PC Mag has come out with a review of VV4 and states:

"Unfortunately, performance was disappointing. On a 1-minute test file, which included a 30-second chroma key and spinning title, Vegas took 4 minutes 52 seconds to render before writing to tape, compared with 1:42 for Premiere. Encoding the same project to MPEG-2 took 6:20 for Vegas compared with 3:51 for Premiere."

They liked virtually everything else about it (I did a search and it doesn't look like this review has hit DVinfo yet). For the record I HATE Premiere, but those numbers are pretty bad on a relativistic scale. Any thoughts Vegas fans? Too bad they couldn't do a third compare to Edition 4.5 or 5.0. I'm still on the fence between VV and Edition.

Here's the link. At least for MPEG encoding one question is do they use the same encoder and what kind of quality differences resulted, etc.

http://www.pcmag.com/article2/0,4149,1020103,00.asp

Premiere is the dog. They (CNET) got it wrong.

He was looking for a reason not to give it a 5/5--My guess is that he didn't know how to optimize for rendering---I am not saying that Vegas is a speed demon when it comes to rendering before output---but it certainly is as fast as Premiere, Avid and FCP.

SO who has done a benchmark comparison of the NLE's for rendering speed?

I know that one was done a while back for Mac vs PC rendering the same After Effects projects, and Mac was blown away.

WHOA!

I went looking for the old posting of the benchmark duel, and found they have done an update of the face-off.

See "Benchmark Duel: Mac vs.PC Round 2"

at www.digitalvideoediting.com

The Mac loses again!

The topic here is Vegas is slow, or so says PC magazine. Let's stay on topic, please.

Getting back to the PC Mag review, is there any sense that the conclusion is unfair or that the test video was not appropriate?

Is the reviewer's use of a chroma key something that could slow Vegas down substantially?

I was not expecting such a performance disparity between Vegas and Premiere. Comments on this forum suggest that it is not so. What, specifically, could have produced these results during PC Mag's tests?

This is how I see the difference between Premiere and VV. With VV I can work and "fire for effect" or check my work easily and quickly which keeps the workflow generating. At the end of a editing sequence, I tell VV to render. Which doesn't really matter to me how long, but that it is done in excellent quality. Premiere is the complete opposite. I would have to work render to check which is forever and then work and work and render, though by the time I finish assembling ther eis no need to render, Premiere's bench time is over 3 times longer than the bench time in VV. VV took 4 hours to render a production I created that was an hour and half long in MPEG2 and as an AVI, it may be slower but like Stuart said, which encoder because Premiere and VV can use different encoders...regardless with VV I'm in bed and sleeping or eating or playing with the XBOX or whatever while it rendering so it really doesn't matter to me. :) Stuart has a good point, TMPEG vs Procoder or Mainconcept or whichever render times will vary.

been using Vegas since fall 2000..
i spend my time editing .. i watch previews in RT ( PIV 2.4) if i have 4-5 layers then i do ram preview ...

for me i find my work-flow is very FAST using Vegas.

i don't know if Vegas is fast/slow on rendering as i render when i'm sleeping. i start rendering before i go to bed and when i awake in the AM it's finished. or i do rendering while i'm out shopping/ at a movie etc... and i would render this way using any NLE ...

i've used premiere .. and based on the review i'll give premiere the faster rendering time ( though i never saw it) BUT the time that i've used premiere all that extra time i gained from rendering was wasted on re-booting that interrupted my work-flow ...

One of the problems I had with the review is that he didn't really mention just what he rendered and how his PC was configured. That can make a big difference. Generally the rule of thumb is that the more complex the fx, the longer it takes to render. For instance if you do a motion blur you are going to increase your time.

I think the point is well taken---If you are churning out product under strict time constraints then you worry about rendering. If you can do it while you sleep then who really cares---Its kind of like building the car then waiting for the paint to dry.

While PC Mag is pretty good on computers, I would rather get the views from other experienced professional videographer users rather that pc users.

Nathan, I agree with you and that's surely what we're getting here.

It still doesn't answer the basic question: Is the PC Mag review correct about rendering? If not, why?

Jan Ozer may not be a professional videographer, but he has been doing an excellent job covering digital video for PC Mag. In this case, he examined and mentioned all the key areas of the product (DVD, AC-3, color correction, audio) and compared them to Premiere, which lost in every category. The only ding was for performance, and I think he's the kind of guy who is perfectly capable of analyzing that.

But anyone can make a mistake. Did he?

Will

True, I did not read the article, and in many ways I would rather have my system configured by a PC geek just to get some of the bugs out.

One of the things you want to be sure of is that the reviewer is using the product the way a pro would. Further, are their shortcuts that improve the product more?

I know in the NLE I use (Cinestream) that using the titler can be an unwieldly process. Many would use Photoshop or other products to do their titling and integrated with CS. Also there were other tricks that might require something like AE in other NLEs, that were already in CS.

In general everything else being equal, if they are close in performance, I would regard them as roughtly the same.

Hey all, I'm the PC Magazine reviewer who reported the slow render times for Vegas. I appreciate you all pointing out this could be a mistake, but I need some more input to track down any possible errors. I'll be reviewing Premiere 7 soon, and can post any updated results then.

Just for the record, my rendering test included two major clip segments:

The first had four layers
Bottom video
20 second chroma key
20 second spinning logo (360 degree spin)
A title with animation

Then I dissolved into clip 2 and performed a 40 second 320x240 image pan from the upper left to the lower right. In used original audio in all tracks with one background audio track. Nothing exotic, but some fairly common design elements. I performed all tests on a Pentium IV 3.06 GB with 1 GB RAM running XP pro.

Just so you know, before I run bad performance numbers I check the popular literature and with the vendor. I starting thinking performance might be a problem when reading on forums that users were opening multiple instances of vegas in the background to render while editing in the foreground.

I also spoke to Sonic, who felt that the high level of filtering they performed might be slowing them down. That said, I saw no obvious qualitative difference between Premiere and Vegas.

I also worked with Sonic to verify that my configuration was sound, even buying a new disk drive hoping to improve performance. Nothing changed the results.

Just so you know, I'm aware that both programs used the main concept MPEG encoder. In head to head pure encoding trials without any rendering, performance was similar.

So - I'd be interested in hearing your thoughts about any possible errors in my test procedure.

Thanks in advance.

Jan Ozer

Jan, how great of you to chime in. Do you have plans to review Pinnacle Edition 4.5/5 also? That would be a third likely comparison.

I will be reviewing Edition for EMedia if not PC Magazine. Once Premiere comes out, I'll update all the performance numbers for all programs.

This should all occur over the next three months. Essentially starting over lets us adjust test parameters, so if there are defects or omissions that you folks point out, we can address them then.

Gosh, are you a fallen lawyer? Me too.

Jan

Actually I'm the General Counsel of a multi-billion dollar global manufacturing company. But that's just my day job -- I hope to trade up to working or retiring to the video/film world some day, which is my real joy. Of course until that happens I'll stick to income-producing jobs ;-)

I think what would be helpful would be a performance review across the three NLEs on the same PC processing the same files, etc. -- similar to what you've already done. You might want to post your intended approach and let the Board folks chime in on any problems they see (there's lots of expertise around here). Functionality is obviously a different matter (e.g., advanced background and GPU processing on Edition 5 v. Vegas' superior audio features; Edition's virtually impenatrable user interface v. Vegas' intuitiveness).

I don't know that many folks actually choose between Premiere (more of an introductory NLE product) and Vegas, which has higher end features and complexity. (Probably also why your review sparked such a strong reaction to Vegas fans.) That's why I think the comparison to Edition 5 is important.

oops. Income is definitely good.

I agree that Vegas has vastly superior tools than Premiere, which will be clear in the Emedia review which is 2000 words. Tough to be effusive in 400 words for PC Magazine.

I'm under NDA with Adobe on the next version of Premiere, so I can't go into details. But Premiere 7 could be very relevant to the Vegas market.

I'm was thinking about using the procedure described above unless someone tells me how and why it's incomplete, biased or otherwise flawed.

Jan

hmm...
main concept mpg plugin isnt a bad one, but i do find it has the tendency to leave afew artifacts, particualrly combing, but thats besides the point of this thread...

Main concept is a third party plugin which works.. yet is is slow and when running multiple tracks it WILL bog down...
u can speed it up by resetting the Priority within task amanger, but that defeats the purpose of multi tasking your machine...
Another way to speed it up is to run a Dynamic Ram Preview prior to rendering, this will fill ur ram and give you a bit of a headstart.

In this case, RAM is the all important fact, probably more so than your HDD config.

besides the slow mpg conversion, if you intend on creating edits within Vegas, then porting across thru to DVD Architect, i would actualy recommend using AVI.. then allowing DVD A to convert the mpg for you.
Forget rendering to mpg unless u really need to... who wants to work in a lossey format anyway?

AVI rendering is MUCH faster in vegas....

On a side, note, i actualy prefer finishing my video then allowing it to render overnight.. the environmental conditions are ideal and there no chance in the machine overheating

One thing to note, if u use 2 HDD, make sure u have Vegas installed on your main partition, THEN in another partition have your Vegas temp files, then on the Physically Seperated drive save your finished render..
sounds tedious, but both drives are running at optimal speeds as your temp files are being accessed on one drive while they render within the other..

to tell you the truth, i dont care if it is slow, IMO this is THE best SW editor on the market....
the integration with other Sonic Foundry Apps (sound forge comes to mnd) puts this app in another league... and the inlusion of 5.1 as well as DX plugins (and VST plugs if u have an adapter)

Sure Premiere has its Avid (pardon the pun) followers, but in the end, the only thing keeping Premiere alive right now is the HW and 3rd party support.
There hasnt really been anything innovative about it for a couple of years now...

If only Sonic Foundry bought out some HW support for it..... then it would REALLY wipe the floor with everything else...

I been using Vegas since it first came out, when it was originally designed for music videos.
Back then it was good, now its even better...

Jan, could you post the actual VEG file somewhere so we could look at it? Media is not necessary. It might help us determine if there was an incorrect setting.

Edward:

Both publications frown on doing this.

Going forward, however, when I look at both Edition and Premiere, I will post the results, probably in general terms (not specific times but relative), and the veg file for review.

In the meantime, has anyone done any testing that would shed some light on whether my results are accurate? I understand most folks don't have both applications, but I'm sure some do.

Jan

Jan, you could try reading here:

http://www.sonicfoundry.com/forums/ShowMessage.asp?MessageID=177815&Replies=11&Page=1

The only reason for the request of the VEG file is that there are some very simple things that people (often accidentally) do that can slow down rendering. For example: track "level" sliders not at 100%, resample on when not needed, rendering at BEST vs. GOOD (good is as good as best in MANY cases), unneeded or misused envelopes, etc.... They are just easier to find by looking directly at the VEG file.

If you would like someone to look over the veg file without it being posted on the web, you can find my e-mail through the newsletters at www.jetdv.com/tts.

Thanks, but I'm going to decline for reasons previously stated, and work with you and other interested folks proactively.

Keep in mind that in order to be objective, we can't produce at less than full quality for some editors and full quality for others. For example, with Premiere, I used high quality resampling, though perhaps Adobe would say that wasn't necessary.

Jan

"Keep in mind that in order to be objective, we can't produce at less than full quality for some editors and full quality for others. "


Definitely would never ask that. However, full quality does not necessarily mean rendering at "BEST". "GOOD" is still full quality and is the mode in which a weekly airing TV show I have been editing is rendered.

Here's the official difference:
"Good" uses bi-linear scaling. (Best compromise between speed and quality. This method will produce good results in most cases).

"Best" uses bi-cubic/Integration scaling. (Best image resizing algorithm available in Vegas. Quality differences will be most noticable when using very large stills or stretching small sources)

If you are using high-resolution stills that are being scaled down to video, use Best and not Good to prevent flicker causing aliasing.