just like to point out some facts:

Dpreview used the canon 50/1.4 @ f8 (please correct me if i'm wrong)

The MEASURED resolution of the 50mm/1.4 at f8 is 70 line pairs per mm (50% MTF).
here: http://www.wlcastleman.com/equip/reviews/50mm/index.htm
(please read in detail how is this measured)

50D (based on 15Mpixel) resolution is 105 line pairs per mm
40D (based on 10Mpixel) resolution is 86 line pairs per mm

Theoretical (http://www.luminous-landscape.com/tutorials/resolution.shtml)
for f8 is 86 line pairs per mm (50% MTF)

And you don't expect the Canon 50/1.4 @f8 to outresolve the theoretical resolution do you?

You decide for yourself.


update:

Please note that 50% MTF is "perceived image sharpness" as defined in the first link and absolute resolution is higher. The image can be sharpened to "bring out" the details so that it can be "perceived" by human eyes.

one line pairs per mm means one white and one black line per mm.

Update on digital sensor resolution:

(approximately only) In pixel term is 2 pixels per mm (one pixel for white and one pixel for black.
For its accurate treatment please refer to the link in the post
http://www.fredmiranda.com/forum/topic/703015/0#6332889

So as the 50D is outresolving a sharp prime at it's sharpest aperture the test is flawed? Would you have just prefered them to say that the camera is outresolving the canon lens lineup? Oh, they did. What use is a camera that cannot reach it's potential in the real world?

Beni wrote:
So as the 50D is outresolving a sharp prime at it's sharpest aperture the test is flawed? Would you have just prefered them to say that the camera is outresolving the canon lens lineup? Oh, they did. What use is a camera that cannot reach it's potential in the real world?


sharpest for the 50mm is at f4.

DPreview could have used a better (non-canon) 50mm @f4 or f2.8. It was afterall a review and comparison of the camera.

I hope Canon is reading this and very soon we'll have a new improved range of canon primes. hopefully!

If the 50D will outresolve the 50 1.4 @ f/8 there will be a lot of lenses (even expensive L lenses) that are going to be outresolved by the 50D. The 50 1.4 @ f/8 is equally sharp or in some situations even sharper than many current Canon lenses (like the 24-70L, 24-105L, 50 1.8, 35L, etc). I think it is a good thing that DPreview used a lens like the 50 1.4. This way it will prove the point that you need very, very, very sharp glass to keep up with the 50D. Not to mention higher shutterspeeds to counter camera shake... 15MP on a 1.6x crop sensor is just too much.

To put a positive twist on it: Now the camera is not the limiting factor anymore. I think that's quite cool, actually.

Antje

Antje wrote:
To put a positive twist on it: Now the camera is not the limiting factor anymore. I think that's quite cool, actually.

Antje


that's right!

YES! Let's hold their feet to the fire for better lenses! Or if you build it (15Mp+ crop sensors) they (better line pair lenses) will come?

If the 15mp of the 50D outresolves canons best lenses what do 1Ds and 5D mkII users do with their 21mp?

freetime101 wrote:
If the 15mp of the 50D outresolves canons best lenses what do 1Ds and 5D mkII users do with their 21mp?

The 50D has a pixel density of 214/mm the 5DII only 156/mm same as the 20D

makron wrote:
Beni wrote:
So as the 50D is outresolving a sharp prime at it's sharpest aperture the test is flawed? Would you have just prefered them to say that the camera is outresolving the canon lens lineup? Oh, they did. What use is a camera that cannot reach it's potential in the real world?


sharpest for the 50mm is at f4.

DPreview could have used a better (non-canon) 50mm @f4 or f2.8. It was afterall a review and comparison of the camera.

I hope Canon is reading this and very soon we'll have a new improved range of canon primes. hopefully!

Not in my experience. I also find it a whole heck of a lot sharper at f4-5.6 than any of the three 24-70L's I've owned or my 70-200 f4L IS.

I actually prefer a real world test. Yes they could use ZE glass at the optimum aperture but that ain't going to tell you what using the camera with the vast majority of popular canon lenses is going to be like.

Of course if you are quibbling about apertures then to even get close to the expected resolution you better have a very solid tripod, MLU, etc.

I think the conclusion is true, unless you want it as a special needs camera (only the best of the best lenses), don't bother, keep your 40D.

It does put it's price advantage over the D300 into perspective though. It's cheaper but only has better IQ if you pay through the nose for L primes by which point its a lesser spec'd and far more expensive camera with only a marginal resolution advantage.

If you want a camera with better IQ and noise than the 40D - get a 5D mkI. Incredible chip, dirt cheap and will blow your socks off for IQ. According to Canon the AF is better as well!

Anyone know of a quick easy (maybe tabular) reference listing the measured resolution of the various Canon lenses?

I thought people say a few MP increase makes very little difference? Well the 450D is not that far behind from the 50D in terms of pixel density and I don't think it outresolves any lenses at all, not even the 18-55mm IS kit lens at f/8.

I don't get it... first people say pixel peeping is bad, total sharpness is not everything, etc etc. but now people are complaining about 50D outresolving a sharp lens.

freetime101 wrote:
If the 15mp of the 50D outresolves canons best lenses what do 1Ds and 5D mkII users do with their 21mp?

Thy probably use them to make photographs, and don't waste their time quibbling over nonsensical internet postings...

freetime101 wrote:
If the 15mp of the 50D outresolves canons best lenses what do 1Ds and 5D mkII users do with their 21mp?

FF "looks through" more of the lens than 1.6 crop.

You need to look at LP/mm and pixel densities to put them on an equal playing field.

My wife's 50D is doing just great regardless of anyones reviews. I now reach for her camera over my 5D about 1/2 the time. And I don't give a second thought to LP/mm or how well the 40D does for noise. It seems like people are in a gear bashing mood lately. I have to go to the nikon forums to hear someone appreciate canon gear.

It is a poor test, flawed in many ways the worst of which is the raw software used. First they used a beta version of ACR 4.6, then reposted with the latest 4.6. Anyone fooling with the 50D knows that ACR needs 5.1 and then it is really still very poor compared to DPP just as ACR couldn't properly handle the 40D for six months after it came out. The lenses is an issue but not the primary one here IMO. People are posting comparisons this morning at DPR of the difference between ACR and DPP and of course they are significant.

Frankly I would rather as MOLSON alluded to look at proper photographs taken by photographers who are after quality and use the right tools available at the moment. Some of the quality 50D pictures I have seen certainly speak for themselves. DPR has become the ultimate bit twitling pixel peeking forum, you have to take it with a box of salt, I would say a pinch but it is way beyond that now.

okay now here's my question... for all you of who did the testing...

AM I TAKING ADVANTAGE IF I BUY A 50D UNIT OVER THE OLD XXD MODELS?

forget about the money\price for now and its handling..let's concentrate on image quality and noise for now.

or WE CAN JUST SAY 50d IS A COMPLETE GARBAGE?

RCicala wrote:
Anyone know of a quick easy (maybe tabular) reference listing the measured resolution of the various Canon lenses?

I second that request, just for curiosities sake.

I think the EYES of EOS III has it... but that's only the MTF charts.

Sorry, the DPR test may be flawed, but so is your reasoning

If the lens resolution is 70lp/mm, then the total system resolution can never exceed that value.

In your example, the total resolution of the 50D would be:
1/70 + 1/105 = about 1/42 also a total resolution of: 42 lp/mm

the total resolution of the 40D would be:
1/70 + 1/86 = about 39 lp/mm

That's pretty close or not

The point is that a sensor with more pixels, always will give a higher total resolution, regardless of the lens used. Thus: one can still use the same lens to compare different sensors.

Oh, and one more thing: IIRC, you cannot just derive the sensor resolution from the number of horizontal pixels.


If you want to learn more about resolution see this site:
http://www.normankoren.com/

freetime101 wrote:
If the 15mp of the 50D outresolves canons best lenses what do 1Ds and 5D mkII users do with their 21mp?

It doesn't, see my previous post

So does this camera take pictures?

Antje wrote:
To put a positive twist on it: Now the camera is not the limiting factor anymore. I think that's quite cool, actually.

Antje
Way to go for the positive spin! I'm curious. Let me try to phrase my question in big chalky lines: Since the camera will out-resolve a very sharp lens by almost a factor of 2 will this lessen well known digital artifacts like fringing? Seems the more pixels you have covering the image the less ambiguity you'd have with the de-mosaicking process.

I think that many people are trying to make is that the 50D isn't that much of an improvement over the 40d, image quality wise. Most users have commented that the high ISO capability is more or less the same.

It may be difficult to accept this fact after dropping down cash for the newest and latest Canon model, but there is lots of truth in the user reviews and DPreview that we've been getting.

It's not that the 50D is garbage, or doesn't work, it's that the marginal improvement over the 40D may not be justified by the price difference, especially since 40Ds are going for such bargain prices now.

In my experience, I felt the same way after shooting with a 40D and a 50D- that the IQ is so similar, in fact I felt the pictures from the 40D came out more crisp and with more "punch." Of course, that is a matter of opinion, and also dependent on the in-camera JPG conversion, or your choice of RAW converter. But under the same circumstances, I found the performance of the 50D to be marginally better at best, than the 40D. I also concur with the other users, that the claimed high ISO improvement, was an overstatement.

Don't get me wrong, I really like the new screen and body material, but you can't help but admit that to get the actual benefit out of the 50D, which presses lenses to their optical limits, you have to get the best lenses. Otherwise, you would be better off with a 40D and the same glass. If you can't take advantage of the sensor with good glass, then you really aren't getting your money's worth.

The argument is which lenses actually limit the 50D's performance- but there is no question, that the 50D is a marginal improvement over the 40D if you have anything but the sharpest lenses.


crypt wrote:
My wife's 50D is doing just great regardless of anyones reviews. I now reach for her camera over my 5D about 1/2 the time. And I don't give a second thought to LP/mm or how well the 40D does for noise. It seems like people are in a gear bashing mood lately. I have to go to the nikon forums to hear someone appreciate canon gear.

freetime101 wrote:
If the 15mp of the 50D outresolves canons best lenses what do 1Ds and 5D mkII users do with their 21mp?
The FF 21MP pixel density is way below the resolving power of most Canon lenses. Nobody complained of lenses being outresolved with a 30D or 20D, so logically the same lenses, assuming their outer circle is similar to their inner 1.6x, shouldn't have a problem.