Archived from groups: rec.video.desktop (
More info?)
Did notice a 7200rpm laptop drive just cant find the link
On Fri, 23 Apr 2004 13:24:09 GMT, dhaynie@jersey.net (Dave Haynie)
wrote:
>On 22 Apr 2004 20:08:04 -0700, karlengel@excite.com (Karl Engel)
>wrote:
>
>>Conventional wisdom has it that 7200 rpm drives are required for high
>>data thruput apps like video & audio, & thus most notebook drives
>>aren't up to the task for intensive use being 4200 or 5400 rpm.
>
>>made me wonder though: if notebook drives are smaller, with data
>>packed closer together, and the read/write head moves at the same
>>speed as in conventional drives, can't the data be accessed just as
>>fast at lower rpms since the distances to travel are smaller?
>
>In a sense. Actually, the effect of the rotational velocity on media,
>especially video, has not been the key issue in a long, long time. I
>have nice 7200 RPM drive here, from the late 80s, a CDC Wren V, fast
>for its day, but fairly useless for video things. Any old 5400rpm or
>4200rpm drive made today would beat the pants of this one, for most
>things.
>
>Part of the issue is to know what makes a drive fast. The real answer,
>as you seem to intuit, is simple: how many bits are read by the
>head(s) per second. You can get more bits per second by making the
>platter spin faster, by adding platters that are read in parallel, or
>by increasing the density of each platter.
>
>That CDC Wren, a full height 5.25" drive, contains a whole mess of
>platters, and can probably sustain about 5MB/s on a good day (that's
>about the limit of my Amiga 3000 anyway, the system it's on). Modern
>3.5" ATA drives these days sport a single 3.5" platter up to about
>40GB or so. That's an amazing increase in bit density. So you can get
>maybe 40MB/s out of that drive. Video doesn't really care much about
>the drives rotational speed, it cares about the effective streaming
>speed.
>
>The one place rotational velocity still kind of matters is audio.
>Audio, even at 24-bit and 96kHz, isn't a "high data throughput"
>application, by modern standards. Each channel is only 288KB/s, or
>0.288MB/s. When you add up 32 or 64 of these, that's still not a huge
>challenge to a modern drive -- under 20MB/s throughput.
>
>The issue here, though, isn't the data throughput, but as long as
>you're working in mono or stereo tracks in different files, seek time.
>You'll find that, with a large number of tracks, the limitation these
>days IS the seek time, not the drive's linear throughput. And given
>the same platter density, a faster spin will give you faster seeks.
>
>>Is it
>>fair to compare the rpm of notebook drives directly with external or
>>conventional drives?
>
>Well, I'd compare all the specs. Yes, you can compare RPMs directly,
>but you're probably missing some of the issues. On the plus side, as
>you mention, a smaller drive implies less distance, so your seeks
>might be expected to go faster, given the same technology elsewhere.
>But also consider that laptop drives are not optimized for speed, but
>for power consumption. So it's quite possible you won't see any
>advantage here. If I'm worried about laptop drive throughput, I'll
>find out who made my laptop's drive and look up the specific data
>sheet on it. That will give you what you really need: average seek
>time, maximum seek time, and the range (inner and outer) of the
>expected raw throughput (obviously, the drive maker can't tell you the
>effect of your particular OS and file system). Or simply run a decent
>hard drive benchmark on your home and laptop systems, ideally
>something that's minicing what you will really do (eg, many streams
>for multitrack audio, one high bitrate stream for video).
>
>Personally, my laptop is more than a match to my recording interface
>(Tascam USB-122); as long as it can take in two channels at 24-bit and
>48kHz, I'll be happy using it in the field. I really wouldn't think of
>doing anything too "studio" on the laptop. I've occasionally done
>video editing with it, but that's with an external Firewire drive (for
>capacity, not really speed), and of course, faster operations in
>editing are nice, but not the critical thing you need for realtime
>work.
>
>
>Dave Haynie | Chief Toady, Frog Pond Media Consulting
>dhaynie@jersey.net| Take Back Freedom! Bush no more in 2004!
>"Deathbed Vigil" now on DVD! See
http://www.frogpondmedia.com