Canon 50D incredibly noisy?

[jani]

I think it's fascinating that my now four year old 20D and my pal's 1D MkII hold up so well in DxOMark.

[ejmartin]

Resolution vs pixel level noise is always the tradeoff for a fixed size sensor, if one's criterion is noise at the pixel level.  That is simply because noise power rises with increasing fineness of scale in the image.  All other things being equal, decreasing the pixel size to increase resolution simply samples the scene at finer scales, and noise at those finer scales is necessarily higher due to the physics.  But pixel level noise is a moving target as one changes the pixel size, because of the scale dependence of noise; noise at fixed scale remains surprising constant as pixel size is decreased.

For instance, here's a plot of noise power vs image scale (spatial frequency) for a 40D (red) and 50D (blue); test images from Imaging-Resource; ISO 1600, converted from RAW in DPP:



The horizontal axis is scale in the image, the vertical axis is noise power; so each data point is a measure of the amount of noise at a particular scale.  The somewhat arbitrary units for the horizontal axis put the Nyquist frequency (the limit of resolution) at 256 for the 50D, and at 209 for the 40D.  The two cameras are more or less the same up to the point where the 40D stops resolving, while the 50D climbs a bit higher.  So, the pixel-level noise in the 50D is higher than that of the 40D, simply because it resolves more.   At comparable scales in the image, noise is identical for all practical purposes.  The higher noise at higher scales is subsumed in a finer "grain pattern" of the 50D, which some find more pleasing.  If on the other hand one is intent on comparing images at 100% view on a monitor, one will essentially be looking at the 50D noise power at 256 and the 40D noise power at 209, which is a biased comparison which will stack the deck against the smaller pixel camera.

If one were to downsample the 50D image to the 40D pixel dimensions, one gets the following



Original images as before; orange points are downsampling in PS using Bicubic, black points are downsampling with ImageMagick using the Lanczos filter.  It may seem that the downsampled 50D is a bit worse than the 40D (though extremely little for Lanczos), but it also has more resolution since the 40D has to contend with its AA filter, while the 50D's AA filter cutoff is out of the picture in the downsampled image (also, downsampling mitigates some rather nasty interpolation artifacts in DPP).

So the noise and the resolution come hand in hand -- one simply has to decide what one wants (if you want the latter without the former, increase exposure, or get a bigger sensor).  And as the above shows, one can always generate the lower resolution data from the higher resolution data, but one can't recover detail that wasn't recorded.

[stewarthemley]

Before you get the knickers in a knot...

No problem there, Panopeeper - I don't wear them!

[joofa]

Removed.

[BernardLanguillier]

Do they offer any rational scientific basis, or just the usual misguided comparisons of per pixel noise and DR levels and 100% on-screen images, along with the currently fashionable dogma that more, smaller photosites on the same sensor size always gives worse image quality?

I ask because of the specific suggestion that staying with the 40D would be better, which is flatly contradicted by the observations in Bob Atkins' review of the 50D at http://www.bobatkins.com/photography/digit...D_review_1.html

These guys are widely seen as being the best photo magazines in Europe. They combine both real world images, prints, and DxO analysis tools.

They are saying that there is more detail at low ISO with the 50D compared to the 40D, but that higher ISO image quality is roughly on par. All things considered they are saying that the 50D is a bit better, but not enough to justify an upgrade for existing 40D owners.

I feel that the 50D is undergoing a bit of an unfair treatement overall, partially because people keep sticking to ACR as a test raw converter, although it appears more and more that it is one of the worse offerings around. Also, there is globally a clear trend towards Nikon/Sony and away from Canon that contributes also to people doubting the ability of Canon to release valuable new releases.

My view is that as much as there used to be too much overlooking of Canon shortcoming 2 years ago, there is now too much criticism being cast.

Cheers,
Bernard

[Ray]

They are saying that there is more detail at low ISO with the 50D compared to the 40D, but that higher ISO image quality is roughly on par. All things considered they are saying that the 50D is a bit better, but not enough to justify an upgrade for existing 40D owners.

That sounds about right to me, Bernard. It's in accordance with my own tests. I would agree that image quality alone is not sufficient reason to upgrade. Or to be more precise, image quality directly related to the increased pixel count is not sufficient reason to upgrade from a 40D. But there are at least a couple of sweeteners that Canon have thrown into the mix, namely, a higher resolution LiveView LCD screen which is just amazing at 10x magnification with a 400mm lens attached (on tripod of course and preferably with IS enabled), and an autofocus micro-adjustment feature. Both of these additional features will potentially contribute to better image quality over and above what can be achieved merely with the increase in pixel count.

Edit: As BJL mentioned, Bob Atkins in his review of the 50D demonstrated that, even at ISO 3200, the 50D image lends itself to a degree of noise reduction which can lower noise to the 40D level, and even better, whilst still retaining the resolution edge over the 40D.

I haven't tested this myself and it looks like Chasseur D'Image haven't either, but there's a certain logic to the process. We know that noise reduction programs tend to destroy fine detail. We know also that sacrificing resolution by downsampling can reduce apparent noise, resulting in high-ISO 21mp 1Ds3 images having similarly low noise to a 12mp D3 image, for example, when downsampled.

It looks as though a similar process applies to high-ISO 50D images when some degree of resolution is sacrificed, not by downsampling, but by the application of a noise reduction routine. It looks as though this might be a more successful or more efficient way of equalising noise than downsampling.

This probably deserves more investigation because, if the effect is consistent, it means that the 1Ds3 and 5D2, with appropriate noise reduction, have lower noise than the D3   .

Of course, I'm aware of the counter argument. Any noise reduction that is applied to the 5D2 or 1Ds3 can also be applied to the D3 image to reduce its noise even further, but alas! not without also reducing D3 resolution. The resolution gap remains and the bottom line might well be, after all noise reduction options have been explored, that a 5D3 image at high ISO can have either lower noise than the D3 but equal resolution, or equal noise but marginally higher resolution.

Something to consider  

[BJL]

They are saying that there is more detail at low ISO with the 50D compared to the 40D, but that higher ISO image quality is roughly on par. All things considered they are saying that the 50D is a bit better, but not enough to justify an upgrade for existing 40D owners.

Ah, that is a different and far more reasonable story than other options like the 40D being preferable to the 50D, or the 50D having "too much resolution": some IQ advantages, no disadvantages, not enough gain for most to replace their only one year old 40D!

I find strange this habit of judging a new camera by this standard of being worth upgrading from the previous model in the same product line, from only a year or two earlier. Surely the far more common likely customers are
1. those with a camera three or more years old, for whom the technological progress is quite significant.
2. those upgrading from a lower level camera: with the 50D, say an entry level DSLR, or even a digicam.
Or am in the minority for still using a four year old DSLR while I wait for the must-have upgrade?!

[BernardLanguillier]

Or am in the minority for still using a four year old DSLR while I wait for the must-have upgrade?!

Are you saying four years old, like in "4" years old???  

Cheers,
Bernard

[jani]

Or am in the minority for still using a four year old DSLR while I wait for the must-have upgrade?!

I suspect you are, but then again, so am I.

The upgrades have been just short of worth it for the past few years, and it seems like I'll be sitting on this fence while the grass grows to my knees.

[BJL]

Are you saying four years old, like in "4" years old???

As in, I admit to still using my horrendously obsolete (so they say) Olympus E-1, purchased in early 2004! (The E-30 is at last tempting me.)

I am more comfortable about being in the majority when it comes to replacing my DSLR less often than every two years, meaning skipping at least one models in the sequence of updates of "amateur" SLR product lines like 10D->20D->30D->40D->50D.

[stewarthemley]

My first digital camera was an Olympus E something or something E. It was the one that appeared about the same time as the Canon D30. Only went up to 400iso, had a fixed lens that you screwed a wide or tele bit on, had a silent (totally) shutter or you could tell it to make a noise like a shutter! And it took bloody brilliant images. I have a theatre client who still has some 20x16's on her wall and they look every bit as good as my 1DS3 and now 5D2 shots. Noisier of course but the noise was so film-like it added to the atmosphere. I think it was only about 4mp but it blew up a treat. So I suppose I shouldn't make too many disparaging remarks about the 50D. And now I'm sending my own thread off topic!

[Ray]

As in, I admit to still using my horrendously obsolete (so they say) Olympus E-1, purchased in early 2004! (The E-30 is at last tempting me.)

I am more comfortable about being in the majority when it comes to replacing my DSLR less often than every two years, meaning skipping at least one models in the sequence of updates of "amateur" SLR product lines like 10D->20D->30D->40D->50D.

BJL,
I feel I should offer you my 40D at a mate's price since I obviously no longer need it. But I probably would not be doing you a favour because, I presume, you have no Canon lenses.

I think I've mentioned before, I bought the 40D in unusual circumstances. It wasn't a sufficiently major upgrade from my 20D to get me interested. However, whilst shopping for an EF-S 17-55/2.8 lens in Bangkok, a beautiful Thai shop assistant offered me a very attractive price on a 40D body. The price was not only significantly less than the best internet price I'd seen in Australia, but the shop assitant was also very charming.

Unfortunately, that 17-55mm lens which has an excellent Photozone rating at F2.8, did not autofocus accurately on the 40D, and was therefore almost useless. I bought it because it was supposedly sharp at F2.8 and therefore an excellent walk-around lens in low-light conditions where flash would be an intrusion.

[jjj]

I did some shots yesterday on my mate's 40D and was very surprised at how very noisy they were. And I used to use a 20D some years back and the images look poor in comparison to that camera. Very surprised by that.
Could have been a poor lens though as well as I use L lens and the lens on 40D was a 70-300mm Canon, but it was the noise that surprised me most. The shots looked like a P+S camera!

[jani]

I did some shots yesterday on my mate's 40D and was very surprised at how very noisy they were. And I used to use a 20D some years back and the images look poor in comparison to that camera. Very surprised by that.
Could have been a poor lens though as well as I use L lens and the lens on 40D was a 70-300mm Canon, but it was the noise that surprised me most. The shots looked like a P+S camera!

This is quite similar to the impression I had when looking at the sample photographs at DPReview and other places that offered them.

A small trial at an exposition in Oslo showed that while there might be other reasons for upgrading, there were no significant differences, at least for my uses. 2 megapixels extra, no improvement in noise (and seemingly lower sensitivity, removing 1/3 stop advantage in the first place).

I've been saving for a 1-series camera for a couple of years, and thought that the 1D MkIII would be the camera for me (the silent shooting mode is pretty good!), but alas, it's not. My money's still in the bank -- and the bank isn't broke!

I've actually considered purchasing a used Hasselblad 503, a couple of lenses and using chemical film instead.

[Ray]

I did some shots yesterday on my mate's 40D and was very surprised at how very noisy they were. And I used to use a 20D some years back and the images look poor in comparison to that camera. Very surprised by that.
Could have been a poor lens though as well as I use L lens and the lens on 40D was a 70-300mm Canon, but it was the noise that surprised me most. The shots looked like a P+S camera!

Perhaps we should request that Michael create a special section for silly comments  .

Here's a 40D image which I would entitle, "The Four Fours"   . Canon 40D, 4000mm lens, ISO 400, 1/4000th sec exposure.

[attachment=10305:The_four_fours.jpg]

[jjj]

Perhaps we should request that Michael create a special section for silly comments  .

Like this response.  
I also thought the screen quality on back was horrible, looked like something off a Hasselblad.
Maybe he has a lemon.