Picture quality 30d vs 50d

[Don Madej]

I will be going to Bosque del Apache late in november and I am considering the purchase of a canon 50d. I currently own a 30d. Should I expect a noticeable difference in picture quality? I know that there are other beneficial improvements,but I am mainly concerned with picture quality. As you all know, money is a little tight these days.

Thanks in advance.

[Joerg Rockenberger]

I don't have experience with the 50D but I recently had to revert back to my 30D while my 40D was at Canon for a Error 99 repair and I noticed a significant difference in quality. Both with regard to the AF performance but also with regard to the exposure. It seemed that the 40D got more accurate colors and also more dynamic range.

So, unless my 30D is sub-par and/or your 30D is stellar I'd assume that you can expect a significant improvement by upgrading to the 50D. Only two ways to find out: rent or buy one.

Good luck. JR

[Doug Brown]

The 50D is a major step up from the 30D IMO in just about every way you can imagine. I have a good friend who went down to Bosque last weekend and was amazed at how well it tracked the cranes in low early morning light. High ISOs are much more usable too. The extra megapixels come in handy. And I think it does a better job of metering scenes.

[Krijn Trimbos]

What about compared to the 40D....???

I just read the review of dpreview which compared noise reduction between the 40D and 50D and the 50D came up short cause of the increased MPs

http://www.dpreview.com/reviews/canoneos50d/

[Doug Brown]

Read the review. Don't agree with the findings; also don't agree with Canon's claim of 1.5 stops of improved high ISO noise. 50D is better than the 40D, at least for bird photography IMO, in particular because of the improved AF performance.

[Doug Brown]

Check out some of the updated info over on DPR: http://forums.dpreview.com/forums/re...ssage=29857356

[Peter Hawrylyshyn]

Krijn -
There's another detailed technical comparison of the 40D vs 50D at:
http://astrosurf.com/buil/50d/test.htm
Their findings and conclusions are somewhat different to dpreview

The last page at ( http://theory.uchicago.edu/~ejm/pix/...300_40D_tests/ ) has detailed "technical" comparison of 50D to several other Canon and Nikon cameras. I believe the author is a BPN member
http://theory.uchicago.edu/~ejm/pix/...300_40D_tests/

Putting all the technical spec's aside - best test is try the camera for yourself against your 30D or a 40D at your local camera shop, then compare images in PS. I sold my 30D backup camera to get a 50D - have not been disappointed.

[Roger Clark]

What about compared to the 40D....???

I just read the review of dpreview which compared noise reduction between the 40D and 50D and the 50D came up short cause of the increased MPs

http://www.dpreview.com/reviews/canoneos50d/

The dpreview noise statistics are dependent on the software used which can be different between different cameras. There are methods to get the real sensor data which some people use. Results are summarized at:

Digital Sensor Performance Summary
http://www.clarkvision.com/imagedeta...rmance.summary

Two main factors dictate sensor performance: 1) how many photons a pixel collects, called the full well capacity, and 2) the noise floor, called the sensor read noise, both expressed in electrons. We now have data at ISO 400 for the 50D, courtesy Christian Buil at: http://astrosurf.com/buil/50d/test.htm

Some full well capacities at ISO 400 and read noise values (in electrons):

Camera _ Full Well _ read Noise _ megapixels _ Unity Gain ISO _ Dynamic Range
__________________________________________________ ___(Full Well / read Noise)

30D _____ 12800 ______ 7.5 ______ 8.2 _______ 1200 ________ 10.7 stops
40D _____ 10800 ______ 6.6 _____ 10.1 _______ 1300 ________ 10.7
50D ______ 8400 ______ 5 _______15.0 _______ 880 _________ 10.7

(I tried to make this table somewhat readable; spaces get compressed so it is hard to line up the columns.)

The Unity gain ISO is a measure of high ISO performance. While the gain is less on the 50D, that is mitigated by some degree by the greater pixel count. Higher unity gain ISO is better as is higher Full Well. Lower read noise is better. So there are pluses and minuses in the trend.
But the uniformity of the fixed pattern noise in the Christian Buil tests show the 40D has a much smoother background. Based on that, if I was going for a high ISO performance alone, I would probably choose the 40D. But for other applications, the higher megapixel count of the 50D looks very nice.

Basically, you don't get something for nothing, but sometimes a trade is good.

Roger

[kevinmat]

Doug is absolutely right, the 50D is a far better camera than the 30D/40D - you really need to use it to see the difference. I could never go back - it's that big of a difference - yet on paper it's difficult to see how so. I was always annoyed at the metering differences between my 40D and 5D, the 40D never seemed to get it right most of the time requiring ex comp set as standard. The 50D is usually spot on. The AF is definitely improved as well (though this may be due to the Digic IV chip).

The Dpreview review was not done by Phil - and it shows - and also looks as though it was rushed. I would not be surprised if they at least adjust their findings in the future. Most of the negatives are due to the use of ACR 4.6 which does not do well with the new 50D. Many RAW shooters are saying that for the first time that they are getting significantly better results from jpeg than RAW out of the camera. I'm not so sure - but I do undertsand what they are saying.

One caveat though - looking at 50D files at 100% takes some getting used to over previous XXD models - not sure I'm used to that still.

[Chris Starbuck]

Most of the negatives are due to the use of ACR 4.6 which does not do well with the new 50D. Many RAW shooters are saying that for the first time that they are getting significantly better results from jpeg than RAW out of the camera. I'm not so sure - but I do undertsand what they are saying.

I've only had my 50D a few days, and haven't had a chance to shoot anything real world outdoors; have mostly just been looking at noise vs ISO in indoor test shots. One thing I found almost immediately is that ACR (5.1 in my case) doesn't know about or use the new in-camera high-ISO noise reduction settings. That's probably why people think their in-camera jpegs (which have those settings applied) are better than the corresponding raw files (which don't). The feature certainly will make the in-camera jpegs more usable at high ISO, but IMO not enough to outweigh the advantages of raw.

Up to now I haven't used a 3rd party NR plugin, because I was always pretty satisfied with ACR's ability to reduce noise in my 20D images, at least up to ISO 800. But I don't think it handles noise as well in my brief tests of my 50D. IMO it's probably OK at ISO 400, but above that it's just not satisfactory, and I'm in the process of deciding on which NR plugin to buy.

[Rene A]

The 50D is good up to Iso 1600..& it's loves good lenses

500 & 1.4x is soft but without 1.4 great same for 300 f2.8

It also seems that firmware 1.0.3 has improve the noise

& lens lockup m& errors..

I have not had any more so far

Renate

[Jack Faller]

I have had a 50 D for a few days. As the others have mentioned, you should find the 50D a big improvement over the 30D. The improvements of the brightness of the image in the viewfinder and the info visible in it are impressive, but the improved autofocus for tracking of birds in flight is especially useful. The improvements in the ISO noise are not qualitatively obvious to me relative to the 40D. I still get a fair amount of noise at ISO 800, but I am using ACR 4.6. I hope this is an artifact of using 4.6. Generally I turn down both noise reduction sliders in ACR and use Dfine 2 for noise reduction. You can download Dfine2 and try it out for free.(http://www.niksoftware.com/dfine)

[Roger Clark]

500 & 1.4x is soft but without 1.4 great same for 300 f2.8

The softness might be expected due to diffraction. The modulation transfer function (MTF), which is a measure of contrast versus spatial detail is at about 40% at f/4 at the spacing of the 50D pixels (4.7 microns). At f/8, it's 2 pixels for 40% MTF and at the pixel level 0% contrast. So a perfect lens at f/8 would definitely appear soft at the pixel level with the 50D pixel spacing. See:
http://www.clarkvision.com/imagedeta...ml#Diffraction

Roger

[arash_hazeghi]

Depends on your shooting style and expectations, image quality is somewhat subjective after all. In pure technical terms 50D has more pixels but pixels are smaller by a factor of ~2 which means lower S/N ratio at least in terms of shot noise, also more non-uniform dark current results in more visible noise in uniformly colored areas such as a smooth BG. In many cases 50D shots will look softer than those of 20/30D because of diffraction effects and also the fact that Canon uses a gapless oxide layer (microlens) with more abberations. In good light high contrast scene with the best lenses you might salvage more detail and get better prints but at higher ISOs you may not. I recommend you rent the camera for a few days before buying and see if you like it, I did the same and end up returning it but that's me, I only shoot in RAW mode and I don't like NR, you might actually like it, especially now that the price has dropped to $1200 for body.

Best

[Arthur Morris]

I am never one for tests or for the test results of others and I do not understand any of the technical terms above, but after using the 50D for a week, often with the 1.4 X TC and the 500, I can say that I have never seen images that jump off of the monitor for both sharpness and color. Yes, sharpness. Yes, with the 500 and the 1.4X. So now you can all see--with no disrespect intended to those who know lots more than I about technical stuff, why I do not do tests or listen to experts. I simply create images...

The Whimbrel image here was created with the 500, the 1.4X TC, and the 50D...

[Sabyasachi Patra]

I am never one for tests or for the test results of others and I do not understand any of the technical terms above, but after using the 50D for a week, often with the 1.4 X TC and the 500, I can say that I have never seen images that jump off of the monitor for both sharpness and color. Yes, sharpness. Yes, with the 500 and the 1.4X. So now you can all see--with no disrespect intended to those who know lots more than I about technical stuff, why I do not do tests or listen to experts. I simply create images...

This reminds me of the bumblebee example.

According to the laws of physics, a bumblebee cannot fly as its wings are not large enough, it is not aerodynamically built, etc etc. The problem is that nobody told the bumblebee about the laws of physics. So what does the bumblebee do?... it just goes and flies.

Cheers,
Sabyasachi

[Roger Clark]

:):):) (With all the smileys. I'm poking fun here.)

I am never one for tests or for the test results of others .....
.....but after using the 50D for a week, often with the 1.4 X TC and the 500, I can say that I have never seen images that jump off of the monitor for both sharpness and color. Yes, sharpness. Yes, with the 500 and the 1.4X. So now you can all see--with no disrespect intended to those who know lots more than I about technical stuff, why I do not do tests or listen to experts. I simply create images...

With all due respect, those images are your tests. Not really any different than others taking images, just some technical people might make quantitative measurements compared to your visual inspection. Both methods can produce interesting and valid results.:)

This reminds me of the bumblebee example.
According to the laws of physics, a bumblebee cannot fly as its wings are not large enough, it is not aerodynamically built, etc etc. The problem is that nobody told the bumblebee about the laws of physics. So what does the bumblebee do?... it just goes and flies.

I have heard this one too. But with all due respect, the person who came up with that obviously didn't know physics! :) The bumblebee does not violate the laws of physics.

Same with cameras. We often hear of some fantastic new claim (like the Nikon D3 high ISO performance), but when the performance was analyzed, it fell closely where physics said it should. Nothing magic, just a well-engineered design. The Canon 50D is also falling just where physics says it should regarding noise performance.

Sabyasachi, on another topic, Chandrayaan-1 is nicely in lunar orbit and doing well (congratulations to the people of India) . Our imaging system should be turned on soon. I'm still not sure when I'll be coming to India, but probably the first half of next year,

Roger

[Sabyasachi Patra]

:
Sabyasachi, on another topic, Chandrayaan-1 is nicely in lunar orbit and doing well (congratulations to the people of India) . Our imaging system should be turned on soon. I'm still not sure when I'll be coming to India, but probably the first half of next year,

Roger

Roger,

It is a terrific achievement and such collaborations among countries for space research (and I believe all research) is the way to go. On a lighter note, I hope one of the scientists don't enter these satellite images in the BBC Wildlife Photographer of the Year Competition. :D:D:D
Let me know when you are in India.

On the issue of Expert opinion and analytical findings, let me tell you a story. When I was doing my Management education more than a decade back, we were taught to go by rigorous analysis of findings through various statistical models till we gain sufficient knowledge. We were told that when you become an expert, you can use your knowledge and come with a solution without resorting to rigorous analysis. An expert musician can tell you whether an instrument is good or not with one glance. I had read somewhere about expert opinion of Japanese camera designers being given more importance by Nikon and Canon, don't know where.

To my eyes, the 50D shots by Doug and Artie look great and I would be happy if I can't extract the full juice out of this camera. I believe these days the cameras are so good that they are much better than the average photographer. However, I would have loved the high ISO performance as Canon had claimed, as I only use the NR feature in Lightroom.

Cheers,
Sabyasachi

PS: By the way, I appreciate your efforts in analysing the sensor capability of these cameras, and have sent the links to like minded people.