5DS/R 2-3 second delay in writng to CF card

I'm using a Lexar 64 MB 1066X card, brand new, but this problem happens with an older 1066X as well. With RAW, between shutter button push and display of image on the LCD, there is a 2-3 second delay. Actually, the first shot in any series is 2 seconds, the next 3 seconds, and it alternates that way for as long as I shoot photos. I timed this several times with a stop watch. The delay is less for small JPEG, but still 1-2 seconds.

Yes, bigger files 75 MB vs 25 MB with my 5D3, but the 5D3 gives me instantaneous LCD images. With the S/R, it's a pain to sit there and wait. The clouds or light in a landscape will change in that time if I want to adjust exposure, composition, etc., and retake the shot. I suspect there would be even more cloud movement in a 5 shot pano series. And it's just one more delay when I'm trying to speed workflow. I think the same would be true in a studio.

Do you find the same delay in your S/R? If not, I think I may have a bad copy and will exchange it.

I've read about this. To be fair, its writing tremendous files to the card.

Frodge wrote in post #17665447
I've read about this. To be fair, its writing tremendous files to the card.

hi frodge, may i ask where you read this?

I believe I read it on this forum. I believe David Arbogast was talking about it.

Frodge wrote in post #17665447
I've read about this. To be fair, its writing tremendous files to the card.

Perhaps fair to Canon, but from my point of view, why didn't they speed up the buffer? I'm still asking how many owners show the same delay?

Frodge wrote in post #17665477
I believe I read it on this forum. I believe David Arbogast was talking about it.

Xyclopx was the first to post about it in the 5DSR announcement thread, then David and I posted. But that's still just two cameras so far. If it's standard function for this camera, I'd like to see other reports of it.

tvphotog wrote in post #17665392
I'm using a Lexar 64 MB 1066X card, brand new, but this problem happens with an older 1066X as well. With RAW, between shutter button push and display of image on the LCD, there is a 2-3 second delay. Actually, the first shot in any series is 2 seconds, the next 3 seconds, and it alternates that way for as long as I shoot photos. I timed this several times with a stop watch. The delay is less for small JPEG, but still 1-2 seconds.

Yes, bigger files 75 MB vs 25 MB with my 5D3, but the 5D3 gives me instantaneous LCD images. With the S/R, it's a pain to sit there and wait. The clouds or light in a landscape will change in that time if I want to adjust exposure, composition, etc., and retake the shot. I suspect there would be even more cloud movement in a 5 shot pano series. And it's just one more delay when I'm trying to speed workflow. I think the same would be true in a studio.

Do you find the same delay in your S/R? If not, I think I may have a bad copy and will exchange it.

I've written about this many times already - even wrote a review on the camera here on day 2 of the release stating this. It is what it is. They should have upgraded the buffer - they didn't. It is 'a' reason why you might not use it as your main camera at a wedding.

If you think that it is bad, if you try and review 'zoomed-in' files on the Sony A7Rii, it is 5x worse.

huge files or not, Canon should have provided enough horsepower for the camera to handle the file size.
Do medium format cameras have this issue? Is this a "comes with the territory" issue or did Canon drop the ball, so to speak?

Fair might not have been the correct way to phrase it. Seems like the camera was rushed to production. Canon IMHO is cutting many corners lately and rushing their camerad out to production. They are huge files, but I do agree the buffer should be able to handle it.

Frodge wrote in post #17665542
Fair might not have been the correct way to phrase it. Seems like the camera was rushed to production. Canon IMHO is cutting many corners lately and rushing their camerad out to production. They are huge files, but I do agree the buffer should be able to handle it.

Is there a firmware fix for a buffer, or is it a physical device?

Something appears quite different than the review at http://www.the-digital-picture.com/Reviews/Ca​non-EOS-5Ds.aspx reports. Once there scroll down to the section titled Frame Rate, Buffer Depth, Shutter Sound

John from PA wrote in post #17665600
Something appears quite different than the review at http://www.the-digital-picture.com/Reviews/Ca​non-EOS-5Ds.aspx reports. Once there scroll down to the section titled Frame Rate, Buffer Depth, Shutter Sound

With due respect, I can't relate the fps data to the delay in showing an LCD image after snapping the shutter once. No matter what Bryan writes, that time is still 2-3 seconds.

I called Canon yesterday about this issue. The tech I spoke to pulled an S/R from their shelf, put in a fast card, and got the same results as I did. She tried it with RAW in the 5D3 and the image appeared immediately, as was the case with my 5D3. She was surprised that the S/R was so slow despite the new processor, though she said she wasn't aware of anyone else who had called to ask about it.

John from PA wrote in post #17665600
Something appears quite different than the review at http://www.the-digital-picture.com/Reviews/Ca​non-EOS-5Ds.aspx reports. Once there scroll down to the section titled Frame Rate, Buffer Depth, Shutter Sound

so, frame rate is not the same thing as what we are talking about here. i just did some tests between my 5d3 and 5dsr side by side....

for the tests, both cameras on M with shutter speed > 1/200th, ISO 6400, different apertures (different lenses), manual focus, IS off

5d3 = SanDisk Extreme Pro 128gb 160mb/s
5dsr = Lexar Professional 64gb 1066x
(i believe these are similar speeds)

Frame Rate, Buffer:

1. put both cameras in high speed mode. held down shutter. the rate does seem to be as advertised: 5d3 = ~6 frames/s, 5dsr ~5. I didn't count the buffer size, but with both cameras in RAW only, at max res, they both easily took > 10 frames, probably closer to 20 before I saw the speed slow. so, I would say there's no frame rate problem.

Time Until First Preview (what we were talking about):

2. JPG only, high res: both cameras highest res JPG. the 5d3 was about ~1s. the 5dsr took a a little over 2s, let's say ~2.25s.
3. JPG, low res: with both cameras at lowest res JPG, the 5d3 = a little bit less than ~1s, let's say ~0.75s. the 5dsr = ~1.25s

4. RAW only, high res: similar results to #2. maybe a tad faster?
5. RAW, low res: similar to #3. maybe the 5dsr was a little slower than with JPG, ~1.75s.

6. Max res RAW + max res JPG: the 5d3 = ~1s, the 5dsr = ~2s.

Now............. yeah, there's some error here, and I didn't time these with a stop watch.

My conclusions:

1. Combining RAW and JPG vs just RAW didn't make an appreciable difference. (maybe because the bottle neck is making the JPG preview, which has to be done either way?)
2. The 5dsr is always slower than the 5d3, even at the same MP settings. I didn't write the test results, but I did test with matching MPs, and the 5dsr was always about 2x as slow.
3. The 5d3 is not as fast as everyone perceives. It still takes 1s for the preview to show up, and not really "instant."
4. The 5dsr is about 2x as slow as the 5d3. this is generally true at all similar settings.
5. Resulting file size doesn't affect speed as much as expected

My hypothesis:

I'm just completely guessing, but perhaps the bottle neck is that for some reason the implementation must process that massive amount of data from the sensor, and the bottleneck isn't really the buffer or card bus. And maybe just the sheer amount of data originating from the sensor is causing some havoc internally. Otherwise have no clue why the 5dsr has to be 2x as slow as the 5ds when previewing the same res results.

Xyclopx wrote in post #17665637
I'm just completely guessing, but perhaps the bottle neck is that for some reason the implementation must process that massive amount of data from the sensor, and the bottleneck isn't really the buffer or card bus. And maybe just the sheer amount of data originating from the sensor is causing some havoc internally. Otherwise have no clue why the 5dsr has to be 2x as slow as the 5ds when previewing the same res results.

With mine, I clearly get a 2 second lag alternating with a 3 second lag, uniformly, with a stop watch. And I thought the new processor was supposed to handle the large files faster, I thought that was part of the hype about its speed.

John from PA wrote in post #17665600
Something appears quite different than the review at http://www.the-digital-picture.com/Reviews/Ca​non-EOS-5Ds.aspx reports. Once there scroll down to the section titled Frame Rate, Buffer Depth, Shutter Sound

I just received a response to an email I sent to Bryan Carnathan, knowing that he's shooting with two 5DS (or S/R's, I don't remember). He says that he tested his cameras after getting my email, and finds the same 2 second lag.

Use a sd card and it seems to speed it up...not sure if its a hardware issue of if a firmware can fix it.

Watch this vid and report back if it does make a differences or not.

http://www.youtube.com​/watch?v=rMpxa4VMJrI

if the sd card is faster then it does not make any sense as cf cards are usually much faster. Well at least they is a workaround. Remember if it does work for you do report back here so that other 5ds/r owner will be aware of this issue.