7200 rpm: Internal vs. External

The Kooky Scientist · Post by **The Kooky Scientist** » Fri Feb 04, 2005 2:14 pm

Recently upgraded my internal HD to a 7200 rpm. Wondering if running all Live samples of the internal disc is better/faster than using my 7200 rpm external Firewire HD. Any way to test ? Both methods seem acceptable.

gaspode · Post by **gaspode** » Fri Feb 04, 2005 3:27 pm

The internal ought to be faster as it doesn't have to do multiple levels of indirection through the firewire bus/ide bus to access the data.

That being said though, the difference may be relatively small... both should be pretty decent.

elektrovert · Post by **elektrovert** » Fri Feb 04, 2005 4:05 pm

What about daisychaining a Firewire Harddrive and a Firewire Audiophile card??

will the side effects from doing that be too much??

muthafunka · Post by **muthafunka** » Fri Feb 04, 2005 4:29 pm

Could be barking up the wrong tree here but won't the internal disc also be in use by the OS? I thought for this reason alone conventional audio wisdom dictated that an external HD was always preferable.
What's up anyways Kooky? Any release info??

The Kooky Scientist · Post by **The Kooky Scientist** » Fri Feb 04, 2005 4:31 pm

I think this is scheduled for release in March.

http://www.studio-dust.com/new/03_work/ ... ence03.htm

telepathic regards,
the kooky scientist

AdamJay · Post by **AdamJay** » Fri Feb 04, 2005 6:53 pm

internal will be faster reading AND writing.
if on PC you can test with Fresh Diagnose
if on Mac you can test with Xbench.

kid music · Post by **kid music** » Sat Feb 05, 2005 8:53 am

Hard drive performance affects track count. So the benchmark test I've used with other programs (and I'm sure would work with LIVE) is to keep recording tracks, say 2 or 4 at a time (record anything, even "air"), onto the internal drive. Just record 2, then record 2 more while playing back the first 2, etc.

Eventually the hard drive won't be able to keep up. Make a note as to how many tracks are safely recorded without any performance problems, like dropouts. Then do the same thing, but record to the external drive.

Technically, the external drive should outperform the internal drive if the internal drive is where you have your OS and the LIVE program. But you may have some bottleneck with the external drive, such as an IRQ issue with the firewire port.

-kid music

AdamJay · Post by **AdamJay** » Sat Feb 05, 2005 7:25 pm

kid music wrote: Technically, the external drive should outperform the internal drive if the internal drive is where you have your OS and the LIVE program. But you may have some bottleneck with the external drive, such as an IRQ issue with the firewire port.

-kid music

well, if you wanna get "technical"

the fact is the firewire bus is 400mbps (50MB/sec). any laptop made in the last 2.5 years will have 100MB/sec ide bandwidth, and machines before that 66MB/sec.

Live is loaded into ram, and almost all of the disk access going on during Live is for audio files. you will get better performance with the internal disc.

for instance, i've tested my 5400rpm/8mb cache internal Fujitsu Drive against my 7200rpm/8mb cache Firewire Hitachi Drive.
the internal drive, though slower rotational speed, reads 2% more data per second. That is marginal, but it does say alot for the importance of the ide bus. a Fatter pipeline means a slower drive operates faster, go figure.

but there are so many factors; firewire chipsets, and the fact that every hd is different are just a few. you should really test it yourself. Xbench for Mac and FreshDiagnose for PC are both free programs that will test bandwidth of HD's.

i just load my whole set into ram anyway so for me it doesn't matter so much. To increase track counts, you should load your smallest clips into ram anyway. This will help you far more than that extra 2% bandwidth you get by changing which HD the audio is streaming from.

kid music · Post by **kid music** » Sat Feb 05, 2005 11:07 pm

You are correct. I'm just suggesting a practice I've done with ProTools and Cubase for years: create your own benchmark test on your own system with the program you use, using the routing, efffects, etc., that you typically use. In other words, a "real world" test. I then save this file and its audio, and periodically run the test again to see if my system has slowed down for some reason.

In my experience in the past, keeping audio files on a separate physical disk from the OS and programs has resulted in better performance. But I can see how that wouldn't be the case for every system.

-kid music