Canon [Image comparison] 12 years old camera kills, seeing is believing (Canon 40D)

jyc860923

Top Veteran
Location
Shenyang, China
Name
贾一川
This comparison is based on only one photo shot on each camera as I'm crazy busy these days, as much as I could to make a fair comparison regarding the 3D pop, the vibe, the micro contrast or tonality as some may call it, I believed the lower density, bigger pixels helps to make more vibrant photos especially with the 40D being early enough to have only 10MP and late enough to shoot 14bit raw. The result, as you will see in below, is not good news for my newer EOS M50, or rather the many new cameras it may represent. The comparison is made with the new to me, a used copy of Canon EOS 40D.

First of all, these two images are the best quality raw photos taken with first, the 40D, and second with the M50 both with a 50mm 1.8 STM lens in P mode, both cameras chose 1/250s, f5.0 although I had to manually set the 40D to use ISO100 as its auto ISO defaults to 400, I then took both photos into lightroom, enabled lens profiles, adjusted WB to be exactly the same number, everything else the same.

Now keep in mind when judging image quality the forum compression plays a factor here.

40D
IMG_7472.jpg
Join to see EXIF info for this image (if available)


Now the M50
IMG_4493.jpg
Join to see EXIF info for this image (if available)


And don't jump to conclusions yet, let's see some side by side comparisons, I've downsized the M50 photo to match the 40D resolution. Left 40D, right M50.

1.png


2.png


End of the comparison. I think the result speaks for itself. I'd really love to see more photos of the 40D kind.
 
Last edited:
In the Leica world- many stand with the Leica M9 as being the camera that has given them their best images. Some state the M8 gives the "most crisp" image.

Manufacturers do not release technical data as they used to ten and more years ago. Looking at the images: the spectral response of the sensors would be telling. Other factors include The noise reduction algorithms build into the sensor itself, and thickness of the filter stack.
 
I don't disagree with your findings in any way - I like the 40D image better - but I think the focus points aren't at the exact same spot; I think the M50 image is actually "misfocused" (focus point further back) as far as the reference plane is concerned - but that in itself would be quite interesting since CDAF should lock on more accurately - in theory, that is (but certainly true as far as my experience with FF DSLRs and mirrorless bodies goes). Anyhow, colour output from the 40D seems somewhat warmer and "stronger" (with bolder tones) than the M50's.

As for the Leica thing, I won't argue - I like the M10 files best. But I find the ones from the M8 fantastically pleasing - they require (but also tolerate) less work than the ones from the M10. Is it a resolution thing? I don't know, but honestly, after experiencing the M10's output, I'm less and less convinced that it's mostly sensor tech. You *can* get rich files from CMOS sensors - as long as you're prepared to compromise when it comes to other aspects (like DR and super-high ISO performance). My take: CCDs were (pleasingly!) limited, CMOS sensors can be tweaked. But that's just my guess from observation only. I'm very happy to be able to use both alongside each other.

I think it'd be very interesting to get more modern sensor tech in lower resolution sensors - something like the A7s/II from other brands, maybe? Makes me wonder about the Sony's RAW output, really ...

M.
 
Makes me wonder about the Sony's RAW output, really ...
I can send you some :). However, only for Sony A7 and A7R2. IMHO comparing color and tone rendition of raw files from different cameras is quite similar to opening Pandora's box or a can of worms, take your pick. I make my own camera profiles for Lightroom with X-rite's Colorchecker Passport, which makes a real difference and can bring the output of various cameras close to each other. Comparing the output of raw files with standard profiles of Lightroom/Camera Raw doesn't make sense to me. Oops, just peeked into Pandora's box :).
 
I can send you some :). However, only for Sony A7 and A7R2. IMHO comparing color and tone rendition of raw files from different cameras is quite similar to opening Pandora's box or a can of worms, take your pick. I make my own camera profiles for Lightroom with X-rite's Colorchecker Passport, which makes a real difference and can bring the output of various cameras close to each other. Comparing the output of raw files with standard profiles of Lightroom/Camera Raw doesn't make sense to me. Oops, just peeked into Pandora's box :).
Thanks, but I'm really only interested in the A7S - I own the A7 II myself, and while it produces nice enough files, it's pretty middle-of-the-road as far as FF 24MP file quality is concerned. That's not to say it's bad in any way - I find FF 24MP spot on as far as advantages go, and the Sony's a fine example. But I like the files from the M10 better (and then some), and in good light, even the Canon G1X III with its APS-C sensor competes well (and the Nikon D5500's files were better than those). Both my Nikons (D750 and Z6) outperform in as far as pure IQ is concerned.

As for camera profiles, that's a great idea and definitely speeds things up no end - but I'm very much a hands-on guy, so every image is processed on its own terms with a minimum of tools and presets. It's very silly and old-fashioned, I know, but I simply don't have the big junk of spare time to dig into the finer points of creating such profiles. But this also means that I compare unprocessed and unprofiled RAW files if I ever do so; in darktable, I always revert to "Original" and then set WB manually (for direct comparisons, the most adequate colour temperature will be set, but I don't do that on a regular basis). I don't mind cameras having their own "personalities", really - that's what I mean if I say files from one camera are easier to process than those from another. It's a different approach - I'm no pro, so uniformity is not what I seek because it actually doesn't matter. I have my preferences (in many respects), so I'm fine with that showing through.

M.
 
In the Leica world- many stand with the Leica M9 as being the camera that has given them their best images. Some state the M8 gives the "most crisp" image.

Manufacturers do not release technical data as they used to ten and more years ago. Looking at the images: the spectral response of the sensors would be telling. Other factors include The noise reduction algorithms build into the sensor itself, and thickness of the filter stack.
Speaking of the Leica "M"s, now I'm starting to think of a debate weeks ago, about CCD vs CMOS, I think whatever edge the CCD sensors may or may not have is more likely due to them being lower pixel count sensors.
 
Last edited:
I don't disagree with your findings in any way - I like the 40D image better - but I think the focus points aren't at the exact same spot; I think the M50 image is actually "misfocused" (focus point further back) as far as the reference plane is concerned - but that in itself would be quite interesting since CDAF should lock on more accurately - in theory, that is (but certainly true as far as my experience with FF DSLRs and mirrorless bodies goes). Anyhow, colour output from the 40D seems somewhat warmer and "stronger" (with bolder tones) than the M50's.

As for the Leica thing, I won't argue - I like the M10 files best. But I find the ones from the M8 fantastically pleasing - they require (but also tolerate) less work than the ones from the M10. Is it a resolution thing? I don't know, but honestly, after experiencing the M10's output, I'm less and less convinced that it's mostly sensor tech. You *can* get rich files from CMOS sensors - as long as you're prepared to compromise when it comes to other aspects (like DR and super-high ISO performance). My take: CCDs were (pleasingly!) limited, CMOS sensors can be tweaked. But that's just my guess from observation only. I'm very happy to be able to use both alongside each other.

I think it'd be very interesting to get more modern sensor tech in lower resolution sensors - something like the A7s/II from other brands, maybe? Makes me wonder about the Sony's RAW output, really ...

M.
I had the same doubt about focus point and technically the chance they are 100% identical is extremely slim, but I did focused carefully and this is typically how people shoot landscapes anyway, far objects, stopped down aperture and choosing the same AF area.

I want to see something like an A7S III from Sony just so to know the industry isn't ruined by megapixels yet, and the discussions I've read don't usually agree on Sony having the best colour so I really hope they build a stronger colour science on the awesome hardware.
 
they require (but also tolerate) less work
that's exactly my experience with the Olympus E1. In the right conditions its results are spot-on - in conditions that exceed its dynamic range or sensitivity there's not much you can do.

I guess the newer cameras (and their colour profiles) deliver "flatter"/more neutral results - leaving the work to us to make the images nice and contrasty and colourful...
The older cameras often deliver contrasty&colourful straight out of camera... (with - on the flipside - no way of making the images flatter/more neutral if required...)
 
Thanks, but I'm really only interested in the A7S - I own the A7 II myself, and while it produces nice enough files, it's pretty middle-of-the-road as far as FF 24MP file quality is concerned. That's not to say it's bad in any way - I find FF 24MP spot on as far as advantages go, and the Sony's a fine example. But I like the files from the M10 better (and then some), and in good light, even the Canon G1X III with its APS-C sensor competes well (and the Nikon D5500's files were better than those). Both my Nikons (D750 and Z6) outperform in as far as pure IQ is concerned.

As for camera profiles, that's a great idea and definitely speeds things up no end - but I'm very much a hands-on guy, so every image is processed on its own terms with a minimum of tools and presets. It's very silly and old-fashioned, I know, but I simply don't have the big junk of spare time to dig into the finer points of creating such profiles. But this also means that I compare unprocessed and unprofiled RAW files if I ever do so; in darktable, I always revert to "Original" and then set WB manually (for direct comparisons, the most adequate colour temperature will be set, but I don't do that on a regular basis). I don't mind cameras having their own "personalities", really - that's what I mean if I say files from one camera are easier to process than those from another. It's a different approach - I'm no pro, so uniformity is not what I seek because it actually doesn't matter. I have my preferences (in many respects), so I'm fine with that showing through.

M.
Fair enough. It took me quite some time to get the color rendition I wanted from my Sony A7R2. Not the camera's fault, I just always need months and months to fine-tune the post-processing of a specific camera to my taste, even with using a tool like X-rite Colorchecker Passport. Another reason for me to stick to essentially one camera.
 
Fair enough. It took me quite some time to get the color rendition I wanted from my Sony A7R2. Not the camera's fault, I just always need months and months to fine-tune the post-processing of a specific camera to my taste, even with using a tool like X-rite Colorchecker Passport. Another reason for me to stick to essentially one camera.
Don't get me wrong, I'm deeply impressed by such dedication - I just can't muster it! And as you indicate, shooting with multiple cameras would multiply the task ...

M.
 
Back
Top