It almost get 4 more shots (RAW) before it slow down, even though after that the speed is about same. (which I cannot explain)
A few interesting observations concerning the interesting results of your test.
1. Formatting the card seems to affect how many images can be written to it during the faster buffer-not-full period near the beginning of the burst.
2. Formatting does not appear to affect the speed at which images are written to the card in either case.
3. In a 30-second test such as that used in this thread, your formatted card would appear to write "more images in 30 seconds" though its pre-buffer-full speed doesn't improve nor does its post-buffer-full speed improve. It is entirely, in this case, a matter of effective buffer depth.
We will see tomorrow when I do a 30 sec test with cards formatted this way. I don't belive there will be any differnce. But we will see. In my older 1 series bodies they always had normal formatting and low level formatting for SD cards. But that didn't make any difference.
According to Transcend this formatting tool is just a way to format and recover cards that the camera and PC can't read.
gdanmitchell wrote:
A few interesting observations concerning the interesting results of your test.
1. Formatting the card seems to affect how many images can be written to it during the faster buffer-not-full period near the beginning of the burst.
2. Formatting does not appear to affect the speed at which images are written to the card in either case.
3. In a 30-second test such as that used in this thread, your formatted card would appear to write "more images in 30 seconds" though its pre-buffer-full speed doesn't improve nor does its post-buffer-full speed improve. It is entirely, in this case, a matter of effective buffer depth.
Lars Johnsson wrote:
We will see tomorrow when I do a 30 sec test with cards formatted this way. I don't belive there will be any differnce. But we will see. In my older 1 series bodies they always had normal formatting and low level formatting for SD cards. But that didn't make any difference.
According to Transcend this formatting tool is just a way to format and recover cards that the camera and PC can't read.
Interesting thread going on here, thanks for taking the time to compile those results. My curiosity has always been, "just how does a 1-series camera RAW burst behave once it slows down due to the buffer being filled?"...it is the one answer that the numbers just don't reveal. Now, Rick Fisher's calculations begin to shed some light on this camera behavior, but short of actually seeing and hearing the 1Dx in action at that point, one is left to wonder just what the camera can do for them, speed-wise, at that point.
Well, my curiosity got the best of me and I went searching online to see if anyone had posted some field results showing the 1Dx's behavior at full-buffer. I found these three videos (they're short, less than 30-seconds each!) shooting with the 1Dx with the best-performing card on your list, the Lexar 1000x...
So, my question is, "are these clips accurately portraying the 1Dx's slow-down behavior when the buffer becomes full?"...in other words, "initial long burst to buffer fill, pause, short burst, pause, short burst, pause, short burst, pause, short burst, etc"? Rick's calculations led me to believe the camera was giving an approximately steady (constant) 4fps at buffer saturation.
Jimmy:
Yes that's how it sounds when the buffer is full or nearly full with a very fast card. With a slow card it's a lot worse. And the camera maybe only take one shot every second. Or one shot in two seconds.
I downloaded and formatted more than ten cards with the Transcend Formatting Tool today. First I formatted them in my 1DX and shoot a 30 sec test so I could compare. It didn't make any difference for speed or number of shots. But while doing that test I discovered why the Transcend 32 GB UDMA 7 400x card had such a low number compared to the 64GB version of it. I have six of those cards. And one is a bit slower than the other five cards. So I tried all of those cards and then updated the number in my test to the number I got with the other five cards. It's still slower than the 64GB version but not like before.
Most of the time the larger version of a card is faster than the same small sized card when I have tested cards.
Lars Johnsson wrote:
I downloaded and formatted more than ten cards with the Transcend Formatting Tool today. First I formatted them in my 1DX and shoot a 30 sec test so I could compare. It didn't make any difference for speed or number of shots. But while doing that test I discovered why the Transcend 32 GB UDMA 7 400x card had such a low number compared to the 64GB version of it. I have six of those cards. And one is a bit slower than the other five cards. So I tried all of those cards and then updated the number in my test to the number I got with the other five cards. It's still slower than the 64GB version but not like before.
Most of the time the larger version of a card is faster than the same small sized card when I have tested cards....Show more →
Thank you so much for your test. I guess there is no magic bullet there.
For that test I got:
Lexar 1000x 32GB ex FAT - 116.96MB/s READ, 91.77 MB/s WRITE
Hoodman Steel 1000x 64GB exFAT - 109.67MB/s READ, 88.24MB/s WRITE
For a different Lexar 1000x 32GB (newest) FAT32 - 116.64MB/s, 87.76MB/s
And that same card as exFAT - 115.18MB/s,89.18MB/s
And same but test repeated - 115.20MB/s, 92.41MB/s
And repeated again - 115.22MB/s,89.78MB/s
So write speed tends to vary a bit test to test with CF.
Another new Lexar 1000x 32GB exFAT: 116-120.36MB/s READ (only first test was slow on read), 88-93.60MB/s WRITE
Yeah it seems like the very first test is always a touch slower, other tests vary, mostly for writes, but tend to all be at least a touch faster than first test.
(For curiosity:
For my Buffalo Drive over USB 3.0 I got: 190.82MB/s, 128.70MB/s
For my main C: drive I got: 70.54MB/s, 69.32MB/s
For my alternate C: drive I got: 93.65MB/s, 80.79MB/s
which is curious since Vista is on the alternate drive and it always behaves like the HD is slower than a sloth and W7 on the other drives acts like the HD is much faster
For a regular external drive over USB 3.0 I got: 97.30MB/s,90.45MB/s
an internal photo HD: 60.27MB/s,57.30MB/s)
ML reports 116MB/s for Lexar 1000x 32GB and Hoodman Steel 1000x 64GB write speed. I don't think that speed can be maintained for realistic situations though.
For me: benchmark tests are not interesting. It doesn't say much about the performace when you put the card in your camera. And it doesn't say much about the speed when downloading the pics from a card reader either.
Even if the card have 100+mb/s in the test, it could have 20mb/s in your camera and card-reader
