How long to render?

Discussion in 'Video Cameras' started by Jim McLean, Sep 6, 2006.

  1. Jim McLean

    Jerry Guest

    <snip other detailed comments>

    I've been saying that for years! Indeed if Tony is honest he would
    agree, he has all but admitted that he has come to computer video
    editing from a broader use of computers rather than the other was
    around IYSWIM.

    No flames intended.
     
    Jerry, Sep 8, 2006
    #21
    1. Advertisements

  2. Jim McLean

    Jerry Guest

    No, like a fool I was letting M$ spiel chucker do the work rather
    than reach for the dictionary! :~(
     
    Jerry, Sep 8, 2006
    #22
    1. Advertisements

  3. Jim McLean

    Jerry Guest

    Not a direct answer to your question but would the Canopus
    'FireCoder' (circa £269.08) card be of any use, although you will
    need a PCIe slot?

    Second from last product on this URL
    http://www.planetdv.net/Content/By_Manufacturer/Canopus/
     
    Jerry, Sep 8, 2006
    #23
  4. And what will that tell us about disk performance as a limiting factor
    when rendering video?
     
    Laurence Payne, Sep 8, 2006
    #24
  5. Jim McLean

    Tony Morgan Guest

    It does give relative performance metrics for C: and the second drive.
    And if you'd care to run a comparison using the Control Panel's
    Performance console while actually rendering [1] - you'll see that the
    console reflects what Sandra told you.

    [1] Obviously you're going to have to change your rendering engine's
    input/output drive to do the comparison.
     
    Tony Morgan, Sep 8, 2006
    #25
  6. Jim McLean

    Dave R Guest

    I believe Laurence's point is that there is a significat difference
    between outright disc speed tests, and speed of rendering video.

    Benchmarks do: read block of data; write block of data.

    Video rendering does: read block of data; process data - render new
    frames; write out different block of data.

    The limiting factor in the latter is like to be the time taken to render,
    which is a lot slower than the doing nothing of benchmark programs.
     
    Dave R, Sep 8, 2006
    #26
  7. And what will that tell us about disk performance as a limiting factor
    when rendering video?
     
    Laurence Payne, Sep 8, 2006
    #27
  8. Indeed. As long as the drive(s) can provide and store data as fast
    as is needed, it's irrelevant whether a different configuration could
    provide it any faster.

    This point is often made regarding using slow laptop drives for video
    capture. In fact, it's often made by our own Tony Morgan :)
     
    Laurence Payne, Sep 8, 2006
    #28
  9. Jim McLean

    Jerry Guest

    But surely the HDD (or it's motherboard controller) is now the
    limiting factor now that we have hyper-threading and duel core CPU's
    etc. It is now quite possible to obtain a real-time, full frame
    preview purely on CPU power alone (meaning the HDD is reading only) -
    no one has ever suggested that one can obtain a real-time render,
    especially to the same drive that one is reading off, this surely
    means that reading *and* writing to the same disk whilst rendering
    has to be slower than reading from one and writing to another IYSWIM?
     
    Jerry, Sep 8, 2006
    #29
  10. Jim McLean

    Tony Morgan Guest

    In message <>, Laurence Payne
    Ahah... But with my new Dell XPS M1710 notebook it's noticeably faster
    rendering than my old desktop with the same CPU Clock Speed. Although I
    do concede that I'm using an external USB2 HD for capturing and
    rendering the video.

    But your point is taken Laurence.

    Two or three of my friends, who are the most vociferous about slow
    rendering sit for five or ten minutes then get bored and start doing
    something else on their computer. One even starts on-line gaming and
    moans just as loudly that his game runs so slowly :)
     
    Tony Morgan, Sep 8, 2006
    #30
  11. Jim McLean

    Jerry Guest

    You mean you have some?!! :~)

    who are the most vociferous about slow
    Indeed...
     
    Jerry, Sep 8, 2006
    #31
  12. In some circumstances you can achieve a considerably FASTER than
    real-time render. But I still doubt that disk performance would be a
    limiting factor.
     
    Laurence Payne, Sep 8, 2006
    #32
  13. And as the notebook has an Intel Core Duo processor, an unqualified
    statement "same CPU clock speed" is thoroughly misleading, isn't it?
    NAUGHTY Tony :)

    And I'd strongly suspect that there's less chance of a glitch
    capturing to the internal drive than the external :)


    When you get a chance, can you render an few minutes of video using
    different combinations of internal and external drives? Relative
    times taken would be very interesting.
     
    Laurence Payne, Sep 8, 2006
    #33
  14. Jim McLean

    Andy Champ Guest

    Almost certainly yes. I'm sending them an enquiry!

    Thanks

    Andy
     
    Andy Champ, Sep 8, 2006
    #34
  15. Jim McLean

    Just D Guest

    Ok, what about real numbers instead of just talking and arguing? I could
    test that on my own machine having 2 Gigs RAM and a very powerful Terabyte
    RAID made with 4 different 7200 hard drives and able to produce 100
    MBytes/sec up and down. I know that when I use a separate hard drive then
    the rendering goes a little bit faster even in this case because 100 is
    actually a linear speed, when I read/write to the same disk then it doesn't
    go smoothly, the noise is much louder, the difference is the very first
    percents although, not higher. If I test same on my laptop with one built-in
    OR from my built-in to external USB2 drive then the second combination works
    much faster. The built-in HD is able to provide 36-41 MBytes/sec "only", but
    the searching/positioning takes time and it's much better when I use another
    physical device to store the target file which is smaller and USB2 is not
    critical for that case. Sorry, have no time right now, but probably I could
    start some test rendering of 10-20 GBytes piece to compare both machines in
    both ways. Realistically on 3.2 HT laptop 1 G RAM, 2.8 Dual Core Laptop 2 G
    RAM and 3.2 HT crazy RAID server 2 G RAM = in al cases the speed is high
    enough, I see in Vegas that it's even higher than the real time, people walk
    faster. :) The quality using for this rendering was set to the best, NTSC,
    etc.

    Just D.

     
    Just D, Nov 29, 2006
    #35
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.