When I edited my last project I could only afford 1 gig of ram. Suffice it to say, the project became a nightmare to edit as it became more complex on the time line (naturally.) So this time around, for my next big project, I have a little bit of cash to upgrade my system. My question is, should I upgrade my Ram, or should I upgrade to a raid setup? My CPU is a Core Duo @ 2.4 ghz, by the way. Thanks.

For big projects it can sometimes be a good practice to work in smaller chunks/chapters/scenes.

When I edited my last project I could only afford 1 gig of ram. Suffice it to say, the project became a nightmare to edit as it became more complex on the time line (naturally.) So this time around, for my next big project, I have a little bit of cash to upgrade my system. My question is, should I upgrade my Ram, or should I upgrade to a raid setup? My CPU is a Core Duo @ 2.4 ghz, by the way. Thanks.

Sorry to break up the "Glenn" party but I'll chime in here... :)

Where did you see performance suffer the most in your previous project? Also, what is your existing drive and is it at least a drive that is not the same as your boot/OS drive? 1Gb of RAM should be sufficient. I operate with 2Gb and find it's enough. The price of RAM right now is dirt chep and i'd consider maybe doing both. 320Gb SATA2 drives can be had for as little as $70/ea and 2Gb of RAM can be bought for $100.

Jon

Sorry to break up the "Glenn" party but I'll chime in here... :)

Jon

PARTY!!!!!!!!

For big projects it can sometimes be a good practice to work in smaller chunks/chapters/scenes.

For the most part, I did just this. But there comes a time when you must put all the pieces together on one time line, and this is when things choke up.

Sorry to break up the "Glenn" party but I'll chime in here... :)

Where did you see performance suffer the most in your previous project? Also, what is your existing drive and is it at least a drive that is not the same as your boot/OS drive? 1Gb of RAM should be sufficient. I operate with 2Gb and find it's enough. The price of RAM right now is dirt chep and i'd consider maybe doing both. 320Gb SATA2 drives can be had for as little as $70/ea and 2Gb of RAM can be bought for $100.

Jon

This is my system:

Core Duo @ 2.4ghz
1 gig Dual Channel Ram
Three IDE drives, one main for the OS/Vegas and the other one does Audio while the third drive does video only.

By the way, the footage is HDV 1440x1080 24f (Canon) footage. The Audio is a mixture of 16 bit and 24 bit.

I just wanna put about $500 bucks into my system, and thus I was wondering if that money would be better spent on Ram or Raid.

RAM is for pennies at the moment. Play.com are selling 1Gb of the stuff for $18 ($36) delivered!

Since you have $500. Why not spend $100 on good Dual Channel 2X1Gb Sticks. Put them in, pull your existing RAM out and see how it goes. Your next step might be to buy a 6850 that runs at 3Ghz or maybe a Quad Core for $290.

In my testing, low latency RAM doesn't improve performance noticeably. You can pay a premium for RAM and get lower latency stuff (and overclock your existing RAM)... but there isn't a measurable difference.

You do see some small improvements from not crippling memory bandwidth... though this probably isn't something that is noticeable.

To not cripple memory bandwidth, use PAIRs of identical RAM.

http://www.dvinfo.net/conf/showthread.php?s=&threadid=18841

2- Another way of working is just to render out your finished segments, and then stick em into a master sequence.

This is my system:

Core Duo @ 2.4ghz
1 gig Dual Channel Ram
Three IDE drives, one main for the OS/Vegas and the other one does Audio while the third drive does video only.

By the way, the footage is HDV 1440x1080 24f (Canon) footage. The Audio is a mixture of 16 bit and 24 bit.

I just wanna put about $500 bucks into my system, and thus I was wondering if that money would be better spent on Ram or Raid.


Add some ram and get gearshift to do the major editing on dv files and switch at the end of the project to render when you are done. Great thing is being able to render a final dv in a very short time to see how the whole thing plays and if you missed or need to rework anything before the long render.

Stephen Eastwood
http://www.StephenEastwood.com

Quick test: download the freeware utility called RamPage. It installs in your System Tray and shows you how much ram is sitting unused. If you load your project and it drops to "0," then you can speed things up with more ram.

BTW I think you'll be surprised how little ram is needed for many Vegas projects. I only recently went from 1GB to 2GB of ram because I use DeShaker frequently, and on HD clips the second pass of this excellent program can soak up 400MB of ram all by itself.

I also second the motion of using GearShift. I shoot exclusively in HDV, bring in m2t clips with HDVSplit, and immediately generate proxies with GearShift. I edit with the DV proxies, and only for the final render do I GearShift back to m2t. Works great.

Honestly I cannot understand how gearshift is not made mandatory and licensed by sony for every Vegas out there. Its a real speed facilitator even with a dual quad running things, no matter what its faster to do anything especially multiple layers and effects.

Stephen Eastwood
http://www.StephenEastwood.com

Honestly I cannot understand how gearshift is not made mandatory and licensed by sony for every Vegas out there.

maybe cos we aren't all working with HD? I like the idea of not having to pay for things until I need them!

GearShift sounds amazing. It reminds me of old school film editing where you would make a "work print/copy" of the original negative to base your cuts on. Pretty neat stuff. I wish I would have known about this for my last project, it would have saved me a hell of a lot of time.

I'm going to ask a dumb question...

If Gearshift creates a SD version of the .m2t file, does it maintain the same 16:9 aspect ratio of HDV source footage?

Let me get this straight... you capture via HDSplit and it creates an .m2t file (can you use the Sony Vegas video capture utility instead? Does it matter how you get the .m2t file?). Then you load gearshift and create a "DV" equivelent of your .m2t file and start editing away. When the project is all done, you say "render - but use the source .m2t file instead of the DV files"

Is that right?

Jon

Yes and Yes, and you can switch back and fourth from mt2 and smaller dv files anytime throughout to see whats happening on the full size files.

Yes and Yes, and you can switch back and fourth from mt2 and smaller dv files anytime throughout to see whats happening on the full size files.

dang... that sounds like a great app.... and seeing DSE's tutorial on it makes me want to give it a shot.... too bad I don't have any HD footage yet.

I just would like to know how one goes about adding plug-in effects to their original HDV footage using GearShift. This might not seem like a big deal, but when you have a large project on the time line, with many cuts, that include numerous plug ins through many shots, it seems like it might get complicated. No?

duplicate post

Glenn:

There are two functions of GearShift:

(1) It makes it easy to create proxies of m2t clips.

(2) It makes it very easy to switch between the m2t clips on a timeline and their DV proxies you made in step #1.

GearShift is essentially a script that swaps an m2t clip with the equivalent DV clip. It does this automatically for every clip on your timeline.

You do not have to render DV proxies in step #1. You can tell GearShift to render to any file format for which you have a codec. I've played a lot with the PicVideo Mjpeg codec, but concluded recently that the proxies are just not as easy to work with as plain ol' DV. The problem with DV is that going from NTSC DV to MPEG2 (for making a DVD at the end of the process) really kills chroma resolution, so in my final MPEG2 render I use the m2t (HDV) clips as the source. PAL DV doesn't have this issue--you can render to MPEG2 directly from the DV proxy.

GearShift also automatically handles the color space conversion between the HDV world and the SD world. Finally, it builds proxies in the widescreen format, so the resulting proxy looks nearly identical in the preview window (just a very slight horizontal shrink, and, of course, a large difference in sharpness). Every dissolve and color correction made to the proxy will be automatically made to the HDV source upon shifting back.

My first big project years ago was done with DV using Pinnacle's Studio7. Back then (2001) a 40GB hard drive was really BIG. In order to fit my entire project into it, I used a similar proxy process in Studio7. On final render, the program would tell me to load a particular DV tape in the attached camcorder, and it would spend an hour shuttling back and forth, finding the equivalent DV master for the lower-rez proxy on the timeline. It was a fascinating process to watch.

Keep in mind that there is a price to be paid for using GearShift--you need to render additional files, and that takes time and space. Still, I can set it up to render away while I'm off doing something else. No big deal.