Optimal external drive for lap top as main scratch disk

Michael Barrette · March 16th, 2007, 12:11 PM

I am guessing capturing on firewire and interfacing with computer by USB is perfectly fine for the capture process. Then, while rendering and editing, you'd probably want to go back to the firewire to computer... just a guess.

As for VISTA, my lap top came with it, but yes, I am definitely thinking of going back to XP. Vista is a little more fancy visually and I do like the explore functions etc. but overall there is really no reason why I should take up 60% of a one gig chip just to run my system...

cheers,
Michael

Harm Millaard · March 16th, 2007, 01:43 PM

John,

What tests did you use?

If you consider transfer rates for NLE work not overly important, what are the measures that do impact performance in your opinion? AFAIK transfer rates are the MOST important measure for NLE, not sound level or idle power consumption.

John Miller · March 21st, 2007, 04:43 PM

Sorry for not replying sooner - I've just seen this (!)

I used our DV processing software to do burn timecode, convert aspect ratio and some color correction on a DV AVI file, writing the processed version to a new file on a separate drive. (i.e., I had two drives - for the FireWire case, I used two separate interfaces - one on the mobo, another on a PCI card).

Basically, I ran the process once with both drives connected via FireWire and then repeated it with both connected via USB2.0. The difference in overall processing time was about 1%. The total time for the processing was about 125 seconds. Without any processing - effectively just a straight copy but using the DirectShow infrastructure - the time was about 70 seconds. A direct file copy via Windows Explorer took about 40 seconds. I have to say I expected to see a significant difference but was pleasantly surprised that using USB2.0 did not seem to have a negative impact.

When I get a chance, I'll run some simple comparisons using GraphEdit - basically, copy one DV AVI to another using FireWire and then USB2.0. I'll compare those times to the direct copy via Explorer.

March 16th, 2007, 12:11 PM	#16
Michael Barrette Regular Crew Join Date: Feb 2007 Location: calgary, Canada Posts: 48	I am guessing capturing on firewire and interfacing with computer by USB is perfectly fine for the capture process. Then, while rendering and editing, you'd probably want to go back to the firewire to computer... just a guess. As for VISTA, my lap top came with it, but yes, I am definitely thinking of going back to XP. Vista is a little more fancy visually and I do like the explore functions etc. but overall there is really no reason why I should take up 60% of a one gig chip just to run my system... cheers, Michael

March 16th, 2007, 01:43 PM	#17
Harm Millaard Trustee Join Date: Aug 2006 Location: Rotterdam, Netherlands Posts: 1,832	John, What tests did you use? If you consider transfer rates for NLE work not overly important, what are the measures that do impact performance in your opinion? AFAIK transfer rates are the MOST important measure for NLE, not sound level or idle power consumption.

March 21st, 2007, 04:43 PM	#18
John Miller Major Player Join Date: Oct 2005 Location: Hillsborough, NC, USA Posts: 968	Sorry for not replying sooner - I've just seen this (!) I used our DV processing software to do burn timecode, convert aspect ratio and some color correction on a DV AVI file, writing the processed version to a new file on a separate drive. (i.e., I had two drives - for the FireWire case, I used two separate interfaces - one on the mobo, another on a PCI card). Basically, I ran the process once with both drives connected via FireWire and then repeated it with both connected via USB2.0. The difference in overall processing time was about 1%. The total time for the processing was about 125 seconds. Without any processing - effectively just a straight copy but using the DirectShow infrastructure - the time was about 70 seconds. A direct file copy via Windows Explorer took about 40 seconds. I have to say I expected to see a significant difference but was pleasantly surprised that using USB2.0 did not seem to have a negative impact. When I get a chance, I'll run some simple comparisons using GraphEdit - basically, copy one DV AVI to another using FireWire and then USB2.0. I'll compare those times to the direct copy via Explorer.