Comments

GmElliott wrote on 4/19/2005, 7:11 AM
"duh"?

Don't worry William, we've tried rendering out to MPG2 (Mainconcept) and to AVI (Vegas codec). Both are marginally slower than Vegas 5.
Jay Gladwell wrote on 4/19/2005, 8:03 AM

Maybe I did something wrong, but here's what I got.

In Vegas 5, I took a 1-minute talking head, normalized the audio, added color curves and color correction. Then I rendered to .avi, it took 4:05.

In Vegas t, I took the same 1-minute talking head, normalized the audio, added color curves and color correction, same as above. Then I rendered to .avi, it took 4:05.

I guess I shoudl be happy that there is no loss in render time, but I am disappointed that there is no increase as I was expecting.


rmack350 wrote on 4/19/2005, 8:40 AM
I haven't tried anything but a 2 pass mpeg render but that was significantly slower for me. 3 and a half minutes on V5 and 4 minutes on V6.

I'd be curious to know if this can be tweaked away. Otherwise I may put V6 back in the box for a while. At least until I can get/afford a Dual Core Athlon64

Ath64-3200+ skt939
1 GB RAM

Rob Mack
rmack350 wrote on 4/19/2005, 8:44 AM
Okay Bill, but if the Mainconcept isn't multithreaded, should it be slower in V6?

That's a pretty big pill to swallow.

Rob Mack
cbrillow wrote on 4/19/2005, 8:50 AM
I'm with you, GmElliot...

Attention. Attention, Sony -- cleanup in the rendering aisle.

Can we please have some definitive word on what to expect with non-HT or non-dual processor systems?

I've never had a gripe about rendering times, figure it just goes with the territory. But I'm not anxious to go backwards and have it take LONGER...
FrigidNDEditing wrote on 4/19/2005, 8:50 AM
I have seen a delay in render times in MPG - haven't tried AVI - but it's been a pretty short project - may make a diff. on longer veg's - it was just a gradient that was animated. It flew in the first part, but then really slowed down later on. Don't know what the deal is, but with only having a single proc and the inability to upgrade in my Laptop to anything else - I have to say that I am a little concerned with it a little bit - but considering that some of my jobs are hourly :) it could help me make a little more money even!!! :)

All in all though - I would have to say that the end, V6 was worth the upgrade in my book.

Dave
BrianStanding wrote on 4/19/2005, 9:09 AM
Just a thought here: has everyone who is seeing longer render times reset all their preferences and file properties to the same as they were in V5? I'm thinking of:
- preview/render files set to A/V drive, rather than "My Documents" default in File Properties?;
- audio set to 16-bit, 48,000 if using DV files?
- media files and render folders on different drives, if possible?

There's a bunch of tweaks that have been recommended on this forum in the past.

Don't mean to insult anyone's intelligence, here. I know I've often forgotten things that I thought later should have been obvious.
zstevek wrote on 4/19/2005, 9:51 AM
Lets do a apples to apples comparison.

I have a P4 3.0 GHz with HT and I am getting a 40% increase in performance rendering a particular AVI to MPEG-2 on V6 vs. V5

I have XP HOME SP2 installed.

I know SP1 had some funny things going on with the way the OS handled Hyper threading. Maybe the people not noticing an improvment haven't upgraded to XP SP2 yet??
jlafferty wrote on 4/19/2005, 10:22 AM
Just to fill out some details -- I'm running my dual rig with SP1a.
GmElliott wrote on 4/19/2005, 10:37 AM
Just for the record- I am indeed using SP1. I won't upgrade to SP2 until it's CONFIRMED that this IS, indeed, the issue.
smurph wrote on 4/19/2005, 11:06 AM
I'm getting the exact same rendering times out to NTSC DV .avi between V5 and V6. System is a 3.4C HT / XP Pro SP2. Using DSE's "rendertest" to compare timings seems the most appropriate methodology in this case. For me, the time in either version is 1:17. Guess I'll just have to wait for a real dual proc setup to get that time under 45 seconds...!
skibumm101 wrote on 4/19/2005, 11:09 AM
I did three renders V5, V6 1 thread V6 4 threads. I am using a 3,2 HT nothwood core P4 with 2gb ram and raid drives. using XP Pro SP2

V5 2:21
V6 1 thread= :34
v6 4 threads= :40
Bill Ravens wrote on 4/19/2005, 11:41 AM
FWIW....I haven't had any luck getting the network rendering to work in V6. It keeps crashing (as in cold rebooting) my host computer. This IS frustrating. Anyone else having problems with network rendering in V6?
jlafferty wrote on 4/19/2005, 11:46 AM
Can't help you there, Bill.

I just did a render test myself:

V5: 2:04
V6 (4 threads): 1:41
V6 (1 thread): 1:45

Not as big a leap as my other test, but noticably faster nonetheless. Tried going up to 12 threads -- no luck :D

- jim
SimonW wrote on 4/19/2005, 12:11 PM
I just did a test on my system. 3.2ghz P4 HT.

I rendered a 22 sec clip out with the chroma blur filter set to max. Clip was 16:9 anamorphic DV25 pro-scan, and I rendered it out to a DVD Architect MPEG2 file with the aspect set to 16:9 pro-scan, 2 passes, and audio turned on.

Vegas 5: 1 min 55 sec.
Vegas 6 (render threads set to 4): 2 minutes exactly.
Vegas 6 (render threads set to 1): 1 minute 52 sec.
Vegas 6 set to high priority in Task manager with 1 render thread: 1 min 52 sec.
rcrawfor42 wrote on 4/19/2005, 4:10 PM
OK, my test is complete. The project is a 7:36 long song from a concert; the source files are on a firewire drive, the target was a local drive. The target format was DV. There was color correction applied, as well as some transitions across the cuts. Both projects were identical; I pulled up a backup of the V5 project and had already saved a V6 copy last night.

V6: 30 minutes

V5: 33 minutes

This is on a single processor 2GHz P4, 512MB RAM, no hyperthreading.

No real speed-up, but no slow-down, either. I'm satisfied that V6 is no slower.
clearvu wrote on 4/19/2005, 4:20 PM
My system: 3.06ghz P4 HT

Rendered a veg file using V5 & V6. Results as follows:

Vegas 5: 3:16
Vegas 6: 4:13 (threads set to 4)
Vegas 6: 4:03 (threads set to 1)

What gives??? Increased speed was one major draw to me purchasing version 6!

According to MY calculations, V6 is NOT faster than V5
FrigidNDEditing wrote on 4/19/2005, 4:56 PM
Render Times: ala Dave – P4 3.2 HT Proc going from an animated 10 sec. generated media to MPEG2 and AVI in V6 with 1, 2, and 4 threads, and in V5.

V6:
To MPEG2:
1 thread – 29sec
2 threads – 29sec
3 threads – 28sec
4 threads – 28sec
To AVI:
1 thread – 29sec
2 threads – 28sec
3 threads – 28sec
4 threads – 29sec

V5:
To MPEG2:
32sec.
To AVI:
31sec.

Let it be known that for some reason my computer did do an mpeg2 in V6 with 1 thread at 59 seconds at one point (but I retried it later, and it consistently gave the time posted previously).

Dave
B_JM wrote on 4/19/2005, 6:07 PM
the mainconcept mpeg2 encoder is multithreaded - it was in vegas 5 also ....but the only way you would know is if you encoded straight to mpeg and did NOTHING else in vegas ..

B_JM wrote on 4/19/2005, 6:09 PM
Bill Ravens , I have been running network rendering for over a month w/ vegas 6 with no problems ... often left running for days at a time ..

Phil_A wrote on 4/19/2005, 6:27 PM
With an HT processor, DON'T FORGET to make sure that HT is enabled in BIOS.
GmElliott wrote on 4/22/2005, 1:13 PM
I still don't have the answer. I can understand not seeing the performance increase beings I only have a single HT processor and not two physical processors but why is slightly slower?!

If I'm running a P4 3.0ghz w/HT...what should I set my thread count to?
jlafferty wrote on 4/22/2005, 7:03 PM
There's threads here that address your question -- see the "a message about performance gains" thread to figure it out. Speaking generally, your gains are dependent upon the project. Sometimes you'll see V6 marginally slower than V5, but never -- so the hope is -- drastically slower. Other times you'll see it marginally to drastically faster. Again, this is relative to the project in question.

Personally, I've seen one project gain about 10% speed, while two others were around the 40% increase area. Nothing slower yet.

- jim
Jameson_Prod wrote on 4/23/2005, 6:36 AM
Just to add my .02 worth for those interested....

A real world test with my current project (6:30 highlight video, 5 - 10 sec. clips, basic crossfades, text overlays, lower third back ground, 1 video filter, a form of PIP, 5 video tracks, 2 audio tracks with volume events). Computer is P4 2.8HT with 512 RAM using 2 threads on an ASUS P4800 mobo. From timeline straight to mpg.

Default mpg template
V5 - 19:55
V6 - 15:39

My mpg template (increased video aspects for LCD projector playback)
V5 - 29:42
V6 - 25:13

Roughly 20 - 25% improvement on everything I have toyed with.