garyvot wrote:
Insightful comments so far... at least about human nature!
WRT your question, the (very) short answer is that the a 50D RAW capture has more noise than the 40D at the pixel level. However, this is mitigated by two things:
1) Enhanced noise reduction in Canon's Digic IV procesing chain (including in DPP). This means that files out of the camera or DPP will show a bit less noise than 40D files at similar ISOs (though they may appear slightly softened and less saturated by NR)
2) Increased resolution, which reduces apparent noise at reproduction size vs. a lower resolution camera..................................snip....Show more →
Gary, you're not alone in this misconception since Art Morris feels the same way. Up until now at least, no Canon body does noise reduction in RAW.
It would take some serious pixel peeping to see worse noise in the Canon 50D. DxOMark says the 40D and 50D are nearly identical. The 50D is the same Canon technology with more pixels. If you don't want high MP images, the 50D has various options for saving lower MP images.
The 50D is less expensive than the Nikon 300, so that's not a fair comparison.
hasanahmad wrote:
So i go through a lot photography forums and there is a dislike of Canon 50D. Some even say the high ISO performance of the 40D is better than 50D. So what is true? Why is it that the latest in the double digit D models is getting the crap while 40D is being the 'preferred' option?
Okay, you hear it stated here first:
The Basic Rule of Canon XXD Cameras: Even-Numbered XXD Canons Do Not Suck
(Yes, I know that technically they are all even numbers, but I hope someone gets the pop culture reference.)
Chris Beaumont wrote:
Interesting....I bought my 40D in May 08, was yours an early one ?
Plastic hinge for mine as well. Dont know its age as I got it 2nd hand but the guy in the store said it was not that old which looks like he was telling the truth
brainiac wrote:
Every time I have seen someone complain about 50D noise relative to 40D it has gradually emerged that she was comparing pixel for pixel. The mistake of comparing these two cameras at 100% without allowing for the greater enlargement of the 50D file is rife.
Bingo!
People keep making this mistake when comparing cameras with the same sensor format but different numbers of photosites. I saw - and still often see - people making comparisons of 100% magnification crops from a 5D and a 5D2 with the same lens and concluding that the 5D is sharper. Duh!
Andrew J wrote:
Gary, you're not alone in this misconception since Art Morris feels the same way. Up until now at least, no Canon body does noise reduction in RAW.
It's not a misconception, but I think you misunderstood me. I am well aware that no noise reduction is performed on the RAW data by the camera (thank goodness).
My point is that Canon's "RAW processing chain" (e.g., processing to JPEG in the camera or in DPP 3.6) *always* applies noise reduction to the image, even at 100 ISO. (In DPP you can disable this, but the default setting is to apply the camera NR settings.)
Canon's "low noise" claims for its Digic IV cameras are based not only on native sensor performance, but on the use of their own processing chain (Digic IV or DPP). For this reason, people using DPP on a casual basis usually report a different experience than do those who use ACR... The fact that camera JPEGs always get NR also accounts for (some of) the criticism of 50D per-pixel sharpness.
People keep making this mistake when comparing cameras with the same sensor format but different numbers of photosites. I saw - and still often see - people making comparisons of 100% magnification crops from a 5D and a 5D2 with the same lens and concluding that the 5D is sharper. Duh!
+1
Dan
I agree there are a lot of misconceptions about this, but I do think that some of the better-informed critics have valid concerns about sensor-level performance being degraded by ever-higher pixel densities.
Ostensibly one reason for the higher pixel density is greater croppability, but if shooting at ISO 800 renders ugly blotching and huge gobs of noise at 100% (without NR), you are not left with much room to crop...
I'd like Canon and Nikon to push the pixel density but not at the expense of seriously degraded sensor noise, otherwise at high ISOs you are effectively left with a 6 or 8 mp camera, and you might as well be using a 20D.
Ostensibly one reason for the higher pixel density is greater croppability, but if shooting at ISO 800 renders ugly blotching and huge gobs of noise at 100% (without NR), you are not left with much room to crop...
No, you test it as you print it. Greater croppability does not mean 100% without noise reduction.
RDKirk wrote:
No, you test it as you print it. Greater croppability does not mean 100% without noise reduction.
NR degrades fine detail and sharpness, even using the best tools. And at some ISO, the amount of detail obscured by sensor noise cannot be recovered after NR and the quality of the image is reduced to the point where it is no longer acceptible even at reproduction size.
This threshold varies, of course, depending on too many factors to discuss, and each person's perception of "acceptible quality" is likely different as well.
It degrades that sensor's fine detail and sharpness, but that sensor may have characteristics that enable it to withstand the degradation with more grace than another sensor does. A larger format, for instance, can make this possible.
