197 of 243 people found the following review helpful
4x as expensive as Transcend Class 10... but worth it if you transfer LOTS of data. Frame advance rate increase is 100% myth.,
Verified Purchase(What's this?)
This review is from: SanDisk Extreme Pro 16GB SDHC UHS-1 Flash Memory Card Speed Up To 95MB/s- SDSDXPA-016G-X46 (Personal Computers)
UPDATE: May 31, 2013 - I have recently moved from Windows 7 (what I was running when I wrote the original review and the previous update) to Windows 8 and have noticed that Nikon file uploads from this card are now substantially faster. I have not done in-depth testing but my Windows 8 uploads of my Nikon files now transfer at approximately 60MB/s sustained. MUCH better than the sustained 20ishMB/s for my D800E uploads under Windows 7 and still substantially faster than my D5100/D7000 uploads under Win7.
And lol, I don't want to get into any debates about how crappy Win8 is. Yes, out of the box it is downright unusable. But, for just $5 you can instantly purchase and download Stardock's Start8 product that will return the Start Button and Start MENU (as opposed to Win8's Start SCREEN) and you can disable most of the really bothersome corner actions and swipe actions from Start8. Then, Win8 is actually pretty tight! It's fast and has WAY better sleep/resume functionality than Win7. My Win8 computers wake up faster than my Galaxy Note 10.1 tablet! However, with Win8, scrollbar contrast is absolutely horrible in browsers and there is no setting or even RegEdit to remedy this. Stardock's WindowBlinds (essentially custom skins for the User Interface) for $10 MAY remedy this and I will be looking into this shortly... So, use the failure of out-of-the-box Win8 as an awesome opportunity to get really good touchscreen laptop or even desktop hardware at really discounted pricing then spend $5 for Start8 and you're all set! ;)
UPDATE (This update was written prior to my May 31, 2013 Update above): As I now have a Nikon D800E I thought I'd share my results for that particular camera (and presumably the D800 as well) as this card will be a likely choice for D800/E users. As mentioned in my original review below, there appears to be an issue with Nikon files and transfer speed being relatively slow with this card. This problem is worse with the D800E. I shoot either Uncompressed RAW + Large Basic JPEG or Lossless Compressed RAW + Large Basic JPEG and approximately 75% of the time my D800E files transfer at 18-22MB/second. This is no faster than transfers from a Transcend Class 6 card. The other 25% or so of the time I get transfer rates as high as 38MB/second. Better, but pretty poor performance from a claimed 95MB/second card. I have yet to sustain a 45MB/second transfer with my D800E files.
In all fairness, I have not shot with any other cards in my D800E and don't know if slower cards will transfer slower than this card does.
ORIGINAL REVIEW BELOW:
First, let's be clear about transfer speeds. And for the record, all speeds I mention in this review are my ACTUAL MEASURED speeds, not manufacturer-claimed or hypothetical limits. I'm talking my real world experience. I use a Transcend TS-RDF8K USB 3.0 card reader and I transferred to a benchmarked 514MB/s write, 551MB/s read Mushkin Chronos solid state drive. See 3rd to last paragraph for camera frame advance rate information and the last paragraph for USB 2.0, USB 3.0 and FireWire information.
My main concern with memory card speed is for computer uploads after a day or night of shooting... First, let me get the ATTO Disk Benchmark numbers out of the way. Doing the 1GB Total Length test, from 64K to 8192K samples, the read speed is consistently at 85MB/s for reads and 71MB/s to 77MB/s for writes. I never got even 86MB/s or more (aside from the initial data transfer rate spikes) once during my testing and feel SanDisk is lying about the card's ability to hit 95MB/s. It never even hit 86MB/s a single time. So, that's benchmarking. Now on to the real world.
I have some conflicting results with these cards (I have 3 of the SanDisk 16GB "95MB/s" cards; 1 for each of 3 cameras):
When transferring files from my Canon S100 card, I routinely hold 80-81MB/s uploads to my computer. Not bad at all! Especially being that they are real world numbers. However, they're supposed to be 95MB/s cards. The only time I ever see 95MB+/s is the MOMENT they start transferring data. Same as any other card, there is that initial spike and then the numbers drop fast. So, 80MB/s is a nice fast upload even though I paid for 95MB/s. I feel like I'm getting shorted by 16%. :(
When transferring from my Nikon D5100 card, after the initial spike I only sustain, *gulp*, about 40-45MB/s uploads. :( VERY disappointing! That's only 33-50% faster than the 30MB/s cards that cost WAY less and not even double the speed of the Transcend Class 6 or 10 cards (Transcend Class 6 is the same speed as their Class 10) that cost 1/4 what these cards do. I'd like to blame SanDisk for this but in all honesty, I think it is something with the Nikon files. I don't see how this is possible but when I put my Nikon files on the Canon S100's card, I get the same 40-45MB/s transfer speeds. Therefore, it's not an individual card's idiosyncrasy. It's probably something with those Nikon files. Makes no sense to me, but I can't figure any other reason. It's not the card, because they all benchmark within 1% of each other and handle Canon S100 files like the other cards and Nikon files like the other cards. If anyone can help me out with this Nikon slow speed issue, please comment here. Thanks. :)
Uploading Nikon files TO the card (write speed) from my computer, I get about 60MB/s. Strange that this is faster than the read speed...
REGARDING CAMERA FRAME ADVANCE RATE... I've reviewed the SanDisk 30MB/s (real life 30MB/s computer uploads with USB 3.0) card and Transcend Class 10 cards (real life 25MB/s computer uploads with USB 3.0) in the past and I found there to be literally only approximately a 0.1% SanDisk 30MB/s card frame advance rate advantage in both my Nikon D90 and D7000. Such a minuscule "advantage" could easily be attributable to my stopwatch button-pressing. I informally tested the "95MB/s" card in my D7000 in Continuous High advance shooting "Lossless Compressed" RAW files only, not RAW+JPEG. I got 9 frames (buffer capacity) at rated FPS of 6FPS and then jerky buffer-restricted advance at 1.5 frames per second. My Class 6 and 10 cards give me the same 9 frames in 1.5 seconds and then continue after the buffer is exhausted at 0.7FPS. In other words, the "95MB/s" card gets you an extra 0.8FPS after the buffer is exhausted. So, card speed means pretty much nothing when shooting RAW files. I don't know about Class 2 or 4 cards potentially slowing things down, but who cares about such slow cards that nobody has anyway? ;) However, though card speed does nothing for RAW FPS, what it may do is raise the JPEG quality and/or size that a camera can shoot in while maintaining maximum FPS (i.e. hypothetically 6FPS forever in Normal Medium JPEG with the "95MB/s" card vs. Normal Small JPEG with a typical Class 10 card). RAW frame advance rate is ALL about the camera's buffer and data output rate, NOT the memory card. PERIOD. It is a myth that card speed matters for RAW frame advance rate.
So, is it worth it? To me, yes. Even at only 40MB/s, these cards save me a lot of time when uploading several to many GB per transfer. At 80MB/s, fuhgeddaboudit, absolutely! If you get paid for photography, time is money and these cards save a lot of time and frustration waiting therefore are worth the money. If you shoot video, you'll save LOTS of time so these cards are totally worth it. So if I love the cards so much, why only 4 stars? Because if you only transfer 250MB per day, these cards will do nothing for you. They won't help your camera in any way and will only save you literally 6.9 seconds per day (250MB takes 10 seconds with a Transcend Class 6 or 10 card or takes 3.1 seconds with a true 80MB/s card like this SanDisk "95MB/s" card). I'm also upset that I usually only get 40-45MB/s uploads and many people are buying this card thinking it will give their DSLRs crazy fast frame advance rates and that is straight-up mythology since it won't even help at all and for them it will be totally wasted money.
A NOTE ON COMPUTER TRANSFER SPEED... If your computer is NOT FireWire and/or USB 3.0 equipped and/or you are using a USB 2.0 card reader on a USB 3.0/FireWire computer, you're NOT going to get better than a hair over 20MB/s transfer speeds regardless of card speed. Just a limitation of USB 2.0, not the card.
Tracked by 7 customers
Sort: Oldest first | Newest first
Showing 1-10 of 50 posts in this discussion
Initial post: Feb 24, 2012 7:42:15 PM PST
Really appreciate the review 7. I just purchased a D3100 for my wife and am really struggling with what card to buy. While it would be nice to use the 1080p video on the camera when we wanted, I bought it strictly for photos. All DSLRs under $1000 seem to have a video function. Based off of Nikon's website, I'm seeing that I can get by for photo and video without performance loss of any kind if I go with a class 6 card that can sustain 20MB/s. Am I correct in this estimation? Again, nice review. Thanks.
In reply to an earlier post on Feb 25, 2012 3:02:53 AM PST
Last edited by the author on Feb 25, 2012 3:04:34 AM PST
@Dude/Jack - Thanks for the compliment and you're welcome too. I also buy cameras for photo, not video, but have to admit that recently I'm finding dedicated instant video recording buttons irresistible to press and am considering video much more than a garbage feature now. Don't be surprised if you get sucked in too. ;)
If you don't have and don't plan on getting a USB 3.0 equipped computer with a USB 3.0 card reader, get a Transcend Class 10 card of whatever size or 32GB Duracell 200x (30MB/s) card. For whatever reason, these Duracell cards have big bang-for-buck at 32GB and aren't as good a bargain at other sizes.
If you are USB 3.0 ready and shoot a lot of photos and/or will be shooting video, get the 32GB Duracell 200x or faster card.
Just remember that card speed is ALL about uploading duration. Any Class 6 or faster card will unlock all of your camera's potential.
Hope that helps. Happy shooting! :)
In reply to an earlier post on Apr 1, 2012 3:55:40 AM PDT
Last edited by the author on Apr 1, 2012 3:55:50 AM PDT
Are you saying, that this extra speed of 50mB/s in this than SANDISK EXTREME PRO 16gb, UHS-I Class 10 45mB/s (LINK-> SanDisk Extreme Pro 16 GB SDHC UHS-1 Flash Memory Card 45MB/s SDSDXP1-016G-X46 )
is of no use for better frame rate (after the buffer fill) and/or recording HD video (highest quality) ?
I plan to use it for Nikon D5100 & D7000 (Both are UHS-I compatible); Yes i also want it to be future-ready
That way, Sandisk Extreme cards or even ULTRA series with just 30mB/s speeds, as more than enough & doesn't affect anything on camera (Just affects transfer times; if one has latest card readers)
I take it that you are professional. Kindly reply in as much detail as possible
Thanks in advance
In reply to an earlier post on Apr 1, 2012 12:02:16 PM PDT
@AmazonMan - Yup. That's exactly what I'm saying. :)
If TRANSFER FROM CARD TO COMPUTER speed is NOT important to you, get the Class 6 or 10 Transcend cards and save TONS of money. Your camera cannot tell the difference between those cards and the SanDisk 95MB/s cards. I knew this before I bought the 95MB/s cards but I DO need that card to computer transfer speed. But yeah, my frame rate and video capture are no different with the 95MB/s cards than they were with my Transcend Class 6 cards.
I have never done frame rate tests with my D5100. I never really use continuous advance in my D5100 so never really thought to do
that test. I can hypothesize that your card won't make a difference because the D5100's frame rate is so slow to start with and also it seems pretty clear to me that the issue limiting frame rate after the buffer is exhausted lies with image processing speed and/or data throughput to the memory card itself. I don't know Nikon's data throughput for their cameras prior to the D4 and D800 but those prior cameras all have USB 2.0 output via cable. IF the data transfer rate to the memory card is at USB 2.0 speed, we will have uncovered the issue. Based on the post-buffer-fill frame advance rate I've measured, this looks to be true. USB 2.0 is really only capable of SD memory card transfers at up to about 20MB/s. So, if my theory is correct then it wouldn't matter if your CARD could transfer at 999TB/s. Because the CAMERA can only do 20MB/s.
My findings on my D90 and D7000 show the same thing. My slowest card (Transcend Class 6) will produce the same frame rate to within 0.1% (sounds like stopwatch button-pressing error to me...) as my fastest card (SanDisk 95MB/s). For the kind of video the D5100 and D7000 shoot in their highest resolution/fastest frame rate, I believe you only need Class 4 to keep up, but I'm not sure about that. I've never had ANY trouble with Class 6 or above.
And by the way, Class 6 is no different than Class 10. On USB 3.0 and FireWire they both transfer at up to about 25MB/s. ONLY cards that specifically state their transfer rate (i.e. 30MB/s, 45MB/s, 95MB/s) are any faster than standard (no MB/s rating) Class 6 cards. My Transcend Class 6 and Class 10 cards transfer at 25MB/s. My SanDisk Extreme III Class 6 30MB/s cards transfer at the exact same rate as my SanDisk Extreme Class 10 30MB/s cards: at 30MB/s. Funny that even SanDisk labels different cards as Class 6 and Class 10 even though they are both advertised as 30MB/s cards that really do transfer at 30MB/s.
And here's a fun fact that I don't understand. Since you're a Nikon shooter, this might matter to you too. For some weird reason, if I'm transferring a lot of NIKON photos from the card to the computer, say 10GB, my sustained transfer rate will be around 45MB/s. The same amount of CANON photo data will transfer at around 80MB/s. I don't understand that at all. Makes me wish I got the 45MB/s cards instead and saved the money. But, lol, the 45MB/s and 95MB/s cards are about the same price. Go figure. None of this makes any sense anyways... Lol. :D
I'm just wondering if the 95MB/s cards will make a frame rate difference in the D4 and D800. Their data output via cable is USB 3.0. So... IF that's the rate data writes to the card, well, we're looking at true 550MB/s maximum transfer rate. And lol, that would barely be enough to keep up with a D4 shooting RAW + Fine Large JPEG in Continuous High. So, SanDisk, when's the 550MB/s card coming out? ;)
In reply to an earlier post on Apr 1, 2012 5:45:15 PM PDT
Amazon Customer says:
I don't know about Nikon cameras, but I can say as far as the Sony a77 is concerned the card speed affecting frame rate is no myth. There is a noticeable difference between my uhs-1 45mb and 95mb cards in how quickly I run out of buffer at 12 frames/sec (13-14 vs 17-18) and how quick successive shots are (1-2 a sec vs 4-5).
In reply to an earlier post on Apr 2, 2012 11:00:16 PM PDT
Last edited by the author on Apr 2, 2012 11:00:43 PM PDT
You said "Because the CAMERA can only do 20MB/s."
You mean to say buffer clearing speed(of camera) is 20mB/s & any card having a write speed above than that won't have any advantage on FPS (Though high speed cards will still be faster in transferring data to PC; if one has latest card readers)
1) Where you found that info about Nikon D5100 and where to find the same speed for D7000?
2) UHS-I is officially described as 50mB/s. Now when Nikon has officially labelled D5100 (D7000 too)as UHS-I supporting..why such slow speeds of 20mB/s?
In reply to an earlier post on Apr 2, 2012 11:34:12 PM PDT
Okay, to be clear, I cannot say that I am CERTAIN about buffer speed or transfer speed. I am theorizing, based on my results/experience, that the data transfer speed from camera to memory card is limited to about 20MB/s. The buffer speed could possibly be insanely fast, but if the camera can only WRITE to the card at 20MB/s, it doesn't matter how fast the buffer speed is. And vice versa, if the transfer speed is insanely fast but the buffer's output speed isn't, you're not going to get fast writes.
Yes, like I said, my D90 and D7000 FPS rates with my Transcend Class 6 card ARE IDENTICAL to my FPS rates with my SanDisk 95MB/s card. However, transferring data to my USB 3.0 capable computer via a USB 3.0 card reader is a whole different story. Transcend Class 6 transfers at 25MB/s and the 95MB/s card can sometimes sustain 80MB/s but with several GB of Nikon files, the sustained rate will be about 45MB/s.
1) I do not KNOW that the D5100 and D7000 have the same speed for anything. I've never tested the effect of different memory cards on the FPS of my D5100. I really don't even use continuous advance on my D5100 so I never thought to test the speed. I would expect the same thing. I don't see any reason that the D5100, which is definitely no FPS speed demon, would have faster writes to the card. I guess it's possible but that would make no sense and I've never tested it.
2) "Compatibility" or "support" doesn't really mean anything. It could just mean that the camera is capable of USING UHS-1 cards. Not necessarily that the camera can utilize the full capabilities that come with UHS-1. All I know for CERTAIN is that it never mattered what memory card I used. From Transcend Class 6 all the way on up to SanDisk 95MB/s, I ALWAYS get the exact same fps on my D90 and D7000. I've also never had any trouble at all shooting video at the highest resolution and fastest frame rate with any memory card I ever used and those cards have always been Transcend Class 6 and faster.
Trust me, for the D7000 and D90 (and probably all Nikons except MAYBE the D3, D3s, D3x, D4, D800, D800e) as long as your card can write at 25MB/s (probably even 20MB/s) or faster you're going to get all the FPS your camera is capable of.
I can't remember exactly what resolution I had to drop to so that there was no lag. You don't get RAW shooting if you want no lag, you MUST shoot in JPEG only. On the D90 I think it was something like Normal Medium or Fine Medium JPEG or lower/smaller. I don't remember the D7000's max resolution for no lag. Probably Normal Small or Normal Medium at the most because the frame rate is so much faster than the D90 and the file sizes are larger.
In reply to an earlier post on Apr 3, 2012 12:11:23 AM PDT
Last edited by the author on Apr 3, 2012 12:14:37 AM PDT
Thanks for reply..i agree to your point 1 & "Compatibility" or "support" thing in point 2... but OVERALL i think you haven't tested/tried things in detail.
Of course it has to be RAW, if speeds/lag/bottle necks in camera/cards need to be tested (to its limit)
There is nice review on SD EXTREME PRO 16mB 45mB/s (including video)...which compares a few cards & gives some good info.. have a look...
In reply to an earlier post on Apr 3, 2012 1:02:04 AM PDT
I think you misunderstood me. The resolutions I listed (i.e. Normal Medium) was the maximum resolution/size combination you could use and still have maximum FPS. I was just going off memory though and I don't remember exactly what the resolutions were that enabled unrestricted FPS.
HOLY SMOKES! Eureka! I checked out your link and the video there. That guy's results were a lot different than mine so I initially thought he was full of it so I retested on my D7000 in RAW only at F1.8 and 1/1000 and manual focus. I got 22 shots in 10 seconds with the 95MB/s card. And this was the surprise for me... I got only 15 shots in 10 seconds with my Transcend Class 6 (25MB/s) card.
I will admit that when I initially tested my 95MB/s card I did so for not much time after the buffer was exhausted. When the advance got jerky I mistakenly assumed that the result was the same as my previous tests which were very thorough in which case my Transcend Class 6 cards performed exactly the same as my SanDisk 30MB/s cards.
Here's the thing though, this kind of jerky frame advance is useless. So, you have to determine the highest resolution that you can shoot at while still maintaining unrestricted FPS. Now I've got to run more tests to determine that resolution.
Here's something to think about... A D7000 14-bit "lossless compressed" RAW file like I was shooting is about 17MB and it takes about 1.5 seconds to use up the 8 or 9 frame buffer. I know it says 10 in the viewfinder but I only get 8 or 9 in real life. So let's say 9 shots in 1.5 seconds and then 13 more in 8.5 seconds. That's 1.529 shots per second after the buffer is toast. 1.529 shots x 17MB per shot = 25.99MB/s. Still, nowhere near 95MB/s. Or even 45MB/s, for that matter.
But hey, I'm still really surprised it made a difference at all! :D Now to find the highest resolution it can maintain full FPS at with no lag... Gotta get back to you on that one. It's a little late here... :)
Posted on Apr 24, 2012 1:20:20 PM PDT
Millennium Arts says:
"7": Thanks for explaining the frame advance rate matter. I have the Panasonic Lumix FZ100 and seek "faster" shot-to-shot speed. Your explanation that a new and faster card is not going to help, and such is very helpful to me. Thanks.