The Canon EOS R3 will be officially announced in September

Jan 29, 2011
10,675
6,121
  • Like
Reactions: 1 user
Upvote 0
Sep 20, 2020
3,052
2,383
Photographic dynamic range, it is differentiated from engineering dynamic range in that it establishes a noise floor limit, SNR or signal to noise ratio, that is corrected for the sensor size. This means it gives a closer to ‘real world’ value.

My big problem with this as a measure is that modern cameras denoise in camera.
In the real world we also denoise in post.
Also not all noise looks the same.
Filmic noise might be perfectly acceptable.

Maybe I am just old fashion but it seems a lot easier to just try cameras out and pick the ones that suit me.
 
Upvote 0
Jan 29, 2011
10,675
6,121
My big problem with this as a measure is that modern cameras denoise in camera.
In the real world we also denoise in post.
Also not all noise looks the same.
Filmic noise might be perfectly acceptable.

Maybe I am just old fashion but it seems a lot easier to just try cameras out and pick the ones that suit me.
Oh there is no doubt that personal testing or use is the only way to work out if a specific camera will work for you personally or not. I also agree it is frustrating that almost everybody now cooks their RAW files, something I think Canon was pushed into because others did it and those comparisons hurt and people don’t understand they are not true comparisons.

But I still find the measures useful for an initial assessment and have found the Bill Claff site to be unbiased and a good indication of what I’d expect to see from real files.

I also like that you can download all the ‘Studio Comparison’ RAW files from DPReview.
 
  • Like
Reactions: 1 user
Upvote 0

entoman

wildlife photography
May 8, 2015
1,998
2,438
UK
Those advantages were more apparent five years ago or so. In real world use today the 45 mp sensor performs very well at high ISOs. At the same time the 1Dx III sensor’s 20mp sustains cropping much better than previous generations. We are fast approaching the point where the old assumptions are relevant only for theoretical debates on geek forums.
That does appear to be the case, although most photography websites are still perpetuating the notion that low resolution sensors (of same generation) have less noise and greater DR...
 
Upvote 0

jd7

CR Pro
Feb 3, 2013
1,059
415
If budget is an issue, maybe going for an Apple or BMW isn't the best idea in the first place.

I understand that the rise in prices is upsetting, but it is to be expected. The market is shrinking, so seeing such a great pace of progress both with the performance of the latest bodies and the glass that goes along with it is something that should be appreciated in itself. But as that R&D has to be payed for by fewer customers and the current world economy causes a bunch of issues from production, over logistics to inflation, a new price structure is inevitable. There's no point in complaining about a corporation wanting to stay in business.
There is no point complaining about a corporation wanting to stay in business, certainly, but I don't understand why the price rises are to be expected (or at least price rises which do more than keep pace with inflation and currency fluctuations). Assuming Canon's R&D budget hasn't suddenly increased dramatically, and that cost of production of a mirrorless camera is similar to that of a DSLR, what's changed? There is a potentially shrinking market, but is there any evidence that the market for higher end cameras is shrinking? Was R&D cost previously shared over a wider range of gear (ie compact cameras, in particular) than now? I don't know to what extent R&D cost was shared, or is shared now (and not just with cameras specifially, but potentially with other things Canon makes in its various divisions), but it does seem possible this is an issue. On the other hand, the more Canon pushes up price, the more the market will likely shrink as people decide it's not worth the price to play. Pushing cameras into the realm of real luxury items which only a relatively small number of people are willing to pay for (which seems to me to be where Canon is heading) doesn't seem obviously a great strategy to me, but I can only assume Canon thinks (based on all of the market data it has, no doubt) that the increased profit per item will more than make up for the lower sales volume.

Anyway, the thing which really gets me is that Sony's camera bodies aren't more expensive than Canon's now, and once you are in the Sony system there are a lot of great value lenses (albeit many are third party) which make the overall cost of ownership pretty similar to what it's been for DSLRs. So, it is possible to have current tech at a much more moderate cost than you can get it with Canon.
 
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,717
1,532
Yorkshire, England
Oh there is no doubt that personal testing or use is the only way to work out if a specific camera will work for you personally or not. I also agree it is frustrating that almost everybody now cooks their RAW files, something I think Canon was pushed into because others did it and those comparisons hurt and people don’t understand they are not true comparisons.

But I still find the measures useful for an initial assessment and have found the Bill Claff site .......
I was looking for information regarding raw files being cooked, curious because of what Canon is doing with the R5, and somehow came across the DPR forums where Bill Claff is quite active. The discussion was around Canon's use of noise reduction in the raw R5 files at low ISO. Somebody asked Bill if he was aware of other camera manufacturers doing the same thing as Canon is with the R5 (and 1DXIII), adding noise reduction to the raw files. Bill's answer was that from the data he analyses he wasn't aware of any other camera manufacturer doing his. Now this really surprised me because I was / am pretty certain when looking at other raw files, that they are !
He was also asked how much noise reduction was being applied on the R5 and his reply was that is was reasonably significant but possibly only in the deep lowlights.
So I find this a bit of a mystery. Is Bill Claff right, and the others have unadulterated raws ? I can't believe it. Or is it limitations in his methodology that is not showing it ? I don't know.
Incidentally when I downloaded some R5 raws from somewhere, can't remember where, and really pulled up the deep shadows to an unrealistic and academic level, they didn't look too pretty to be honest. Certainly saw nothing that made me want to give up the 5DS.
 
Upvote 0
Jan 29, 2011
10,675
6,121
I was looking for information regarding raw files being cooked, curious because of what Canon is doing with the R5, and somehow came across the DPR forums where Bill Claff is quite active. The discussion was around Canon's use of noise reduction in the raw R5 files at low ISO. Somebody asked Bill if he was aware of other camera manufacturers doing the same thing as Canon is with the R5 (and 1DXIII), adding noise reduction to the raw files. Bill's answer was that from the data he analyses he wasn't aware of any other camera manufacturer doing his. Now this really surprised me because I was / am pretty certain when looking at other raw files, that they are !
He was also asked how much noise reduction was being applied on the R5 and his reply was that is was reasonably significant but possibly only in the deep lowlights.
So I find this a bit of a mystery. Is Bill Claff right, and the others have unadulterated raws ? I can't believe it. Or is it limitations in his methodology that is not showing it ? I don't know.
Incidentally when I downloaded some R5 raws from somewhere, can't remember where, and really pulled up the deep shadows to an unrealistic and academic level, they didn't look too pretty to be honest. Certainly saw nothing that made me want to give up the 5DS.
I think Canon are the manufacturer that cooks (uses built in noise reduction) in its RAW files at low iso numbers. Almost all manufacturers cook their high iso RAW files as can be seen by Bill's own graphs. Triangle down or diamond shape indicate noise reduction pre applied. But as you say, I wouldn't rule out limitations in his calculation strategy.


Interestingly the only 'new' camera I could see that doesn't use noise reduction on the RAW files anywhere in the iso range is the Sony α1.

Having said all this it really becomes a bit of an esoteric conversation. We can't 'see' the analog readings coming off the sensor, it is necessarily converted to digital output which we also can't 'see'. If we use programs to demosiac that information in an as is format they look horribly dark and green anyway. So given RAW files have to undergo a regeneration process one wonders what difference pre baked noise reduction makes. Then, of course, we go down the rabbit hole of baked in lens corrections etc etc...

Personally I don't have a problem with baled in NR, I do have a problem with baked in lens corrections. To me the first is something that is going to be done anyway, the later is a way of working around poor lens design.
 
  • Love
Reactions: 1 user
Upvote 0

unfocused

Photos/Photo Book Reviews: www.thecuriouseye.com
Jul 20, 2010
7,184
5,483
70
Springfield, IL
www.thecuriouseye.com
There is no point complaining about a corporation wanting to stay in business, certainly, but I don't understand why the price rises are to be expected (or at least price rises which do more than keep pace with inflation and currency fluctuations). Assuming Canon's R&D budget hasn't suddenly increased dramatically, and that cost of production of a mirrorless camera is similar to that of a DSLR, what's changed? There is a potentially shrinking market, but is there any evidence that the market for higher end cameras is shrinking? Was R&D cost previously shared over a wider range of gear (ie compact cameras, in particular) than now? I don't know to what extent R&D cost was shared, or is shared now (and not just with cameras specifially, but potentially with other things Canon makes in its various divisions), but it does seem possible this is an issue. On the other hand, the more Canon pushes up price, the more the market will likely shrink as people decide it's not worth the price to play. Pushing cameras into the realm of real luxury items which only a relatively small number of people are willing to pay for (which seems to me to be where Canon is heading) doesn't seem obviously a great strategy to me, but I can only assume Canon thinks (based on all of the market data it has, no doubt) that the increased profit per item will more than make up for the lower sales volume.

Anyway, the thing which really gets me is that Sony's camera bodies aren't more expensive than Canon's now, and once you are in the Sony system there are a lot of great value lenses (albeit many are third party) which make the overall cost of ownership pretty similar to what it's been for DSLRs. So, it is possible to have current tech at a much more moderate cost than you can get it with Canon.
First, I think it is a mistake to assume that the last year or two are typical, which they are not.

In a typical cycle, after a year on the market the R5 would have seen a substantial price drop and most of the RF lenses would also be seeing lower prices (Not necessarily official price cuts, but defacto price cuts through rebates, promotions and special deals from retailers.)

Next, I think people make the mistake of looking only at the high end products. The Canon RP is the cheapest full frame camera ever offered by Canon, well below the cost of the 6D, not only at introduction but even today. The R is a fantastic camera and before the pandemic, it was dropping in price. Nearly two years ago, I paid less for the R than it is selling for today.

The RF lens system also includes some bargains. Two supertelephotos for under $1,000 -- something previously unheard of. A full-frame 24-240mm superzoom for under $1,000. An f1.8 wide angle with macro for $100 less than its 35 f2 EF lens.

The R5 is an expensive camera, no denying that, but it is also a sophisticated camera, offering features that no previous Canon DSLR offered.

And, yes, R&D Cost was previously shared over a wider range of gear, plus DSLRs benefitted from years and years of previous embedded development.

Ultimately it comes down to supply and demand. We are in a cycle of high demand and low supply. If Sony products are cheaper or seem to be a better value, that is because the demand is less and the supply is greater. Ultimately, the market sets prices and if Canon products are not perceived by consumers to be a good value, the prices will come down.
 
  • Love
Reactions: 1 user
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,717
1,532
Yorkshire, England
I think Canon are the manufacturer that cooks (uses built in noise reduction) in its RAW files at low iso numbers. Almost all manufacturers cook their high iso RAW files as can be seen by Bill's own graphs.
When I download sample raw files from say the Nikon Z7II for instance, in the low ISO samples there is definitely a very tiny, subtle smearing and smoothing of the data and suspicious lack of shot noise compared with the equivalent Canon, which is recording a tiny (and insignificant) amount of extra detail. This makes me think that with my raw converter programs at least, there is some kind of noise reduction somewhere, somehow, maybe not intended. For instance look how noiseless and smooth at the expense of detail the Fuji X trans raw files are when opened in ACR.
 
Upvote 0
Aug 7, 2018
598
549
Very annoying is that "star eating" noise reduction in some cameras that you can't turn off. If you take a video of the night sky and nail the focus, most stars will only have the size of a single pixel. So the camera treats them as random noise and the stars suddenly disappear. They will only reappear if you slightly change the focus, so that each of them is spread over more than one pixel.
 
  • Like
Reactions: 1 user
Upvote 0
Sep 20, 2020
3,052
2,383
Interestingly the only 'new' camera I could see that doesn't use noise reduction on the RAW files anywhere in the iso range is the Sony α1.
I do not believe that for a second.
DPReview is dead to me.
They seem to review gear that they have no idea how to use in the manner people actually use it.
Blogs of all kinds do this and they are far from the biggest offenders.
I think they review whatever they are sent and companies need better criteria for who they send what to.
 
Upvote 0

H. Jones

Photojournalist
Aug 1, 2014
803
1,637
Personally I don't have a problem with baled in NR, I do have a problem with baked in lens corrections. To me the first is something that is going to be done anyway, the later is a way of working around poor lens design.

I do agree with you on this, but for me it more depends on the context. A cheap $200 18-55mm kit lens, or even a non-pro 24-240mm travel lens needing baked-in corrections as it doesn't cover the full sensor? Sure, why not, I'll accept it if it mens the lens is significantly smaller and easier for its intended task(travel/being super cheap). I don't feel like anyone should really expect miracles from non-L series glass, at the end of the day, so if it accomplishes the small/cheap/light/big zoom flexibility, I'm not too bothered.

My true concern is if Canon applies the same baked-in lens correction to L-series glass, like the 14-35mm, and that the lens doesn't actually cover the sensor at 14mm. That would be undercutting people who *do* care about image quality and expect the most out of their glass. Same thing goes for if they ever make a L-series version of a 24-240mm, that lens needs to be flexible, but it needs to actually be built for the focal lengths intended.
 
Upvote 0

H. Jones

Photojournalist
Aug 1, 2014
803
1,637
I thought the lens correction was just metadata in the RAW files.
It obviously needs to be baked in JPEGs.
Not for the 24-240mm, which is what PrivateByDesign was talking about. That lens doesn't cover the sensor at 24mm, so even the raw files have correction applied to appear to cover the sensor.

Edit: Looking at it again, I may me slightly mistaken. The 24-240 does cover the sensor at 24mm, but is actually closer to a 18-20mm fisheye look at 24mm, which is corrected even in the raw files to appear as 24mm.
 
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,543
4,082
The Netherlands
Not for the 24-240mm, which is what PrivateByDesign was talking about. That lens doesn't cover the sensor at 24mm, so even the raw files have correction applied to appear to cover the sensor.

Edit: Looking at it again, I may me slightly mistaken. The 24-240 does cover the sensor at 24mm, but is actually closer to a 18-20mm fisheye look at 24mm, which is corrected even in the raw files to appear as 24mm.
From what I've seen it isn't lens corrected in the RAW files, open one up with a tool like RawDigger and you should see un-lens-corrected data.
 
Upvote 0

H. Jones

Photojournalist
Aug 1, 2014
803
1,637
From what I've seen it isn't lens corrected in the RAW files, open one up with a tool like RawDigger and you should see un-lens-corrected data.

Yeah, you're right about that. I'm sure this is a conversation that could go on for hours, but I think the bigger point was that the lens isn't actually a 24mm lens unless it's corrected, and that the 24mm uncorrected is basically unusable. The lens would have been designed differently in the film era, where what you see was all you had, which meant lenses had to be (relatively) more corrected on their own as they couldn't depend on correction in the same way.

I know other lenses have extreme distortion, but the 24-240mm is a pretty good example of where Canon chose to emulate 24mm at 24mm using digital corrections rather than design the lens that way from the get-go.
 
Upvote 0