Canon R3 (and Fuji X2HS) RAW buffer tests with CFE

[FormerFearlessPhotog]

Ok so I'm not claiming these tests are an exact science, but I do feel like they represent what to expect in real world shooting conditions fairly well. Both cameras were set to manual with a shutter speed of 1/3200, f/5.6 and auto ISO, uncompressed RAW, which represents how I would generally shoot fast moving wildlife in the real world. The only difference is they were set to manual focus to prevent the cameras from slowing while trying to refocus. Shutter button was held until the camera stopped shooting or slowed. Both cameras set to their respective maximum framerate with ES. The cards shown in the photos are the specific models from each brand, obviously the performance doesn't reflect every card from the same brand. For example the ProGrade card is an older Gold series model.

My conclusion is that all these cards will provide a pretty similar experience in the R3, the difference in write speed, which I was not set up to measure, is probably more significant than the difference in files buffered. By my own observation, the slowest CFE cards, the Angel Bird and ProGrade, took about 3x as long to write the files vs the Delkin and Sony.

The biggest difference between the R3 and the X2HS in these tests seems to be that the R3 will fill the buffer before starting to write files, and the X2HS will write files while the buffer is filling, resulting in effectively unlimited buffer with the fastest cards. The X2HS also had a more dramatic drop in performance when using an SD card. Hopefully this is helpful or at least interesting to someone.

Delkin Black R3: 158 X2HS: 200/581? (Never hit a hard buffer limit, stopped at 581 on second try with camera still shooting)

Sony Tough R3: 149 X2HS: 185 (didn't stop shooting but slowed)

Angelbird R3: 143 X2HS: 143

SanDisk R3: 148 X2HS: 140

Ritz Gear R3: 156 X2HS: 148

ProGrade R3:155 X2HS: 132

Sony Tough SD R3: 148 X2HS: 111

-
Former user of this account. https://www.dpreview.com/members/9243307581/overview

 

[rogalski]

Really interesting test. It’s great that you took the time to prepare it. You wrote that both cameras were set to their respective maximum frame rates with electronic shutter so 30fps for Canon and 40fps for Fuji. I'm curious how the test would have turned out if both were set to 30fps.

I think Fuji might have performed even better, because, as you mentioned, it continues to write files to the card even while the buffer is filling, so reducing by 10fps would likely allow for a significantly longer burst of shots.

I myself as a X-H2S user very rarely use 40fps. Most of the time I shoot at 30fps or 20fps.

 

[FormerFearlessPhotog]

You wrote that both cameras were set to their respective maximum frame rates with electronic shutter so 30fps for Canon and 40fps for Fuji. I'm curious how the test would have turned out if both were set to 30fps

I appreciate the comments, and you make a good point. My initial goal was to test both cameras at their highest performance setting, so 40fps is the most demanding in terms of write speed, but using 30fps would make it a more apples to apples comparison.

When I initially tried using the Fuji with wildlife I used it at 30fps as well because I was also interested in how the AF performed at the same framerate as the R3. I do appreciate that it has a wider selection of options for maximum framerate vs the either 30 or 15 of the R3.

I could do a retest of the Fuji at 30 over the weekend, and it will probably confirm your thoughts. It seems like with a fast enough card the camera effectively does not have a buffer limit. Maybe it would also close the gap in performance when using an SD card.

 

[M4Mobility]

I don't think any Canon cameras use uncompressed Raw.

 

[koenkooi]

I don't think any Canon cameras use uncompressed Raw.

The 'RAW' setting is, for RF bodies, lossless compressed RAW. You can opt for lossy compression by choosing CRAW, which is virtually indistinguishable from non-CRAW, unless you need to lift shadows by 5 stops.

Personally, I can only see a difference between RAW and CRAW for extreme low light shots at night. Running them to DxO PR4, increasing exposure and bumping the shadows a bit more, the CRAW show a bit less detail. That's that at ISO25600.

For all other situation, I use CRAW and enjoy the buffer that now lasts twice as long

 

[FormerFearlessPhotog]

I don't think any Canon cameras use uncompressed Raw.

Yes you're right, I'm referring to the lossless compressed files vs the dedicated CRAW option, since it's a distinction that will have a significant impact on the amount of files buffered.