40D vs 50D, ISO and detail comparison - test results

As a consequence of a discussion on another forum I wanted to finally answer questions, for myself, regarding the performance of the 40D vs the 50D and shot a series of test images for comparison. Having gone to that trouble I thought I might as well share the results here. The following is a direct copy of my post to that other forum (Here's the whole thread from that forum- http://www.birdforum.n​et/showthread.php?t=13​5541)....

I have shot some test images under controlled conditions (Darkened room with 300W of continuous tungsten lighting, tripod, MLU, timer, 10X Live View to focus) using my 40D and 50D at full stop ISOs from 100 to 1600 (and 3200 for the 50D). I used my 85/1.8 prime at f/5.6 and 3m from the subject, to give decent sharpness and a reasonable DOF (+/-13cm) to cover the subject/scene at normal magnification levels. The 1/4 frame crops at 800x533 display at 6"x4" on my monitor, which equates to a full image size of 12"x8", which is just a little over the magnification for the calculated DOF, but close enough. In other words, the 1/4 frame crops should look acceptably sharp throughout the main subjects within the image, assuming focus was accurate.

They were shot raw and processed in DPP using default settings. Each image has been processed in four sizes....

Full image, resized to 800x533;
1/4 frame, resized to 800x533;
1/16 frame, resized to 800x533;
100% crop at 800x533.

Here is the album....

http://picasaweb.googl​e.co.uk/EezyTiger/40DV​s50D?authkey=Gv1sRgCP6​r0Zbe4bPNCg#

If you play through the images as a slideshow you will be hard pushed to notice the differences from one ISO to the next, right up until 1600 ISO, either in terms of noise or detail. 3200 ISO is a bit obvious, viewed at 100%, but a long way beyond redemption with a bit of NR work. EXIF data is in the images and you can view the details for each image by clicking "more info" over on the right of the gallery when viewing each individual picture.

My own opinion is that....

- As far as noise is concerned it's a wash, and just not worth arguing which camera is the better performer.

- For sharpness and detail the 50D is the definite winner, once you start creating larger final image sizes. This is slightly evident on the 1/4 frame crops (equivalent to a 12"x8" uncropped image on my monitor), and more so as the crops get tighter and the enlargements become larger.

- At the end of the day, unless you are cropping or enlarging aggressively it is probably not worth trying to differentiate the two cameras based on image quality, up to 1600 ISO. If you do crop hard and have sharp glass then the 50D will deliver the better results.

As for other features on the 50D, if you shoot to JPEG you may well benefit from new image enhancement options such as ALO and LPIC, plus the finer levels of control over NR. There isn't anything to directly benefit raw shooters specifically, but most everyone may well benefit from AF microadjustment and the higher resolution screen together with a higher resolution preview image. This does give real advantages when checking focus in Live View, and when reviewing images after they've been taken. If you have a static subject, such as a perched bird, then Contrast Detect AF in Live View can yield the absolute sharpest image possible. I have never been able to improve on the results of contrast detect AF when focusing manually at 10X mag (assuming adequate contrast on the subject).

In a nutshell, the 50D is all about delivering the sharpest images possible, with maximum detail. In that respect it beats the 40D, plain and simple. Whether or not you can use that to your advantage is another matter. If you only create small prints or 800x533 web shots, or smaller, then the 40D will serve you perfectly well. If you print big, crop tight, or supply images to stock agencies at maximum resolution, or perhaps larger than that, the 50D will put you ahead of the game.

EDIT : Just so you know what to expect in the album, here's the full scene at 1600 ISO with the 50D....

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

Thanks for throwing this out there. Interesting. nicely done.

I would add that i still do not see photog asking others what camera you took the shot with..... instead: question which lens/camera settings (ISO/f/shutter speed)/ PP technique.

Later,
_Mark

Ziffle wrote in post #7568939
Thanks for throwing this out there. Interesting. nicely done.

I would add that i still do not see photog asking others what camera you took the shot with..... instead: question which lens/camera settings (ISO/f/shutter speed)/ PP technique.

Later,
_Mark

I agree, between the 40D and the 50D there is less to concern one regarding the cameras than over the choice of lens or the photographer's own skills, but many still ask whether they should get the 40D or the 50D, and many still concern themselves with pixel level noise, when there are more important things to consider, and some refuse to believe that the 50D is an improvement on the 40D. Others may consider the 50D to be better than the 40D but not worth the price premium.

The thing is, I wonder how many of these opinionated people have actually used both cameras in anger and squeezed the last drop of performance from either one? I'm not saying I've pushed them to the limit, but wedding photography is demanding enough in terms of light levels and the 50D can deliver very worthwhile results at 3200 ISO. It's time the "per pixel" noise debate got laid to rest. The important thing is - how do the photos look, not the wretched pixels.

Between the 40D and 50D the big questions to be answered are....

- Did the extra resolution come at a price of too much noise? to which the answer is that I think not.
- Does the extra pixel density offer tangible benefits in terms of image detail? to which the answer is that I think it does, and the 50D is equiped with features to make those megapixels count, like AF microadjustment and a higher resolution screen, higher resolution preview image and an additional AF system.

Here's a shot taken today with my 50D, 100-400 zoom and a Kenko 1.4X teleconverter. This was shot at 560mm, f/8 (wide open), 1/640 and 400 ISO. This is a 1/4 frame crop resized to 50%. Notice the thin DOF leaves no margin for error with focus. This was focused using Live View Contrast Detect AF. I suspect with a 400mm (or longer) prime the image might have been sharper, but I think it turned out OK.

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

Hello;
My name is Bob and I found your evaluation interesting; however, I think the more critical test would be to photograph with your sharpest lens over the range of aperture settings at ISO 100 for both cameras. I think that you will find that the 15.1MP 50D is seriously diffraction limited to f5.6 or f8.0 at best if you want to obtain a larger size print like for example a 20"x30".

Hi, Bob, and welcome to the forum

Viewed at the pixel level it is true that the effects of diffraction may appear visible at sooner on the 50D than the 40D when stopping down, but the relevance of diffraction is not considered important at the pixel level, but at the image level and based on how much you will enlarge the sensor image to reach your final reproduction size. Since the 40D and 50D have the same size sensor the diffraction calculations, in terms of circles of confusion, are no different for either one.

I guess the resolution advantage may begin to disappear as diffraction becomes significant, but the 50D image will never be worse through diffraction than the equivalent 40D image. It is generally accepted that lenses are usually sharpest when closed down a couple of stops from wide open, but usually between f/5.6 to f/8 is the sweet spot. I'm quite sure that at apertures of f/8 and wider the 50D will still maintain a lead over the 40D for useable resolution. Beyond f/8, I'm not sure most people would care very much. I almost never stop down below f/8. Mostly I am shooting at f/4 or faster, where my lens allows. The only exception is my tortoise like 100-400 zoom. There may be rare occasions when needing to stop down further, like when using fill flash on a bright, sunny day, but it is rare to need to stop down beyond f/8 for DOF reasons on a crop body. It's mostly only to achieve a specific slowish shutter speed. An ND filter or CPL is another option if stopping down is a concern.

FWIW I shot a series of images with my 40D, quite some time ago, to demonstrate diffraction softening and the results are in an album here....

http://picasaweb.googl​e.co.uk …key=Gv1sRgCMum1​sn3v9iWzgE

In those example diffraction is really not an issue until you get to f/16, and then only when viewing at 100% and being very picky. I don't actually print my images but I suspect that ink smearing and degraded lens sharpness would be as much of a concern as diffraction softening. I think all in all the diffraction thing is a bit of a non issue as far as comparing the 40D and 50D is concerned. However, I'm always one to explore and confirm my understanding so maybe I will perform the test along the lines you suggest. Don't hold your breath though

This article was posted in another thread but relates directly to some of the issues brought up so bears repeating. I won't bother trying to condense the information but leave it to you to read at your leisure (it is short).

http://www.bobatkins.c​om …resolution_conf​usion.html

Thanks, Duncan. I had seen that article a while back. It seems that Bob and I concur in our views. I got some spare time this evening so I have actually run off the diffraction test shots, but only for the 50D. I'm really not interested in repeating the tests for the 40D. I know they will appear equal at regular sizes and favour the 40D when pixel peeping at 100%. So long as the 50D delivers satisfactory results that is more than good enough.

I have uploaded the test results from my 85/1.8 at all apertures from f/1.8 to f/22. I have included 100% crops, 50% crops and the full image, resized to 800x533. The album is here (EXIF is available by clicking "more info", over on the right when viewing each picture)....

http://picasaweb.googl​e.co.uk …hkey=Gv1sRgCImD​h_jvquTOJw

What I find interesting, but unsurprising, is the trend in file sizes as one progresses through the various f/stops, increasing in size at f/8 +/- 1 stop and falling off at the extreme ends of the aperture range. This implies there is more detail/sharpness within the image at these middling f/stops. That is through a combination of increasing DOF bringing more into focus, before diffraction pops its head up and starts to take its toll. This is exactly as Bob described. I have included screen prints of the file listings for the 100% crops, the 50% crops and the full images. Here they are....

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

Tim, have you been nominated for Measurebator status? I hereby nominate you, if someone else hasn't !

Excellent test methods and analysis. I wish even half of the posts would be similarly prudent in elimination of non-contributory variables before inquiring ("Is my lens sharp or not?", accompanied by a hand held shot at 1/8 )

Has anyone done this type of test between the 50D and the 5D Mark II?

tdodd wrote in post #7569143
I agree, between the 40D and the 50D there is less to concern one regarding the cameras than over the choice of lens or the photographer's own skills, but many still ask whether they should get the 40D or the 50D, and many still concern themselves with pixel level noise, when there are more important things to consider, and some refuse to believe that the 50D is an improvement on the 40D. Others may consider the 50D to be better than the 40D but not worth the price premium.

The thing is, I wonder how many of these opinionated people have actually used both cameras in anger and squeezed the last drop of performance from either one? I'm not saying I've pushed them to the limit, but wedding photography is demanding enough in terms of light levels and the 50D can deliver very worthwhile results at 3200 ISO. It's time the "per pixel" noise debate got laid to rest. The important thing is - how do the photos look, not the wretched pixels.

Between the 40D and 50D the big questions to be answered are....

- Did the extra resolution come at a price of too much noise? to which the answer is that I think not.
- Does the extra pixel density offer tangible benefits in terms of image detail? to which the answer is that I think it does, and the 50D is equiped with features to make those megapixels count, like AF microadjustment and a higher resolution screen, higher resolution preview image and an additional AF system.

Here's a shot taken today with my 50D, 100-400 zoom and a Kenko 1.4X teleconverter. This was shot at 560mm, f/8 (wide open), 1/640 and 400 ISO. This is a 1/4 frame crop resized to 50%. Notice the thin DOF leaves no margin for error with focus. This was focused using Live View Contrast Detect AF. I suspect with a 400mm (or longer) prime the image might have been sharper, but I think it turned out OK.

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

I had my 1D3 with me but compared to the 50D the reach was simply not there, so it remained on standby, on the off chance that I switched to BIF photography rather than perched birds.

Very well said. I am in total agreement. I really enjoyed your additional comments:
-extra resolution
-extra pixel density

Later,
_Mark

Why would you run this SAME test between the 50D and the 5DMII? They are simply two different animals and not even in the range (price or sensor) of one another. If a $2700 FF body DID NOT outperform my $1300 DOLLAR APS-C body why would there be a need for the FF camera. This question makes notone iota of sense. Please explain to me why you desire to see this test.

Shooting from the hip here, since I don't have a 5D2, I'd say that in this test, and in the bird picture above, the 50D would beat the 5D2, quite simply because the 5D2 has an even lower pixel density than the 40D, equivalent in fact to the 30D. Thus the 50D will place almost twice as many pixels on your subject for any given focal length.

Now it's true that the 5D2 will have very beautiful pictures, with great dynamic range and colour tone, but it will simply not resolve as much detail as a properly focused 50D with a good lens - e.g. a 400/5.6L or 500/4L. Once the light falls, and the ISO goes up, perhaps the 5D2 will reclaim lost ground, but at low ISOs the 50D will have more detail.

If you want the extra pixels on the 5D2 to count then you'll need the longer glass to make up for the lower pixel density. That comes with a very hefty price tag plus extra bulk and weight. e.g 400/5.6L (640mm equivalent) costs £1009 today. There is no directly matching full frame equivalent but your options would include....

- 400/5.6 + 1.4X telecon for £1263 but that comes up short on focal length (only 560mm vs 640mm) and the aperture is now at f/8 so no AF on the 5D2.

- 500/4 + 1.4X telecon for £5,250. Max aperture is the same and you gain some length to 700mm vs 640mm, but that's a pretty stiff price premium.

- 300/2.8 + 2X telecon for £3,800, but again we're a little short on focal length

- 600/4 - for £7335. I think I'll stop there.

So, yes, feed the 5D2 with the right glass and it can match or beat the 50D on resolution and with higher per pixel (and per picture) IQ, but it will not be a cheap, light, compact solution for wildlife togs.

Let's not forget, that bird was shot at 560mm on my cropper (with zooming capability). On a 5D2 I'd need an 896mm lens to fill the frame equally, and I would not have the zoom facility if I chose Canon glass. To match that photo I'd be looking at the 600/4 + 1.4X, or the 800/5.6, or perhaps the Sigmonster. Any of those solutions is the wrong side of £7,000 and only the Siggy (almost 6Kg!) retains zoom flexibility.

On the flip side, the price to pay for the high resolving power of the 50D is the noise at higher ISOs, and perhaps slightly less dynamc range. So for low light stuff like weddings, with short and medium length lenses, the 5D2 will be killer. In good light the 50D will offer tremendous value for money in terms of whole system costs, taking lens costs into account. For wildlife and sports the 50D is a very cost effective solution. It's no slouch at everyday stuff either. Just don't expect tiny pixels to each be quite as beautiful as large ones. That's the trade off.

EDIT : p.s. it is for the reasons above that I chose to use my 50D to shoot this bird, rather than my 1D3, which I had with me but which remained unused all afternoon, just hanging of my shoulder with a body cap for protection. There was simply nothing big enough or close enough to make it worth bringing the 1D3 into play. I'f there had been larger stuff flying about closer by then I would have switched to the 1D3 for the AF performance, but apart from gulls there was nothing evenly remotely interesting anywhere near me. Perhaps next time I'm shooting on the limit for range I'll try the 1D3 as well and see how the results compare.

tinner18 wrote in post #7571959
Why would you run this SAME test between the 50D and the 5DMII? They are simply two different animals and not even in the range (price or sensor) of one another. If a $2700 FF body DID NOT outperform my $1300 DOLLAR APS-C body why would there be a need for the FF camera. This question makes not one iota of sense. Please explain to me why you desire to see this test.

Simply to learn; to learn how the 50D compares to the 5DII in low light and normal light high ISO situations where I need higher ISO for some speed increase in low light situations, and higher ISO in normal light with moving objects, e.g., birds in flight.

Just simply to learn if the comparison tests have in fact been done and what are the results. For example, is it automatically assumed that at 3200 the 5D will have less noise than the 50D? Perhaps I have even asked the wrong question!

I do not have either camera; I am definitely planning on buying the 5DII prior to October if an upgrade is not released; I will probably but not definitely buy the 50D - jury is still out whether to upgrade the 40 to a 50 or wait for the 60 and just buy the 5DII.

So.........what was the conclusion here? I'm baffled by the techno talk.

Who "won" 40D or 50D?

Hi,

I do not know about YOUR tests; however, the scientific/photographi​c website that I rely on (www.clarkvision.com) provides the following in substance (from http://birdphotographe​rs.net …/showthread.php​?t=29039):

As I understand it, a very important factor is sensor-site (pixel) size. Big pixels like those generally found on full-frame cameras are like big buckets that catch more rain per unit time than small buckets. In our case "rain" is light, and light is the signal we are recording on the sensor. More light captured per unit time means you do not have to turn up the sensor amplifier so much to obtain a particular ISO equivalent, thus noise is proportionately much less in relation to the signal. I noticed this very clearly (night and day as they say) when I recently acquired a Canon 5D, 12.8 mp full-frame body. The pixels are over 8 microns across on this sensor, compared to half that size on my 50D. The noise difference is very noticeable, in fact Canon decided not to even offer high-ISO NR built-in to the 5D. It would be surplus to needs.

So all this is to say that rather than comparing megapixels or sensor size, use the pixel size to make choices about camera bodies. Table 2 of BPN member Roger Clark's excellent web site on the subject gives pixel spacing which is equivalent to size in modern sensors. Here are a few from well-known modern camera bodies in order from biggest to smallest (1 micron = 1 millionth of a metre):

Nikon D3/700: 8.5 microns (µ)
Canon 5D/1D mkII: 8.2 µ
Canon 1Ds mkII/1D mkIII: 7.2 µ
Leica M8: 6.9 µ
Canon 5D mkII: 6.4 µ
Nikon D3x: 5.9 µ
Canon 40D: 5.7 µ
Nikon D300: 5.5 µ
Canon 50D: 4.7 µ
Olympus E3: 4.7 µ
"point-and-shoots" are smaller still

Table established that based upon pixels on the subject, the 40D at 5.7 will have a higher IQ than the 50D at 4.7.

Another very interesting site to look at for comparisons of all of the various Nikon, Canon, and other mfgs sensors is: http://www.dxomark.com​/index.php/eng/DxOMark​-Sensor

So, while at the end of the day IQ is in the eyes of the beholder, the scientific studies give the gold star to the 40D over the 50D on the basis of the sensor IQ.

There are other factors pertaining to the 50D that makes it the overall winner between the two bodies; that is why I am selling my 40D and upgrading to at the least a 50D and at the most a 1D4 if Canon give birth in time for my next trip.