naturerecordists
[Top] [All Lists]

Re: hardware [esp soundcard] question

Subject: Re: hardware [esp soundcard] question
From: Walter Knapp <>
Date: Sun, 08 Dec 2002 20:45:42 -0500
jplanet ten wrote:
> greetings,
>
> i have been listening to and learning from this group
> for a while, and i appreciate your openness and
> willingness to share your experiences.
>
> i'm about to replace a nearly-dead computer.  i want
> to make sure that i 'outfit' it for audio work, which
> i gather means being attentive to memory, storage
> space, and the sound card. [right now i'm learning to
> use cool edit software, and will probably continue to
> do so for the foreseeable future.]  the primary
> constraint is that it's got to be a pc rather than a
> mac, because of other uses.=20

I use a mac, but a couple comments.

Sound processing is expanding in capability, and a lot of this is the
availability of better processors. More can be done. It takes
considerable memory to keep this moving fast. You really don't want your
PC to be be using the disk for virtual memory. If looking to the future,
get lots of processor, and lots of memory, a gig of memory would not be
unreasonable. What will do it now, may not in a few months. Or will be
slower.

As a thought, the software I use when filtering allows me to build a
whole set of filters, and will run the entire set on the sound in
realtime. That could be a dozen or more single filters. I'm using a
400mHz G4 now and can overload it without a lot of effort, particularly
as at the end of the filter stack I have a Sonogram running. The dual
1gHz G4 here is less overloaded, but I can bring even it to it's knees.
That dual G4 easily runs rings around current pentium based PC's. And
next year the new design G4 processor will be out, it doubles the
processing rate at the same clock speed. You can be sure the software
designers will find ways to give it a workout.

Realtime filtering is a huge advantage, when you can adjust while
listening and watching the sonogram. It takes less power to do filters
slower, but you may have to repeat many times to get it right without
the realtime.

Sounds are also big files, and when processing files your sound software
will be keeping very large working files on disk that it needs to access
fast. Go for the highest speed disks you can do, as well as large. The
fastest disks now are Ultra 320 SCSI, I don't know much about them, and
they are bound to be pricey. I find the Ultra 160's I'm using to not be
too limiting. You can do sound or even some video without a Raid array
with disks like these and not have to wait all the time for the disks. I
also prefer to have more than one disk. I'm using 3 Ultra 160 72gig
Quantum drives inside my G4. If going ATA, each bus is limited to two
disks, and they are slower than the above choices. At least make sure
it's the latest, fastest ATA if you go that route. Too many just judge
disks by price.

That kind of space comes in real handy when it's big project time. For
instance, during working on the Georgia Frog CD, I put all 37 CD's worth
of frogcalls I'd recorded on, then sorted new copies as subsets. Several
levels all instantly accessible. Then I added new recordings this year
to the pile, transcribed off several cassettes of recordings from
elsewhere. The amount of time saved having that all on fast HD was of
huge value. And I don't have to rush to get it back off. Several layers
of working files will stay on the disks until the CD is finally
published, just in case.

> so as i sort through what's out there, would you all
> give me some recommend specs to keep in mind?  i'm
> especially concerned about soundcards, because i'm not
> enough of a techie to know if that's something  that
> can be 'upgraded' as you make a purchase.  [i know
> that one can often specify more memory and larger
> harddrives.]

The very name "card" gives it away. They are generally removable and
replaceable. You can also get really good Sound I/O via the USB port, so
it would be fairly hard to get stuck unless you get into some PC's that
are non-standard.

Walt




________________________________________________________________________
________________________________________________________________________

<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU