User Tag List

Thanks useful information Thanks useful information:  4
Results 1 to 8 of 8

Thread: Some memory card info CF and SD.

  1. #1
    A royal pain in the bum! arthurking83's Avatar
    Join Date
    04 Jun 2006
    Location
    the worst house, in the best street
    Posts
    8,777
    Mentioned
    4 Post(s)
    Tagged
    0 Thread(s)

    Some memory card info CF and SD.

    I've had in mind to get myself some proper memory cards for some time now .. you can never have too many I reckon ... so this just happened the other day(or was it today .. work has mushed my brain these past two weeks).

    Anyhow, I got myself a 32G Sandisk Extreme card to get the D800 working at least.
    I have a handful of CF cards, some of which work badly, and others that don't on the D800 .. the real painful card is this noname brand Patriot card in the LX Series.
    A brilliant card on the D300, even tho it's a no name brand, is super fast and allows full use of the D300's ability.
    I've had this card now for about 3 years or so .. maybe more, it's been so long .. it has been my workhorse card since I got it for the D300.

    BUT!! it just doesn't work on the D800. Have no idea as to why, but the camera just reads Card Err on the top LCD.
    So this card is stuck in the D300 and allows my 12 yo son to play at full speed on his D300!(he love this 6fps bizo .. the hoon! )

    So this Patriot LX series card I know is fast, but exactly how fast was always a guess, up until I finally got myself a USB3 card reader now.

    Patriot LX card @ 8Gig:

    via the PC sees 70Mb/s read speeds, and just a whisker over 50Mb/s sustained write speeds.
    This makes sense because on the D300, it allows full use of the buffer, and when that's full, it then continues along at the expected 1fps rate indefinitely until full. That is, you will tire before the camera reaches the 100 burst limit.
    Actually!... I should clarify that last bit. You and I will tire before the 100 frame limit is reached, but my hoon 12 yo son won't

    So back on the topic and reason for this post ... and the reason for this post is that once again, my expectations for an expensive purchase (or more accurately disappointingly excessive financial expenditure!) have not been realized again!
    I read good things about these Lexars, and I can't see it .. all!

    With the Lexar what I do see is massively fast read speeds:

    Lexar Professional 32Gig 800x UDMA7 CF card:

    via the PC this card reads at an astonishing 110Mb/s and hovers at 115Mb/s all too often.
    This is backed up by the fact that if I blink, 2 or 3Gig of image files have been transferred to the PC, where my usual routine is(or was) to start the transfer and then make a cuppa.
    But the time it's cooled down to a drinkable level, and I'm now back at the sink and washing the cup, there may be a few more files that need to be transferred, after which I'm ready to review and process, or not.
    Very painful to transfer 8 Gig of files via USB2 .. and if you have USB3 capability do yourself the favour and get a USB3 reader!

    So again via the PC, I can get up to about 37Mb/s write speeds to this supposedly fast (800x!!) card, which is about 50% slower than the no name Partiot CF card from all that long ago.
    So the USB3 interface is not the issue, the card is. Rob Galbraith reckons he sees about 55Mb/s from the 600x version of this card! with a burst of 29 frames before the D800 buffer fills all the way and slows the shooting rate.
    So I reckon I've probably done something wrong somewhere(which I'll investigate).
    Card was formatted in camera, and on the PC shows 32Kb cluster sizes and all that jazz .. but all I see is an expensive failure so far, in terms of money spent v's any gain in performance.

    But I had to do this anyhow, it's just that now that I know what I know, I wouldn't recommend this Lexar 800x bogus performance memory card based on what I've seen so far.
    (if this changes by way of new info coming to light, I'll edit my post to suit the updated info)

    I've had two 8G CF cards of A-Data fame for almost as long as time itself. These cards were originally purchased for use on a PDA for my mapping program, and speed was not a major issue way back then.
    They were only used due to their capability to contain data, on the D300 this meant a few days worth of shooting easily.
    On the D800, not only is 8G very little in the way of actual frames captured, but the speed of the card became a nightmare.
    A single image would take minutes to write .. and so testing reveals why.

    A-Data 8G CF card:

    via the PC it has only evr achieved a top write speed of about 7Mb/s or less(on a good day ) .. so it's easy to see why it took so long to actually commit a single D800 file to it's memory modules.
    USB3 is wasted on these old slow cards. Read speeds have only ever been at about the 15-17Mb/s, and I initially thought this was due to the USB2 interface, but with the USB3 reader, I see 17-19Mb/s .. not even 20Mb/s. Remembering though, that this is not a criticism, just an observation. The cards were never meant for any camera .. I didn't even have a camera back on '04 when I got them.
    They simply ended up as redundant hardware for a device that no longer exists, and were kept as backup backups .. just in case.

    And I have a Sandisk Extreme SD card of 32Gig capacity.
    This one is also weird.

    Sandisk Extreme 32Gig SD card:
    Formatted in the exFat file system as it's supposed to be, the write speed is a consistent 33Mb/s flicking up to 35 every so often, but my take is that 33Mb/s is a good average.
    This again is via the PC too.
    Read speeds are interesting, and this has now happened on three occasions that I've tried it.
    It starts off at about 35-40Mb/s and hovers there for short while(maybe 30sec or so), and then suddenly jumps up to 95Mb/s for no reason, and then settles down to about 70-75Mb/s for the remainder of the large-ish file transfer.

    I had a stash of about 4Gig that I transferred to and from the PC to each card via Windows Explorer, each time clearing out the prefetch cache .. so whilst not 100% accurate, is a reasonable indicator as to what I see in real life.

    D800 can't manage more than 15 frames in a burst at the highest quality image settings before it chokes under the weight of it's own data, no matter the card used. If anyone has any info as to why this may be, more folks may benefit from this than just myself.
    Even the old slow A-Data manages this, as the 15 frames is a product of the camera's buffer, not the card. But once those 15 frames are shot, then it's matter of waiting for them all to be written to the card itself.
    The D900s will be an old camera before that task is complete, so this is not something I ever expected to have to deal with in real life .. and why the Lexar card was purchased in the first place.
    I can get 19 frames from the buffer if the raw files are set to compressed, and then 1fps after that.

    The real annoyance is the once mighty Patriot card. Faster write speeds than the fast Lexar, that is much newer and much more expensive, even via the supposedly complimentary card reader.
    I'd have expected the Lexar usb3 card reader to work better/faster with a similar branded card than I expect with a foreign card that barely even works with most hardware!

    I've seen these Patriot cards for sale now for about $20 or so for the 16gig versions, but even with all this pseudo praise, the card is useless on a D800 .. just doesn't work.
    And what's even more annoying, I can't format the Patriot via the Lexar reader, only via the USB2 reader I have .. but it writes a lot faster and reads fast enough. This is just weird.

    Summary:

    I don't hesitate to try no name gear. Usually you get what you pay for, and sometimes this can be a pleasant surprise.
    What I hate is the disappointment of a supposedly superior piece of kit purchased at a premium price that doesn't deliver as it's supposed too!

    .. things I need to do .... read up as to how to best extract more performance from the Lexar before I find myself in a situation where it may be required.

    .. and things I'd like to know .. why does this stuff not work for no good reason? What was(and still is) a perfectly great card on the D300, is useless on the D800. Why do Nikon find the need to change stuff like this for no apparent advantage?

    anyhow .. just some info if this helps anyone else too.
    Nikon D800E, D300, D70s
    {Nikon}; -> 50/1.2 : 500/8 : 105/2.8VR Micro : 180/2.8 ais : 105mm f/1.8 ais : 24mm/2 ais
    {Sigma}; ->10-20/4-5.6 : 50/1.4 : 12-24/4.5-5.6II : 150-600mm|S
    {Tamron}; -> 17-50/2.8 : 28-75/2.8 : 70-200/2.8 : 300/2.8 SP MF : 24-70/2.8VC

    {Yongnuo}; -> YN35/2N : YN50/1.8N


  2. #2
    http://steveaxford.smugmug.com/
    Join Date
    19 Nov 2007
    Location
    About in the middle between Byron Bay, Ballina and Lismore
    Posts
    3,150
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Just to throw another complication in. Remember - read and write speed are different, and speeds vary across a single card.
    Perhaps you wrote that and I just didn't read it. It was a very long post.

  3. #3
    Moderately Underexposed
    Join Date
    04 May 2007
    Location
    Marlo, Far East Gippsland
    Posts
    4,902
    Mentioned
    3 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by arthurking83 View Post

    Lexar Professional 32Gig 800x UDMA7 CF card:


    So I reckon I've probably done something wrong somewhere(which I'll investigate).
    Card was formatted in camera, and on the PC shows 32Kb cluster sizes and all that jazz .. but all I see is an expensive failure so far, in terms of money spent v's any gain in performance.

    But I had to do this anyhow, it's just that now that I know what I know, I wouldn't recommend this Lexar 800x bogus performance memory card based on what I've seen so far.
    (if this changes by way of new info coming to light, I'll edit my post to suit the updated info)
    By total coincidence we have just received two 16gb versions of that card through Cheap Chips and they appear to work in the camera at about the same speed as the D700 with a Sandisk extreme 111 (30 mbs) and that is a MASSIVE improvement in the D800.

    Quote Originally Posted by arthurking83 View Post

    D800 can't manage more than 15 frames in a burst at the highest quality image settings before it chokes under the weight of it's own data, no matter the card used. If anyone has any info as to why this may be, more folks may benefit from this than just myself.
    Read http://www.ausphotography.net.au/for...37#post1096537
    Andrew
    Nikon, Fuji, Nikkor, Sigma, Tamron, Tokina and too many other bits and pieces to list.



  4. #4
    Member
    Join Date
    12 Jul 2012
    Location
    Rockyview
    Posts
    2,087
    Mentioned
    10 Post(s)
    Tagged
    0 Thread(s)
    I've got a 32gb 1000x Lexar in the CF format, and the Lexar USB3 Reader. I find it a huge improvement over the Sandisk Ultra II's I was using as far as write in-camera goes. This is in a 5DIII. Whilst I haven't sat the with a stopwatch when uploading to my computer, I can honestly say the difference in that speed caught my eye immediately. The progress bars zip across my screen at several a second. Faster than my 10MP RAW files from my 40D did. Considering the 5DIII files are three times the size (30mb instead of 10mb), this is pretty good. I'll have to some quantitative analysis and get back to you.

  5. #5
    http://steveaxford.smugmug.com/
    Join Date
    19 Nov 2007
    Location
    About in the middle between Byron Bay, Ballina and Lismore
    Posts
    3,150
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    You need to remember that the transfer speed is controlled by the card and the camera. If the cameras maxim speed is X, then no card will make that any faster. Speeds will vary with quite complex queuing algorithms that are very hard to work out without knowing the exact electronics involved. It is easier to rely on real world measurements.

  6. #6
    A royal pain in the bum!
    Threadstarter
    arthurking83's Avatar
    Join Date
    04 Jun 2006
    Location
    the worst house, in the best street
    Posts
    8,777
    Mentioned
    4 Post(s)
    Tagged
    0 Thread(s)
    All the transfer speeds I've quoted were just estimates based on watching the transfer rates via Windows Explorer, not measured with stop watches.

    I didn't have the patience, nor the hardware(i. stopwatch) to measure actual speeds, just watched the green(write) light on the D800.
    The write speeds on the CF Lexar is definitely faster than the SD Sandisk Extreme card in the D800 as the green light appears to extinguish quicker with the Lexar card.

    Rob Galbraith tested the 1000x and 600x Lexars in the D800 and gets some good results from both, much better than I'm seeing.

    Firstly I'd have at least expected the 800x Lexar to read and write faster than the old Patriot branded card I have, and only the read rates is faster by a considerable margin too, as Warbler found.
    Even using Nikon's Transfer 2, the way it transfers D800 files(at about 70+ Mb each) is awesome to watch.
    The SD card shows acceptable transfer times, and the Patriot CF card is also good.

    What's got me stumped is the not only the actual write time differences between the Patriot CF card and the Lexar CF card, but why the Lexar 800x card is slower in a reader manufactured by the same company.
    if the controller has anything to do with the transfer speeds, then you'd expect that the two devices from the same manufacturer would be matched, tuned or optimized in some way.

    Also, while the Patriot card reads and writes speedily via the Lexar USB3 reader, I can't format it. Windows says that the format was not completed and then the card seems to be corrupted in some way, until I format it again in the old USB2 reader .. which works perfect every time.

    Thanks for the D800 transfer speed info Andrew, my manual is in the usual spot, at the bottom of the camera's box which serves as a reminder not to throw the box out(yet!).

    Two things are for certain tho: Lexar USB3 card reader is money well spent, Lexar 800x CF card is not(really).

    The USB3 reader allows at least 70Mb/s read(ie. transfer) times, and while 100+ Mb/s read times are always better, the difference is really not a significant issue.
    Where the advantage is, is that compared to the old USB2 card reader, which maxes itself out at 20Mb/s, this 3x or more improvement in card to PC transfer times is definitely a bonus.

    I got the new CF card simply because I needed a new CF card, as I knew the very slow A-Data card just wasn't up to the task at all, and because I was already purchasing the card reader. But I think this was not money well spent, and I should have saved myself the extra $100, as the CF card's advantages are so minimal.
    Any old 400x non name brand card would have worked just as well in the D800 I reckon.

  7. #7
    Moderately Underexposed
    Join Date
    04 May 2007
    Location
    Marlo, Far East Gippsland
    Posts
    4,902
    Mentioned
    3 Post(s)
    Tagged
    0 Thread(s)
    Just for you Arthur.

    19 lossless compressed NEF files, 734 mb on the Lexar 800x 16gb card through a Lexar usb2 reader run at around 30mbs in windows explorer.

    I'm not really worried about the time it takes to get them to the pc, I just appreciate the time it takes to clear them from the camera to the card and and with the current configuration it is running at about D700 speeds which for me is adequate and certainly way way way faster than the previous cards.

  8. #8
    http://steveaxford.smugmug.com/
    Join Date
    19 Nov 2007
    Location
    About in the middle between Byron Bay, Ballina and Lismore
    Posts
    3,150
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    To Arthur

    Just because you get a certain result with a windows write doesn't mean you will get the same result with a D800 write. You would also have to test the whole card and also with fragmented storage.
    Why would you expect a card and reader from the same company to be optimized??? It may be optimized for something else, not optimized at all, different generation, or the computer/reader interface isn't good. Lots of variables. p.s. The controller IS important. I do know this stuff as I spent 20 years in the industry with a focus on system performance. It's a couple of years out of date, but it was to do with very high end stuff and the basic rules never change.
    Last edited by Steve Axford; 12-12-2012 at 10:14am.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •