DPReview.com is closing April 10th - Find out more

Canon EOS R3 at PhotonsToPhotos

Started Dec 10, 2021 | Discussions
bclaff Forum Pro • Posts: 13,922
Canon EOS R3 at PhotonsToPhotos
28

The analysis took a bit longer than I had hoped due to a variety of issues.

For Photographic Dynamic Range (PDR) the R3 looks substantially like the R5

The Electronic Shutter (ES) penalty is much lower than that for the R5:

It looks like some amount of signal processing (amounting to noise reduction) occurs at all ISO settings. I anticipate a long technical debate about this but I suspect there is little or no practical consequence.

The amount at base ISO (100) looks comparable to the R5

Canon EOS R5

Canon EOS R3

When there is no signal processing these curves are flat rather than dipping in the middle.
Not shown but these curves do flatten for the R5 at higher ISOs but stayed bowed at all ISO settings for the R3.

FWIW, here's the R at ISO 100

Canon EOS R.

Note it is essentially flat dipping slightly on the right hand side.

-- hide signature --

Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

Canon EOS R3 Canon EOS R5
If you believe there are incorrect tags, please send us this post using our feedback form.
Hoka Hey
Hoka Hey Senior Member • Posts: 2,991
Re: Canon EOS R3 at PhotonsToPhotos

As always, thanks Bill.

-- hide signature --

Joe

dmanthree
dmanthree Forum Pro • Posts: 10,302
Re: Canon EOS R3 at PhotonsToPhotos

Thanks! Good work.

Looks pretty good.

-- hide signature --

---on the cutting edge---

buellom Contributing Member • Posts: 800
Re: Canon EOS R3 at PhotonsToPhotos
1

So this means Canon is cooking their raw files of the R6,5,3?

what does the dipping mean? At the sides the lines are higher (compared to the R), then there is this dipping... What does this technically mean? (For someone who knows not much about the underlying technology...)

-- hide signature --

********************
www.freude-am-licht.de
********************

 buellom's gear list:buellom's gear list
Canon EOS R Canon EOS R5 Canon RF 24-105mm F4L IS USM Canon RF 100-500mm F4.5-7.1L IS USM Canon RF 14-35mm F4L IS USM
PicPocket Veteran Member • Posts: 5,897
Re: Canon EOS R3 at PhotonsToPhotos
1

buellom wrote:

So this means Canon is cooking their raw files of the R6,5,3?

I anticipate a long technical debate about this but I suspect there is little or no practical consequence

what does the dipping mean? At the sides the lines are higher (compared to the R), then there is this dipping... What does this technically mean? (For someone who knows not much about the underlying technology...)

If you are really interested

https://www.photonstophotos.net/GeneralTopics/Sensors_&_Raw/Sensor_Analysis_Primer/An_Introduction_to_Energy_Spectra_for_Sensor_Analysis.htm

It isn't about underlying technology in use, but more like basic theory of signal processing. If you don't want to venture into that territory, this may not mean much for you. The above highlighted summarization / interpretation is a better place to start

-- hide signature --
 PicPocket's gear list:PicPocket's gear list
Canon EOS R5 Canon EF 24-70mm F2.8L II USM Sigma 135mm F1.8 Art Canon RF 70-200mm F2.8L IS USM Canon RF 100-500mm F4.5-7.1L IS USM +16 more
Steve Balcombe Forum Pro • Posts: 15,571
Re: Canon EOS R3 at PhotonsToPhotos

Great work as always Bill, thank you!

Interesting result re the ES.

Also interesting to me is the PDR comparison between the R3 and the R5 in the 'normal' shutter modes. Below ISO 400 they are essentially identical. From 400 to 1000 (the range I use by far the most often) the R5 actually wins by a small amount! But above that the R3 wins by an equally small amount.

I'm not in the market for an R3, but studying it closely for clues as to what might trickle down to an R5 Mk II.

Ephemeris
Ephemeris Senior Member • Posts: 1,186
Re: Canon EOS R3 at PhotonsToPhotos

The Electronic Shutter (ES) penalty is much lower than that for the R5:

Is this true? Ignoring <ISO 100 then I can see some variation between 100 and 400. As we move towards 400 the difference reduces.

>400 the differences look small?

Salgado Photo Regular Member • Posts: 377
Re: Canon EOS R3 at PhotonsToPhotos

Bill can you explain how to use this chart to make decisions in the field regarding ISO settings?

Also why is there a dip in the chart between 200 and 400? Very interesting stuff thanks!

 Salgado Photo's gear list:Salgado Photo's gear list
Sony RX100 II Nikon D700 Canon EOS R6
higheronymous
higheronymous Regular Member • Posts: 271
Re: Canon EOS R3 at PhotonsToPhotos
1

buellom wrote:

So this means Canon is cooking their raw files of the R6,5,3?

what does the dipping mean? At the sides the lines are higher (compared to the R), then there is this dipping... What does this technically mean? (For someone who knows not much about the underlying technology...)

Whatever Canon is doing we don't really know as outsiders looking in. There are a number of steps between image capture on the sensor and saving a RAW to the card.

Ephemeris
Ephemeris Senior Member • Posts: 1,186
Re: Canon EOS R3 at PhotonsToPhotos

All RAW data is processed if that's what you mean by cooking.

The jumps are related to the specific gain at those ISO settings.

The chart, as I understand is an indication (nothing more) of the effect to shadow noise as we move through the ISO range. The thoughts are move one stop in ISO and highlights compress one stop. If we have an almost flat line on this chart it means we can recover shadows post processing in the same.or similar way as

OP bclaff Forum Pro • Posts: 13,922
Re: Canon EOS R3 at PhotonsToPhotos
3

Geyzer J. Salgado wrote:

Bill can you explain how to use this chart to make decisions in the field regarding ISO settings?

This chart simply gives you some idea of there the "ISO Invariant" areas are.
Personally I don't ever make ISO setting chooices based on this type of information.

Also why is there a dip in the chart between 200 and 400? Very interesting stuff thanks!

It's a mystery. One theory would be that noise reduction is a little stronger at the ISO setting.

-- hide signature --

Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

TukTuk
TukTuk Contributing Member • Posts: 528
Re: Canon EOS R3 at PhotonsToPhotos
1

buellom wrote:

So this means Canon is cooking

not as much as Pentax in their K3III dSLR ... not there you have what can be called real cooking !

flyingskiguy Junior Member • Posts: 26
Re: Canon EOS R3 at PhotonsToPhotos

bclaff wrote:

Hi Bill, thanks for this. Are there electronic shutter results for the A1 available for comparison? Or does the DR not drop from the mechanical shutter like in Canon bodies?

 flyingskiguy's gear list:flyingskiguy's gear list
Canon EOS R5 Canon EOS R3 Canon RF 28-70mm F2L USM Canon RF 70-200mm F2.8L IS USM Canon RF 85mm F1.2L USM +3 more
OP bclaff Forum Pro • Posts: 13,922
Re: Canon EOS R3 at PhotonsToPhotos
2

flyingskiguy wrote:

bclaff wrote:

Hi Bill, thanks for this. Are there electronic shutter results for the A1 available for comparison? Or does the DR not drop from the mechanical shutter like in Canon bodies?

I only publish ES results when they differ from MS.

That said, I don't think I got Sony ILCE-1 ES measurements.

-- hide signature --

Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

John Sheehy Forum Pro • Posts: 26,688
Re: Canon EOS R3 at PhotonsToPhotos

Steve Balcombe wrote:

Below ISO 400 they are essentially identical. From 400 to 1000 (the range I use by far the most often) the R5 actually wins by a small amount! But above that the R3 wins by an equally small amount.

Triple Conversion Gain? The only other idea I have is that the headroom changes at ISO 1250, but I don't think that DxO saw that, and they actually measure absolute headroom.

-- hide signature --

Beware of correct answers to wrong questions.
John
http://www.pbase.com/image/55384958.jpg

OP bclaff Forum Pro • Posts: 13,922
Re: Canon EOS R3 at PhotonsToPhotos

John Sheehy wrote:

Steve Balcombe wrote:

Below ISO 400 they are essentially identical. From 400 to 1000 (the range I use by far the most often) the R5 actually wins by a small amount! But above that the R3 wins by an equally small amount.

Triple Conversion Gain? ...

No, probably more aggressive noise reduction which appears as lower read noise and higher dynamic range.

-- hide signature --

Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

Keyboard shortcuts:
FForum MMy threads