Reading mode: Light Dark

The image qualities of Sony vs Canon sensors

Go to last post
bclaff said:
As a student would I be provided with raw images from both cameras of identically lit, identical scenes, taken at the same exposure and preferably with the same lens?
That might be interesting.
Click to expand...
DPR attempts to do that with their "Studio Scene". You can compare up to 4 cameras when looking at the full review of the Sony a7Rii:
https://www.dpreview.com/reviews/sony-alpha-7r-ii/5

And compare two cameras from the recent full review of the Canon 5d4:
https://www.dpreview.com/reviews/canon-eos-5d-mark-iv/9

DPR generally uses a high quality prime lens to make the tests, which should provide a relatively "level playing field".
 
Last edited:
andrewD2 said:
BBQue said:
andrewD2 said:
JimKasson said:
Midwest said:
JimKasson said:
Midwest said:
How does 14-bit (Canon) work compared to 12-bit (whoever that is/was)?
Click to expand...
I don't understand the question.

Jim
Click to expand...
To the limited extent of my understanding, unless things have changed, Canon outputs a 14-bit uncompressed RAW image while Sony (at least used to) output a compressed 12-bit RAW image. Perhaps nothing is lost in the compression but I see the opportunity where it might be, given the difference in values possible in 14-bit vs 12-bit. If so that would seem to work against the idea of smoother gradiations in color in Sony images.

And I would caution anyone against using a likely compressed JPG image website to draw IQ conclusion about something so 'fine' as these differences.
Click to expand...
14-bit uncompressed raw is available from several a7x cameras, including the a7RII. The dynamic range is unaffected by craw compression, as shown here:

http://blog.kasson.com/?p=12287

The a7x cameras switch to 12-bit precision with some shutter modes and for bulb setting and LENR.

Jim

--
http://blog.kasson.com
Click to expand...
If you look at real images you can see compressed raw hugely affects the usable dynamic range of SOME images (those with contrasty edges like a stained glass window against dark stone).

Andrew
Click to expand...
It may give artifacts at these edges, and usually only if you severely process the file,
Click to expand...
The ability to push the shadows is what gives you the high dynamic range.
Its NOT "severe" processing. It is processing that uses the full dynamic range for a high dynamic range scene.
What you are basically saying is, "if you don't require use the full dynamic range of the camera you won't notice craw is limiting it". No surprise there.
Andrew
Click to expand...
Well no, you still do have the full dynamic range available. Examples I have seen of these artifacts was by intentionally underexposing 5 stops or so, then trying to recover the shadows by pulling them out 5 stops. That's not a typical photo application.

In a typical, properly exposed image, you may get dot artifacts around edges of extreme contrast, but you are not losing any dynamic range. At least that is my understanding and experience, which is why I continue to use compressed raw, even though uncompressed is available on my camera. Is your experience different?
 
Last edited:
JimKasson said:
andrewD2 said:
JimKasson said:
andrewD2 said:
JimKasson said:
andrewD2 said:
JimKasson said:
Midwest said:
JimKasson said:
Midwest said:
How does 14-bit (Canon) work compared to 12-bit (whoever that is/was)?
Click to expand...
I don't understand the question.

Jim
Click to expand...
To the limited extent of my understanding, unless things have changed, Canon outputs a 14-bit uncompressed RAW image while Sony (at least used to) output a compressed 12-bit RAW image. Perhaps nothing is lost in the compression but I see the opportunity where it might be, given the difference in values possible in 14-bit vs 12-bit. If so that would seem to work against the idea of smoother gradiations in color in Sony images.

And I would caution anyone against using a likely compressed JPG image website to draw IQ conclusion about something so 'fine' as these differences.
Click to expand...
14-bit uncompressed raw is available from several a7x cameras, including the a7RII. The dynamic range is unaffected by craw compression, as shown here:

http://blog.kasson.com/?p=12287

The a7x cameras switch to 12-bit precision with some shutter modes and for bulb setting and LENR.
Click to expand...
If you look at real images you can see compressed raw hugely affects the usable dynamic range of SOME images (those with contrasty edges like a stained glass window against dark stone).

Andrew
Click to expand...
I know what you're talking about, although that's not precisely a DR effect. Can we agree that craw compression affects neither EDR nor PDR?
Click to expand...
I compare DR using sets of bracket images, it is a LOT more useful than the online data if you have the cameras to hand. So if EDR or PDR are not affected by craw then both measurements are not always going to be an accurate indicator of real world DR.

For example compare a 5DIII dynamic range with a NEX 5R. The darkest tones recorded (given the same highlights) are actually deeper into the shadows on the Canon. The structure of the noise on the Canon is the problem, it is in bands that are easily identified by the eye. The range is there, it is just not as usable. DXO ignores the patterned noise and the 5DIII measures not much below the 6D.

Now compare a Canon 6D and compressed raw on the A7RII on a stained glass window image. The 6D shadows push up well enough to show some noise but also texture in the dark walls. The compressed A7RII images can't be pushed as much at all before noticable artifacts that obscure detail and are far more noticable to the eye than evenly distributed noise. The 6D beats the compressed A7RII for useable DR in those craw limited cases. The 6D never beats uncompressed raw on the A7RII for DR.

If you can't reveal the shadows then DR is limited, where that noise in the signal comes from, sensor readout or poor choice of compression algorithm matters not.
Click to expand...
Andrew, I don't disagree with the thrust of your argument. However, let's put craw compression in context. I have made about 800,000 exposures with a7x cameras since I was once of the first civilians to lay hands on an a7R. More than 95% of those have been made with craw compression turned on. I have never had a compressed exposure fail to succeed in accomplishing my photographic purpose, although I have been able to find defects in actual and synthetic images. In the latter case, I implemented the algorithm myself in Matlab.

http://blog.kasson.com/?p=4876
Click to expand...
Took me a week with the A7RII to hit problem senarios with craw. A Cathedral and a neo-gothic library. I'd been to the later with a 6D and the A7RII was producing a worse result because of craw. The sharpness of the highlight / shadow cutoff is important. The 55/1.8 I got with the A7RII produces craw artifacts edges.
Click to expand...
Notice I said that over 95% of my a7x images were made with craw on, not 100%. It's pretty obvious when you might have a problem. If you have an older model camera and can't turn craw off, you can always make two exposures.
Click to expand...
Why don't I use uncompressed raw all the time on, say, my a7RII? It's because Sony does not offer lossless compression. That's the option I use on my Nikon gear. It gets the files to within spitting distance of the Sony craw file size in most cases.

Jim
 
BBQue said:
andrewD2 said:
BBQue said:
andrewD2 said:
JimKasson said:
Midwest said:
JimKasson said:
Midwest said:
How does 14-bit (Canon) work compared to 12-bit (whoever that is/was)?
Click to expand...
I don't understand the question.

Jim
Click to expand...
To the limited extent of my understanding, unless things have changed, Canon outputs a 14-bit uncompressed RAW image while Sony (at least used to) output a compressed 12-bit RAW image. Perhaps nothing is lost in the compression but I see the opportunity where it might be, given the difference in values possible in 14-bit vs 12-bit. If so that would seem to work against the idea of smoother gradiations in color in Sony images.

And I would caution anyone against using a likely compressed JPG image website to draw IQ conclusion about something so 'fine' as these differences.
Click to expand...
14-bit uncompressed raw is available from several a7x cameras, including the a7RII. The dynamic range is unaffected by craw compression, as shown here:

http://blog.kasson.com/?p=12287

The a7x cameras switch to 12-bit precision with some shutter modes and for bulb setting and LENR.

Jim

--
http://blog.kasson.com
Click to expand...
If you look at real images you can see compressed raw hugely affects the usable dynamic range of SOME images (those with contrasty edges like a stained glass window against dark stone).

Andrew
Click to expand...
It may give artifacts at these edges, and usually only if you severely process the file,
Click to expand...
The ability to push the shadows is what gives you the high dynamic range.
Its NOT "severe" processing. It is processing that uses the full dynamic range for a high dynamic range scene.
What you are basically saying is, "if you don't require use the full dynamic range of the camera you won't notice craw is limiting it". No surprise there.
Andrew
Click to expand...
Well no, you still do have the full dynamic range available. Examples I have seen of these artifacts was by intentionally underexposing 5 stops or so, then trying to recover the shadows by pulling them out 5 stops. That's not a typical photo application.

In a typical, properly exposed image, you may get dot artifacts around edges of extreme contrast, but you are not losing any dynamic range. At least that is my understanding and experience, which is why I continue to use compressed raw, even though uncompressed is available on my camera. Is your experience different?
Click to expand...
If you are pushing 5 stops you are only underexposing if you are doing this when the highlights are a way from being clipped. In an extreme DR scene your shadows may be very dark because you are trying to expose for the highlights. That is where the DR comes from, being able to hold the brightest areas and still get good midtones and accepable shadows. It is not done by mistake.

I could say that using my Canon 5DIII and exposing for the subject and ignoring of the sky getting clipped that I'd rarely see banding in the shadows. Or you can fill flash or exclude the sky or use a diffuser etc etc. If I shot the A7RII the same way would rarely, if ever, see artifacts. I bought the A7RII because in lots of situations I'd like to be able to shoot I was unable to do so with the Canon. D750 and D800 photographers were at a competitive advantage. But with craw some of those situations came out worse than with my Canon 6D. It was a camera with high DR in most scenes, poor DR in others. Uncompressed raw was requested, some of my samples got to Sony. Since then zero problems, massive DR in any case. Never looked back and my images have imprived. That you accept 'dot artifacts' in even moderate DR scenes is surprising.

No such thing as a typical photo situation. Wait till you find the back of the brides dress comes to rest with a narrow beam of sunlight on it in an otherwise very dark church. Used to be you'd have to accept the blowout to get a reasonable subject, now you can hold the detail. It isn't a time for fill flash and I can't climb up 40 feet and scrim a church window!

Cheers, Andrew
 
JimKasson said:
JimKasson said:
Why don't I use uncompressed raw all the time on, say, my a7RII? It's because Sony does not offer lossless compression. That's the option I use on my Nikon gear. It gets the files to within spitting distance of the Sony craw file size in most cases.
Click to expand...
Click to expand...
My practice is to use Adobe's DngConverter to convert from .arw's to .dng's. That provides "pragmatic lossless compression" from about 85 MB uRAW's to 40 to 45 MB .dng's.

My impression are that there are some post processing tools that aren't able to use .dng's, so that may be a liability for some.

I have the camera set to uRAW for the great majority of my photos. The exceptions are when I use the PlayMemories TimeLapse app which forces use of cRAW.

Note it is possible to use the free Remote Camera Control app in tether mode to get the equivalent of TimeLapses, but with uRAW instead of cRAW.
 
l_d_allan said:
JimKasson said:
Is this more to your liking?

fc9ce9061d0c4794b5e6d6d21631cab1.jpg.png
Click to expand...
Here's a direct link, which also includes the Canon 6D with older technology (circa 2009'ish).
http://photonstophotos.net/Charts/PDR.htm#Canon EOS 5D Mark IV,Canon EOS 6D,Sony ILCE-7RII

This Canon f.f. semi-defector finds that I now almost never need to use HDR for high contrast scenes with my a7Rii, whereas that happened regularly with my 6D.

I also was often using the hold-your-nose work-around of MagicLantern's Dual-ISO to improve dynamic range of the 6d.

RawDigger (and FastRawViewer) are helpful to help understand what is really going on with your RAW files. You can learn a lot with their trial, and then consider purchase.
Click to expand...
Don't know much on how to read those charts, but if you just compare the points on the lines, then add Sony APS-C and m43. It's very interesting to see the (older model) Sony a6000 not really fair better than the Oly m43 cameras, but you get a lot better lenses for m43! (I still use Sony so I can have both FF and APS-C, but that was still surprising - the Sony a6300 though, pulls ahead)
 
andrewD2 said:
BBQue said:
andrewD2 said:
BBQue said:
andrewD2 said:
JimKasson said:
Midwest said:
JimKasson said:
Midwest said:
How does 14-bit (Canon) work compared to 12-bit (whoever that is/was)?
Click to expand...
I don't understand the question.

Jim
Click to expand...
To the limited extent of my understanding, unless things have changed, Canon outputs a 14-bit uncompressed RAW image while Sony (at least used to) output a compressed 12-bit RAW image. Perhaps nothing is lost in the compression but I see the opportunity where it might be, given the difference in values possible in 14-bit vs 12-bit. If so that would seem to work against the idea of smoother gradiations in color in Sony images.

And I would caution anyone against using a likely compressed JPG image website to draw IQ conclusion about something so 'fine' as these differences.
Click to expand...
14-bit uncompressed raw is available from several a7x cameras, including the a7RII. The dynamic range is unaffected by craw compression, as shown here:

http://blog.kasson.com/?p=12287

The a7x cameras switch to 12-bit precision with some shutter modes and for bulb setting and LENR.

Jim
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
 
JimKasson said:
Cameras see color differently from humans. Well, I failed right there. I'm using the word color in the technical sense.

Cameras see color differently from each other.

Commercial raw converters perform camera-specific color corrections.

Together, both of the two preceding sentences make the colors produced by raw converters look different from camera to camera.

Four one-sentence paragraphs. One technical term. How'd I do?

Jim

Jim
Click to expand...
Very well IMHO, except I do not understand which Jim is actually Joe ;)
 
Iliah Borg said:
Convert several images from both brands (some studio scene, DPR or IR) with the same gamma and without a colour transform. See if you can reliably pick an image from a particular brand based on features other than noise.
Click to expand...
 
Rol Lei Nut said:
Tony Beach said:
chironNYC said:
If one can't see the differences, why would anyone care which camera they were using?
Click to expand...
  • Lenses
  • Features
  • Performance
  • Fanboyism
Click to expand...
+
  • Ergonomics & handling
  • Habit
  • => MARKETING!
Click to expand...
Of course there are other features that mtter. We were talking about whether the sensor matters. It is just wrong to say that the Sony and the Canon sensors don't produce different images. But you are welcome to that view.
 
chironNYC said:
Tony Beach said:
chironNYC said:
If one can't see the differences, why would anyone care which camera they were using?
Click to expand...
  • Lenses
  • Features
  • Performance
  • Fanboyism
Click to expand...
Of course there are other features that mtter. We were talking about whether the sensor matters.
Click to expand...
Re-read your post. If you don't like my answer, next time be more careful what you write. I was deliberate in putting lenses at the top of my list, but I could have put performance ahead of features. Sensors can be a reason, but it's really not as huge a reason as many make it out to be, and many just don't bother to figure out where the differences in colors are coming from which is not from the sensors.
chironNYC said:
It is just wrong to say that the Sony and the Canon sensors don't produce different images.
Click to expand...
You referred to colors, so prove that.
chironNYC said:
But you are welcome to that view.
Click to expand...
You are welcome to change my view, if you can, but I fully expect you to not even try.
 
chironNYC said:
Rol Lei Nut said:
Tony Beach said:
chironNYC said:
If one can't see the differences, why would anyone care which camera they were using?
Click to expand...
  • Lenses
  • Features
  • Performance
  • Fanboyism
Click to expand...
+
  • Ergonomics & handling
  • Habit
  • => MARKETING!
Click to expand...
It is just wrong to say that the Sony and the Canon sensors don't produce different images.
Click to expand...
Once again: Very few - if anyone at all - said that. Different brands of sensors do indeed produce different output in the sense of the RAW data (which is not the same as the image). But the sensor output is not the final image.

What I and others have said is that whatever differencies in sensor output will be overridden by a lot of other variables which also determines the final image, so the difference due to different sensor brands won't be visible in the final image.

I can't say that the sensor didn't play a role at all when I bought my A7S, because the good high ISO properties (along with the ability to use adapted Leica lenses) was important for me. But as far as colors, contrast, gamma and all the rest goes, I didn't mind because I always tweak that to my liking in PP - including making my own presets for the camera in question which are automatically applied to the files upon import in my PP software. The aim is of course to "equalize" the look of my pictures no matter the camera they were taken with which I think is what most people will want. I want pictures to look "my way", not the camera designers way.

If sensor differencies are to be discussed, it will make more sense to look at differencies based on low light sesitivity, CCD vs. CMOS vs. Foveon, FF vs. APS-C, pixel count etc. than brand names.

--
"Good photos is not about fancy gear. It's about how you see as a photographer and about being ressourceful with whatever gear you have." (Alfred Eisenstadt, Life photographer)
 
Last edited:
Tony Beach said:
chironNYC said:
Tony Beach said:
chironNYC said:
If one can't see the differences, why would anyone care which camera they were using?
Click to expand...
  • Lenses
  • Features
  • Performance
  • Fanboyism
Click to expand...
Of course there are other features that mtter. We were talking about whether the sensor matters.
Click to expand...
Re-read your post. If you don't like my answer, next time be more careful what you write. I was deliberate in putting lenses at the top of my list, but I could have put performance ahead of features. Sensors can be a reason, but it's really not as huge a reason as many make it out to be, and many just don't bother to figure out where the differences in colors are coming from which is not from the sensors.
Click to expand...
Sure Tony, but aren't you arguing with the Original Poster about the topic of their thread which is specifically about sensors?
 
bclaff said:
Tony Beach said:
chironNYC said:
Tony Beach said:
chironNYC said:
If one can't see the differences, why would anyone care which camera they were using?
Click to expand...
  • Lenses
  • Features
  • Performance
  • Fanboyism
Click to expand...
Of course there are other features that mtter. We were talking about whether the sensor matters.
Click to expand...
Re-read your post. If you don't like my answer, next time be more careful what you write. I was deliberate in putting lenses at the top of my list, but I could have put performance ahead of features. Sensors can be a reason, but it's really not as huge a reason as many make it out to be, and many just don't bother to figure out where the differences in colors are coming from which is not from the sensors.
Click to expand...
Sure Tony, but aren't you arguing with the Original Poster about the topic of their thread which is specifically about sensors?
Click to expand...
I responded to the post I responded to and not the original post. That said, the OP prefaced the entire discussion with, "I am neither a scientist nor engineer, nor a professional photographer, so there is a lot about the technical aspects of digital imaging that I have only a vague understanding of." That being the case, I strongly doubt there is a worthwhile debate to be had here.
 
You must log in or register to reply here.

Similar threads

M
Image Sensor dynamics ~ 20 stops !!
2
Replies
38
Views
669
J A C S
J A C S
J
Clear Image Zoom vs 2x Converter
2
Replies
20
Views
119
Juanito123
J
Dieseltijger
Endlessly Hesitating… Thinking About Upgrading from the A7RII to the A7RIV or A7RV? Here’s What I Fo
2
Replies
34
Views
75
LarryRC
LarryRC
Batdude
Fujifilm Vs. Canon Image Quality
Replies
13
Views
125
StrugglingforLight
S
joger
can's buy anything at SONY now - cash is piling up ( would spend immediately )
3 4 5
Replies
87
Views
196
joger
joger

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top