Vegas 5.0 is SLOWER on a Dual Core system??? Does this make sense?

[Gary Kleiner]

Dual core processing was optimized starting with Vegas 6.

Gary Kleiner

Learn Vegas and DVD Architect

http://www.VegasTrainingAndTools.com

[Mordy Gilden]

Yes, I thought I read that somewhere as well.
So, as far as I understand it, that would mean that Vegas 6 would be faster than Vegas 5 on a dual-core system because of the optimized code. So that explains why the difference isn’t as dramatic as it could be.
However, that doesn’t explain why it would be slower. My new Core 2 Duo is still faster at just plain number crunching than the older AMD chip according to various benchmarking.
So, even without “optimized code”, it should still outperform the 3 year old CPU at least a little bit.

I read regarding other software (some 3D graphics packages) that having hyperthreading and dual core CPU’s can confuse software not designed for it, and it will use a less-efficient instruction set and task threading. In those cases, they said it is better to disable hyperthreading so that both Cores are utilized instead of the HT on a single core. I guess what I’m looking for is perhaps something similar in Vegas.
Perhaps I’m wrong about how I understand this.

[Terje A. bergesen]

However, that doesn’t explain why it would be slower.

In a way it does, depending on what single core system you upgrade from. Each individual core on a dual core system may be somewhat slower than a single core CPU. Since the rendering takes place on one core only, this may be slower than your AMD single core CPU.

This doesn’t surprise me.

—
Terje A. Bergesen

[Anoni Moose]

With Vegas 6, an Intel e6600 (what I have) should be roughly three times faster rendering than your old AMD 3200+ (I had a 2500+ where the difference was more dramatic). I’m not sure how much slower the e6400 is compared to the e6600, but I think I’d have still expected some improvement, maybe 1.5X with only one core being used. You can try and use the process manager (right click on the vegas process) and set it to like just one of the cores, rather than just randomly switching between them. Don’t know if it’d make any difference (suspect not) but worth a try. Is your disk systems (etc) all the same before/after upgrading?

Selected rendertest.veg results (list is maintained by Glenn Chan):

28s – Intel E6600 Core 2 Duo (Vegas 6d; 2.4ghz, 4MB cache)
SOURCE: Emailed submission.

34s – Intel E6300 Core 2 Duo (Vegas 6d; 35s in Vegas 7b)
SOURCE: Guy Bruner @ https://www.sonymediasoftware.com/forums/ShowMessage.asp?MessageID=488694&Replies=31

93s – AMD64 3200+ (2004, so probably old core)
SOURCE: PH125@ https://mediasoftware.sonypictures.com/forums/ShowMessage.asp?ForumID=4&MessageID=256422
99s is Sid Phillip’s report in the same thread.

95s – AMD64 3000+ (2.00ghz, 512kb cache, single channel, socket 754, 2004 core)
SOURCE: ibliss@ https://www.sonymediasoftware.com/forums/ShowMessage.asp?ForumID=4&MessageID=256422

[Mordy Gilden]

According to many real-world tests that I’ve seen, the E6400 is just slightly behind the 6600 for speed. Each of the Core 2 Duo chips are slightly faster than the one beneath it, however all of them (the E6xxx series at least) appear to be in the same ballpark.

Whatever the case may be, I agree with you that even with non-optimized code, it should still be way ahead of my older Athlon system, or at least I would think so.
Did I mention may old chip wasn’t even a 64 model? Yeah, it was the older Athlon XP 3200+, the one that pre-dated the Athlon 64 series. I bought it in 2004, but it was already dropping in price because they were clearing shelves for the 64’s. Most of the tests in that forum you linked to are conducted with the 64. The difference should be even more dramatic for me(I guess similar to your 2500+ vs E6600).

I tried doing that render again and opened up my process manager. It seems that, as speculated, only about 50% of my CPU is being used, even when the thread priority is raised.
This is very annoying. The computer responds faster overall than my Athlon which is nice, but I bought this to speed up rendering time specifically, and that seems to fail to benefit me.

I’m going to try and see if I can install at least a trial version of Vegas 6 to see how much of a difference “optomized code” really makes here.

[Adam Rose esq.]

why not a trial version of v7? at least you can get it straight away, and not likely to be slower than v6

plus better for HDV, if you be using it, etc etc

🙂

[Mordy Gilden]

Honestly… I’m sure 7 is great, but it costs too much for me right now if I don’t need it. You have to keep in mind that it wasn’t so long ago that I bought 5, not to mention I just bought a new computer, and additionally just bought the Adobe Production Studio because certain projects I’m working on require it.
I wasn’t planning on buying a new version of Vegas as well, but if 5 is “incompatible” with my new hardware, I may have to purchase at least 6 (which I can get for substantially less than 7). So basically, if 6 works that much better than 5, I might have to purchase it. But 7 is out of the question for me right now, so why try it?

As far as HD goes… the stuff I use vegas for generally stays standard Def. Anything HD I’d probably end up doing in Premiere Pro.

[Mordy Gilden]

Ok, I installed Vegas 6 and it did perform much better (roughly twice the speed, which would make sense to those who speculated Vegas 5 only used one of the dual cores).

These are my findings:
Vegas 5 on Athlon XP 3200+
11m 40s

Vegas 5 on Core 2 Duo E6400
12m 15s

-> VEGAS 6 ON CORE 2 DUO E6400
6m 4s

Much better!

Conclusion:
I’m impressed with the speed using Vegas 6, however a little bit appalled that Vegas 5 is so poor at CPU utilization. This seems more like a bug that was fixed in Vegas 6.
I say this because I run plenty of software that predated the dual core chips and they still perform well. For example, Caligari Truespace 5 (released back in 2001) does not have any “dual core optimized code”, however the resource manager reports 100% CPU usage when rendering. Vegas 5 reportedly only used 50% during rendering on a Dual Core. This should not be so.

[Anoni Moose]

How well programs use the dual cores isn’t a matter of “bugs”. Just has to do with how many threads they use and how balanced the processing requirements are between the threads. Seeing as how everything was single-processor (Intel’s “hyperthreading” didn’t count, that wasn’t really useful for reasons we won’t go into here) there wasn’t any real point to making things multithreaded (for performance reasons). But there could be software architectural reasons that things were made that way such that the multi-core processors could take advantage of it. Those that did were “lucky”. Some programs probably can’t be easily made to use more than one core efficiently. Depends upon what is being done. Now that there are quad cores, those using only two threads will only use half the cores, in a couple years, when we’ve ten cores… etc. Having unused cores is good,
btw, your computer remains usable even with the processor-hogging application running. Still some processor left to pay attention to your mouse clicks. 🙂