I guess i'll probably end up being known as the lone champion of NON-defragmenting on this forum. oh well. ;)

Generally defragmenting is nowhere near as necessary now as it used to be. Even if the drive has to seek for every single sector it can still deliver data faster than the data rate of a DV stream, so it's not really necessary for successfull captures or print-to-tape operations. While working on a project in Vegas and rendering, fragmented drives can usually serve up data faster than Vegas can make use of it. On the down side, defragmenting a drive is a LOT of extra wear and tear on the motors. It generates a lot of heat that can be damaging to the drive's components (well, so does anything else, but one would expect other drive accesses to be more necessary). It's also several more chances for data to become corrupted or lost.

I'm not saying that defragmenting doesn't help. However, it's probably not as useful as lots of folks seem to think it is. Yes, it can decrease rendering times and help eliminate playback stuttering, but by extremely small amounts. I don't think the extra heat and wear on the motors is worth the return.

My alternative is to have two large video drives. When i'm done with a project on one and can get rid of most of the large source files, i'll copy what i need to keep into a folder on the other, then erase everything from the drive or even do a quick format. The next time around it's the other drive that gets this treament. It's very fast and easy, minimal wear on the drive, and does a more thourough job than defragmenting would anyway.

I'm of the belief that defragmentation does make a big performance difference, although maybe it's less important on today's fast systems. I use Diskeeper on my lowly less than 1G P3 and it does help.

Interestingly, there was a recent article in PC Magazine (I thihk that was the mag) that showed degramention really didn't help much, if at all.

I'm sure the debate will go on forever. For me, I know it helps performance on my 7200rpm drives.

I agree with Chienworks: Defrag is probably not necessary, but my reasons for saying this are somewhat different. Here's my explanation.

First, if you have no files on your drive, there is nothing to defragment (brilliant, eh?).

Second, if you have a bunch of files on your drive, and you then delete them all, you have just defragmented.

Third, if you have only three big files on your drive, each of which takes up 1/3 of the whole disk, then when you delete one file, you have just defragmented 1/3 of your drive.

My point is that video, by its nature consists of a few very large files. Even if you break your files up into small "scenes," when you delete all the files from a project, it frees up a huge chunk of space. By contrast, you almost never delete a large percentage of files from your C: drive and therefore you never get the benefit of this "natural defragmentation" that occurs when you delete files that consume a large percentage of the total disk space. Thus, my contnetion is that the defrag "problem" is going to be far less with video than with program files (or normal WP and spreadsheet files).

Finally, most modern drives have a pretty good-sized buffer, so even if the head has to seek to find a free sector, it is not going to cause the capture to drop frames.

I have always thought that defrag was overblown. Everyone takes it on faith that they have to do it, but I have not seen, in a long time, any attempt to measure the impact on performance provided by a defrag.

I'd recommend putting this under "if it ain't broke" category. Find something else for you (and your computer) to do.

It's probably not going to change anyone's mind, but take a look at

http://www1.execsoft.com/pdf/NSTL-XP_mddvdk.pdf

The test results given here indicate a significant improvement of access time when the hard drive has been fragmented.

Jack

I recently downloaded diskeeper and ran it (was trying to solve a problem that ended up going away of its own accord).

Diskeeper reported one of my drives badly fragmented then showed a display of what performance would be before and after a defrag. The difference was something like 0.5% - and this was from defragging software :-)

Cheers

Thanks for your comments!
"Chienworks" way of working with 2 large videodisks and
alternatively format them in between seems very good to me.
Aje

It's probably not going to change anyone's mind, but take a look at

Man, I love test results. Thanks, Jack!

This paper describes a test that showed a dramatic performance improvement (almost double the speed) after doing a defrag. However, I am not sure the test reflects the results one would actually see in the real world. In particular, the test used a program called "FragmentFreespace.exe" to create the fragmentation. However, a search in both Teoma and in Google does not turn up any mention of such a program, other than in the above-quoted paper. Therefore, it is an almost certainly an in-house utility they developed for this test. That is OK, except we don't know what it really did, and how well it really simulated real-world defragmentation. Given the severity of the performance hit in performing simple tasks like opening a spreadsheet file, my guess is that this utility created some almost pathalogical defrag cases that probably don't reflect the real world. I'm not saying this just to defend my "don't bother to defrag" position, but just simply because it doesn't reflect any experience I have ever had after doing a defrag.

My conclusion is still this: You should definitely do a defrag on your C: drive after you have installed or de-installed a lot of programs. It can make a difference on load times, not so much because of defrag, but because the defrag utility also moves EXE and DLL files to the faster part of the disk. For video files, for the reasons given in my earlier post, I really don't think you're going to see any measurable difference. For normal data files, like the Office XP applications that were the focus in the above-quoted study, it probably won't hurt anything to defrag once every year or so, but I wouldn't bother doing it more often.