Comments

Grazie wrote on 4/20/2004, 11:45 PM
Spot, I'll get on the case . . BUT! What time is it there? Don't you ever sleep? - Take it easy . . we need YAH! - Grazie
Grazie wrote on 4/20/2004, 11:46 PM
Sorry, Spot, please give me the rendertest veg link .. It seems to have gone "walk-an=bout" from my web favourites . .. G
riredale wrote on 4/20/2004, 11:47 PM
A nearby thread posted V5 as very slightly faster.
Grazie wrote on 4/21/2004, 12:22 AM
Er . . this can't be correct . .I just did a simple no edit 19 second clip with audio, no cuts, just straight to New Track .. .er .. 19 seconds in 3.5 seconds . .is that about correct? I repeated it too . . 19seconds in 3.5?

Okay . . same clip, 24 splits, made these into dissolves, pre-rendered, - that flashed by! - this pre-rendered 24 dissolved clip printed to New Track .. . er . .. 3 seconds ? ? ? Appears to be very fast indeed . . ppphhoarrhh ..

Grazie
Grazie wrote on 4/21/2004, 12:30 AM
Spot, just reapeated with Vegas 4 - reality check here .. it's about the same. Until I do some more serious rendering, I can't add much. Do you have the rendertest link? - Thank you - Grazie
busterkeaton wrote on 4/21/2004, 12:33 AM
Grazie, go to the sundance media group site. It's under 'tools/aids' in the file section. It's the last file in the list.

Vegas 5 is 7 seconds faster for me.
Grazie wrote on 4/21/2004, 12:43 AM
Uh huh ! - Thanks BK . . I'll get it and post my figures too . .

Grazie
Grazie wrote on 4/21/2004, 1:06 AM
I'm getting 1:36. It's about the same as V4 . . maybe 2 seconds less ..

Just found out something else . . I can't copy and paste events from V5 to V4 .. .Wonder if we can do it the other way around . . ?

Grazie
johnmeyer wrote on 4/21/2004, 10:01 AM
Grazie,

I assume these are DV PAL test, right? Have you also done MPEG2 comparisons? I didn't expect any improvement in DV AVI, but I was hoping for some improvement in MPEG2. I'll be doing my own tests in a few days when my package arrives, and I'll post the results then.

planders wrote on 4/21/2004, 10:22 AM
I'm getting 1:20 on the rendertest with V5, 1:26 with V4. Taken as a percentage, that's a pretty significant performance boost if it carries over to bigger projects. Now we'll need a bigger render test that can help evaluate network rendering, too--the current one only ends up rendering on one machine even if you choose to distribute the rendering.
TheHappyFriar wrote on 4/21/2004, 10:28 AM
I got the same mpeg-2 rendertest times. Someone with a faster comp then mine (XP 1800) could just tripple the length of the render test & run that in V 5 & 4.
SonyPJM wrote on 4/21/2004, 11:32 AM
you can reduce the default length of each segment in the options tab of the network render service application... that allows you to distribute shorter render jobs.
StormMarc wrote on 4/21/2004, 11:38 AM
I have not done any of the official tests but I've been working on a monster project that I just ported to V5 and it's rendering faster. This is a new machine (Asus/P4 3 gig/ 2 gigs ram) and V5 appears to be taking better advantage of the power. The ram being used during rendering is higher as well.

Marc
PAW wrote on 4/21/2004, 12:12 PM

I have not used the render tests but I did post yesterday saying that V5 had my dualies with hyperthreading clocking 100%

further looking at it shows it is just with the magic bullett plugin that this occurs other effects are getting 40-50% on rendering to AVI

Paul