Thanks for the reply. So does this mean that for optimum performance I would have my system on one physical drive, the temp files on a second and the data files on a third? I've been using three SATA drives, so could configure this way. How big is larger than required? I'm still a bit confused about how the whole process works. I edit in Vegas then render and the then use DVD Architect and that seems to be rendering the files again? Perhaps I should be asking what kind of set ups other people use. How does workflow relate to physical disc space?

Kit

Thanks for the reply. What do you mean by dynamic memory space? I've been using three physical drives. The first has system and programs. The second has several partitions. The first partition is for the paging file which I have as a fixed size. The third drive is currently used primarily for backup.

Why should partitions be avoided? Aren't they a useful way of organising work? For example, if Acid needs to look for a file it doesn't give the option to search folders, only partitions. By keeping my loops on a separate partition I can cut down on the search time. Partitions are quicker to defragment and check for errors than whole drives. Drives are so huge nowadays it seems strange to keep them as one vast space?

Kit

Some people like partitions, some don't. I'm in the "don't" group. Whenever i've set up multiple partitions i've found myself later on wishing i had set them up differently as some fill up much faster than i had thought. Making the entire drive one partition avoids the problem.

Don't defragment. It doesn't help, and it can actually be harmful.

If you're checking for errors why wouldn't you scan the entire drive? There's nothing physically separate about partitions, so if one partition is developing trouble there's a good chance the problem could be global to the entire drive.

Thanks - how can defragmenting be harmful - do you have a source for this information? Using an application like spinrite to check errors takes a long time. By using partitions the check can be broken into stages. I guess I'm in the do group.

Kit

Don't defragment; don't bother putting your temp files somewhere else; don't compact the registry.

The performance gains from any of these things will be minuscule, if you can measure them at all. I've spent many hours on this stuff, and these things just don't matter.

You can search on "defragment" in these forums and find dozens of posts arguing the benefits. I always issue the same two challenges: 1. Find a credible test done by an entity other than a disk defrag company that shows any benefit. I still haven't found one. 2. Do benchmarks of several things you think will be sped up by virtue of disk defragmenting, such as rendering, program load times, etc. Then, defragment, re-boot, and do the same tests, in the same order. If you find more than a 2-3% difference, I'll be absolutely amazed.

As to why defragmenting can be harmful, it is an amazingly stressful process, with your disk drive going full bore for several hours, in some cases. Virtually no other process moves the heads this much for such a long time.

I use my last computer eight hours a day, every day, for six years and never defragmented anything.

So if I may ask, how do you organise your drives and files. Do you render and capture to the same drive?

Thanks,

Kit

Kit - I think most everyone recommends that Vegas should be installed on your main boot drive along with the Operating system and all other programs. And most everyone then uses a second harddrive for data (video, audio, image files etc.). There should be some slight speed advantage to doing it this way -- but more important is the advantage of being able to make an "image" of your boot drive incase a complete reinstall of the operating system with all programs and drivers etc. is necessary. You do not want to keep DATA that changes frequently on your boot drive.

I leave my Vegas temp files to the default which is your C drive or boot drive. I've always done it that way with no problem. Never really thought about it being faster to use a different drive for the temp folders.

I do keep my data files on a second harddrive. I actually use a drive caddy so I can plug in SATA drives easily and change them out easily. I don't like to spread things out beyond that. I keep ALL of the data and all elements (including the .veg files) for a Vegas project in one place. One project folder with sub-folders for "narration", "music", "still images" etc. This makes it easy to backup a specific project or move it to a different drive and make sure I have all the elements. So if I want to use some footage from a different project, I copy those files over to the new project. This takes extra time, but it's worth it to keep the project completely together.

Thanks for the detailed reply. I've been working in a similar way. What do you do about rendering finished projects?

Kit

I organize by folder names, not by drives. I've got 7 drives between my two twin PCs and i put new folders wherever there's space. I don't consciously choose any particular drive over another. That means sometimes my source is on the same drive i render to, and sometimes not. I've never noticed any significant speed difference either way. When i've actually done the experiment and timed it i might find that rendering to a separate drive saves half a minute out of a 90 minute render. It's not enough difference to think about.

Generally I render to a third drive for safety. I'm not thinking about speed (and it may have advantages) but more about harddrives failing. Why have the rendered versions on the same drive if you don't have to? But, I admit, sometimes I do as a convenience but I like to render to a different drive when possible.

All this house keeping gets to me sometimes. I'm much more interested in my next idea for a piece than worrying about system and harddrive crashes and being organized! .. But -- consider the time you'll waste when it does happen -- and I do mean WHEN it does happen.

I strongly suggest that anyone who wants to improve performance do tests of their own. For instance, since several people here keep asking about where to put files, why not just try out a few things and see if you can detect a difference?

For instance, I just put 1.5 minutes of SD video on the timeline in Vegas 8.0c. I rendered it to the same drive. Since I didn't do anything to it, Vegas "smart rendered" it. For the first test, I rendered to the same drive on which the video resides. Took 12 seconds on my SATA drive. I then re-did the exact same render and it took 9 seconds. It was faster because most or all of the file had been read into RAM memory. If I did the test with a really long file that couldn't be cached in RAM, I wouldn't see much difference between the first and second test.

I then rendered to a completely different drive. This drive also happens to be really fast. This time the test took three seconds.

Finally, I turned on an external USB drive (using an old IDE 7200 RPM drive) and did the same test with that. This time it took 15 seconds, reflecting the relatively slow speed of both the drive and the USB connection.

I then copied the video to my super-fast drive and then rendered to that drive. 15 seconds. I repeated that and got the same result. I then rendered to a second super-fast drive and got four seconds.

So, the results are pretty easy to understand, and there are two easy-to-understand conclusions:

1. The write speed of the drive to which you render makes a big difference. If you render to a USB drive, a slow drive, or a network drive, it will make your render slower.

2. Rendering to the same drive on which your source video resides slows down the render. The most dramatic illustration of this was when I used really fast drives for both the source video and the render. Rendering to the same drive took fifteen seconds, but rendering to a second drive took three seconds.

But, you say, many people have posted over the years in this forum that rendering to a different drive doesn't make much difference. The results I cite above seems to make their statements completely wrong. Are all these people smoking something?

Well, in some situations their statements are partially correct.

How is this possible? Well, I purposely did my "render" so that Vegas was really doing nothing but copying the video (i.e., a "smart render"). In most cases, however, Vegas will be adding fX, compositing, altering audio, adding titles, etc. This takes lots of compute time. For my 1.5 minutes of video, Vegas might take ten minutes to do all this, and of course the rendering time can become amazingly long if you really do lots of things to your video. Thus, if Vegas is going to take ten minutes to manipulate my ninety second video, the difference of nine seconds isn't going to mean much.

I strongly suggest that anyone who wants to improve performance do tests of their own.

John -- well done. I appreciate the effort you put into that post. Practical, sensible advice.

I like to work in small sections. Render each to a loseless AVI format and assemble the sections for the final render. Smart render is a thing of beauty then.