Lars Johnsson wrote:
But what camera are you using According to your profile you own the 40D. To get any good speed out of the newer and faster UDMA cards, you need a good and new camera that supports UDMA 6 cards.
A lot of people here buy those cards and then complain that they are not fast. And they are only fast if your body support that kind of cards
Ok, so he hasn't updated his profile, but he is the OP and is talking about the 7D.
vpexp wrote:
Stanj, comparing the GMonster to the adata 533x would only make more sense to me if I had an adata 533X. I compared what I had on hand. Didn't not feel that it was worth it to invest $ for this "critical" test.
So that people don't accuse me of not being cooperative or productive, I compared the AData 533x and the Sandisk Extreme 60MB/s. The only benchmark that I used was from burst stall til buffer completely empty. adata = 10s, SanDisk 4.5s. Ouch, and FWIW.
stanj wrote:
So that people don't accuse me of not being cooperative or productive, I compared the AData 533x and the Sandisk Extreme 60MB/s. The only benchmark that I used was from burst stall til buffer completely empty. adata = 10s, SanDisk 4.5s. Ouch, and FWIW.
Stan, what cam? (I know you shoot 1DsIII, but are those numbers for that?!?)
stanj wrote:
No, this was in 7D. The fastest thing I could get my hands on short-term.
Oh good, 7D is what interests me. So the 4.5s after "buffer stall" for the Sandisk Extreme 60MB/s... Is "buffer stall" at the end of the initial unrestricted burst, say, 16-20 or so raw-only frames depending on scene complexity? Or do you allow for a few hiccups, and define "buffer stall" a bit further on? Or is this RAW-JPEG?
stanj wrote:
So that people don't accuse me of not being cooperative or productive, I compared the AData 533x and the Sandisk Extreme 60MB/s. The only benchmark that I used was from burst stall til buffer completely empty. adata = 10s, SanDisk 4.5s. Ouch, and FWIW.
paulfeng wrote:
Stan, what cam? (I know you shoot 1DsIII, but are those numbers for that?!?)
stanj wrote:
No, this was in 7D. The fastest thing I could get my hands on short-term.
paulfeng wrote:
Oh good, 7D is what interests me. So the 4.5s after "buffer stall" for the Sandisk Extreme 60MB/s... Is "buffer stall" at the end of the initial unrestricted burst, say, 16-20 or so raw-only frames depending on scene complexity? Or do you allow for a few hiccups, and define "buffer stall" a bit further on? Or is this RAW-JPEG?
Thanks for the report.
This was raw+S JPG (coarse). There's no way I can trigger the stopwatch right on the very first hiccup, as it's 1/10s at most, then the camera fires two more shots. So the timing started when the camera really stopped (hiccup + 2 frames) and ended when the red LED turned off.
stanj wrote:
This was raw+S JPG (coarse). There's no way I can trigger the stopwatch right on the very first hiccup, as it's 1/10s at most, then the camera fires two more shots. So the timing started when the camera really stopped (hiccup + 2 frames) and ended when the red LED turned off.
So... about how many shots is this? 6+2? or better? (Sorry, don't have camera in front of me right now) I know the buffer is not as deep in RAW+JPEG, but don't recall if it gets better with S JPG (coarse) that you used.
Just trying to get a feel for how many MB/sec the card is taking. My testing has been with RAW-only bursts, given the deeper buffer for that.
stanj wrote:
So that people don't accuse me of not being cooperative or productive, I compared the AData 533x and the Sandisk Extreme 60MB/s. The only benchmark that I used was from burst stall til buffer completely empty. adata = 10s, SanDisk 4.5s. Ouch, and FWIW.
paulfeng wrote:
Stan, what cam? (I know you shoot 1DsIII, but are those numbers for that?!?)
stanj wrote:
No, this was in 7D. The fastest thing I could get my hands on short-term.
paulfeng wrote:
Oh good, 7D is what interests me. So the 4.5s after "buffer stall" for the Sandisk Extreme 60MB/s... Is "buffer stall" at the end of the initial unrestricted burst, say, 16-20 or so raw-only frames depending on scene complexity? Or do you allow for a few hiccups, and define "buffer stall" a bit further on? Or is this RAW-JPEG?
Thanks for the report.
stanj wrote:
This was raw+S JPG (coarse). There's no way I can trigger the stopwatch right on the very first hiccup, as it's 1/10s at most, then the camera fires two more shots. So the timing started when the camera really stopped (hiccup + 2 frames) and ended when the red LED turned off.
paulfeng wrote:
So... about how many shots is this? 6+2? or better? (Sorry, don't have camera in front of me right now) I know the buffer is not as deep in RAW+JPEG, but don't recall if it gets better with S JPG (coarse) that you used.
Just trying to get a feel for how many MB/sec the card is taking. My testing has been with RAW-only bursts, given the deeper buffer for that.
In the name of customer service, I got off my butt and walked over to my buddy with his 7D. Raw+S is indeed 6+2, which doesn't suck. Then I tried raw only and subsequently repeated the test 3x - 23 frames. I don't get it. I understand that it would be less (actually I don't really, knowing how the mk1 cameras worked but I digress). My 1Ds3 actually has more buffer raw+S than raw only, which makes absolutely no sense. But 23 frames vs. 8?
ISO 100, MF, 1/8000s @ f2, picture of the sky, in case you're curious. No funny noise reductions turned on.
You know, if your day job is what I think it is, I am a customer.
I got off my butt and walked over to my buddy with his 7D. Raw+S is indeed 6+2, which doesn't suck.
doesn't?
Then I tried raw only and subsequently repeated the test 3x - 23 frames. I don't get it. I understand that it would be less (actually I don't really, knowing how the mk1 cameras worked but I digress). My 1Ds3 actually has more buffer raw+S than raw only, which makes absolutely no sense. But 23 frames vs. 8?
Right, one of the goofy things about the 7D. Formerly (pre-7D... well, pre-50D, before the embedded JPEG in the CR2 became full-rez) I often shot RAW+JPEG, but now I only shoot that way if I know I won't want to stretch the buffer.
Uhhmmmm... did you happen to note how long it took the buffer to clear when filled with 23 RAWs? (That would probably be in the high 400s of MB, so then we could calculate the MB/s...) You know, because you live to serve my inquiries.
paulfeng wrote:
Uhhmmmm... did you happen to note how long it took the buffer to clear when filled with 23 RAWs? (That would probably be in the high 400s of MB, so then we could calculate the MB/s...) You know, because you live to serve my inquiries.
vpexp wrote:
I have a 7d with a 16GB Speedy A-Data and a 32GB GMonster 533X Pro
With the 7d I recorded the following average times AFTER filling the buffer:
Time to preview: A-Data 4.5 Sec, GMonster 3.0 Sec
Time to clear Buffer: A-Data 18 Sec, GMonster 6.5 Sec
YMMV
Say, did you fill the buffer in RAW only shooting, or RAW+JPEG? Big difference in effective buffer "depth" and amount of data being shuffled around. Thanks.
I'm almost certain it was Raw+JPEG. I would do it again to be sure but I sent my 7D back to Amazon. It just would not focus properly with the same lenses that work great on my DsII and 20D. I could make some of them work decent with MA, but was concerned about how that would work over the long haul especially as to different zoom lengths and distances. Since Amazon will take it back I decided to go that route rather than send it to Canon. Now trying to decide if i want to try another 7, go with a 4 or wait for a ds4.
Mike, where did you wind up buying your Photofast cards from? Also, is the www.mydigitaldiscount.com website posted in this thread a reputable place to buy from. I'm in the market for some CF cards for an upcoming vacation, and I need to purchase soon.
According to your profile you own the 40D. To get any good speed out of the newer and faster UDMA cards, you need a good and new camera that supports UDMA 6 cards.
