue for me as well.
toonhorse: I have the 30D, (same sensor as your 20D), and I can tell you the 50D has much more detail than the 30D. At ISO 200, the pix are razor sharp (with good L glass). I have not tested the higher ISO's though, yet.
Well if your willing to believe DPR then by all means don't buy the 50D. I have found it to be a good improvement over the 40D and my lenses are not having any problems with the extra MP.
In fact that was the most attractive aspect of the 50D for me because I find myself cropping quite often. Now I can crop 5MP from the 50D file and Have a 10MP file vs a 5MP when using the 40D. I use it exclusively for wildlife and birds where my 500 or 600 sometimes is just not enough reach so cropping the extra MP really helps.
I plan on getting the 5DMKII for everything else so I can live with some of the 50D's shortcomings.
as to the 50D's premium over the 40D, you're paying extra for the ability to crop more (very useful for wildlife shooters), print big (again, people like wildlife photogs who sell large prints) the micro af adjust (very useful for just about anyone), and to a lesser extent the vga screen (flashy, i suppose).
either you find the 50D's features useful, or you don't. i personally could do with a 40D retro-fitted with micro-af adjust, and that's about it. (actually, i could do with a 40MP medium format system, but that's not in the cards at the moment.)
the 50D will serve many scores of canonists well. ymmv.
davenfl wrote:
ACR is not really handling 50D files well at all just like it took 6 months for them to get the 40D right. DPR even used a beta of the ACR software. If you want to see what the camera can do you use DPP and then make your decision. I am not defending the 50D but in my tests the statements made by DPR are pure bunk. "Just" barely highly recommended and when you look at the flawed images against say the Pentax k20 they are much better. That camera is highly recommended by DPR with no reservations. I don't sell this stuff but I think it is only fair to give each and ever camera an impartial review and I think this one stinks. How can they not have even quality control in their review to not even know what raw software they are using. We are not testing the Mars probe here. ...Show more →
ACR 4.6 vs ACR 5.1 no beta. An ISO 1600 100% crop with Chroma NR, luminance NR set to zero and sharpening set to 0. ACR 5.1 on the left, 4.6 on the right. Little difference. How do we know what dpp may be doing? ACR or Capture 1 is my workflow, not dpp.
I expect a "native" ISO of 1600 might actually fare better. Don't have the camera, was evaluating it as a backup - someone was nice enough to post up RAW.
alanwarp wrote:
Is this ACR 4.6 RC or ACR 4.6 final, because ACR 4.6 final and ACR 5.1 are the same. Same algorithm
Alan
Final. And for both I used the ACR 4.6 camera profile. Just bought PS CS4 to see if the 50D was better in ACR 5.1 (well there were other advantages ) in order to help make a decision Didn't know that, but it was as inevitable as taxes.
I notice my neutral RAW files look cleaner in CS4 ACR 5.1 than DPP. Could be because ACR applies a default tad sharpening.
But I like DPP's picture styles so...... I have use for both.
therock wrote:
I notice my neutral RAW files look cleaner in CS4 ACR 5.1 than DPP. Could be because ACR applies a default tad sharpening.
But I like DPP's picture styles so...... I have use for both.
Or dpp applies a default tad NR
Certainly the neutral applies a more moderate curve, which should result in less noise amplification. But you have to compare apples to apples (or tonal curves to tonal curves) I guess.
Daan B wrote:
I agree... it is rather amateuresque to use the beta/RC version of ACR instead of the official release. We will see if they have to correct some of their findings... My point was merely that I didn't see any difference between ACR4.6 beta/RC and the final release regarding noise myself. YMMV
Right, PS CS 3 and LR 2.1 use both ACR 4.6... ACR 5.1 is limited to PS CS 4. But I think both 4.6 and 5.1 use the same algorythms for the 50D, so it shouldn't matter anyway.
) in order to help make a decision