Billyboy, your cake analogy is the best, most logically simple explanation of the render time issue I have ever heard!

Never thought of it that way before. Oh well, sometimes concepts are so simple, we like to complicate them. Thanks

MainConcept has its new version of its standalone encoder. I was part of the beta test and found it truely magnificent, you could hardly tell the difference with the original avi. It is very quick even with 2 pass VBR. These features are in the Premiere version of the encoder and I guess they will be added to Vegas shortly, or preferably as a seperate plugin encoder as with the opposition.

For the price difference between Vegas and Premiere (it's closest competitor, price wise) you could BUY another mpeg encoder and still be the same/under the price of Premiere. Also, I use Premiere 6, and the mpeg encoding from that version (Ligos encoder), completely suck. Even if I encode file from a DVCPro with the high settings, it only looks as good as the Vegas 4 encoder at mild settings (around 3-4,000mbps). Sad, isn't it? :)

The whole cake making process is a good analogy, but it doesn't answer the original question. Nor does it necesarily mean Vegas can't be looked down on in that area.

If Premiere is twice as fast and Pinnacle is three times as fast then there needs to be some improvement for sure. Assuming the output is equal.

Take a recent project I did for someone of simply transfering 4 hours of home video to a couple of DVD's. The project only required a little amount of color correction in some places and some cutting out of a few breaks in the video.

In Vegas if the total time for editing was 1 hour plus the 4 hours of data capture, that means after 5 hours I could start rendering before compiling the final DVDs. If the same procedures took about 2 hours in Premiere and Pinnacle then I would be looking at 6 hours when rendering started.

Lets say Vegas rendered at 1:1 ratio where one hour of rendering time would result in one hour of final video. The total amount of time needed before I could compile the DVDs would be quite substantial.

Total times
Vegas = 13:00 hours
Premiere = 8:00 hours
Pinnacle = 7:20 hours

Perhaps the newest MainConcept encoder is what makes the difference who knows and if so, then SoFo needs to provide us with an upgrade. If not the SoFo needs to do some optomizing or something. 2:1 ratio is a fairly large gap.

Of course like BB said, PC Mag isn't the most trustworthy anymore when doing testing. They let to many variables slip through.

JBJones

Folks,

It's nice that we all come up with the marketing spin to help reposition the review so "total speed" is considered, rather than just render speed. This is information that would typically go in a SF reviewer's guide.

However, if the render speed really is slow--I haven't seen anyone dispute it (which is something we all want faster), then I'm sure SF will address it. They've done a wonderful job of anting up the competition numerous times. I'm excited to see what they come up with next (we would all love XP optimized, dual processor enabled, multi-threading optimized, GPU optimized, and then the fastest "quality" render engine to boot). Let's see how SF will respond (unless they get so tied up in Sony big company buracracy (usually endless meetings) that they loose time).

Let's let 'em have it SF... we're all rootin' for ya!

Engineers will tell you of these three things:

1) Fast
2) Cheap (inexpensive)
3) High Quality

... you can have two. We have numbers two and three (and all three if you account for total workflow). I feel quite confident that if improvements can be made in rendering speed WITHOUT sacrificing qualtiy, the programmers in Madison will do it.

Actually, PC Magazine is a great publication... if you understand and manage the review process. I received 7 consecutive editor's choice awards from them--not necisarily because of the best product, but the best reviewers guide and review management process. In fact, we were about to loose the editor's choice, but we kept such good tabs on the review, and had such a good long-term relationship that the reviewer called us and gave us a chance to respond. We fixed the issue within two days, plus did a re-build to add in just a little bit more...and won again. I've seen several reviews (especially of Video Factory) that the product was obviously not well positioned with the press, if so, it would have been positive, versus negative.

It is sometimes a PR issue, not a product issue. The reviewers are as thourough as possible, but can only work with the information they have available. In this case, Adobe may have seen an advantage and put the test in their reviewer's guide, effectively changing the ground rules. That's how it works... better luck next time SF.

I don't think MainConcept and Sonic Foundry are talking to each other anymore. The MC encoders (standalone and Premiere plugin) are both up to version 1.4 (including 2-pass VBR) but the SF version is still 1.0.

It would appear to me that we have been abandoned in that respect.

Jake

PC Magazine was one of many computer related magazines once owned by Ziff Davis, now defunk taken over by Cnet. The running joke was if you advertising in their applications you almost always would get a good review. If you didn't, well you didn't.

Look, the typical magazine "reviewer" isn't technically competent to review either software or hardware. It's like asking for a medical opinion from one of actors that plays a "doctor" on ER or some other TV show. Most of the magazine reviewers and free lance writers that know less about computers than Jay Leno.

Of course like BB said, PC Mag isn't the most trustworthy anymore when doing testing. They let to many variables slip through.

They certainly aren't as good as they were back in the 1980's, but they are about the only magazine left that does real product reviews. In addition, Michael Miller is still editor in chief, and he is the ultimate geeky product guy, and that is reflected in how the magazine does things.

All that said, most reviewers are contract people, and therefore you don't get consistent reviews from one product to the next, and they often don't work hard enough on their benchmarks to really find out what a product is capable of doing.

When I posted my original message, I was hoping that someone had some actual hands-on experience that they could share. I got a lot of SoFo cheerleading -- which is great (count me in as a rabid supporter), but am still in the dark. The rendering time does make a difference. For me, lots of projects are cuts-only -- simple affairs. The rendering is therefore a big percentage of the total time. If it could be cut in half, that would make a huge difference (if no loss in quality).

Regardless the review result, Vegas is finally mentioned together with the other well known products in a major magazine. That's an improvement over the PC world review, which did not include vegas at all.

Ok, I did my own rendering test verse Premiere. Here's the fighters:

NTSC DV AVI (and all the propper specs) 2:04.25 All the renders were at 720x480 res and 16bit stereo audio.

Vegas 4.0. MC encoder built in. Rendering Quality: Good. Video Quality: 31. 5,000cbr. 224kbs audio. All the other mpeg-2 settings I left alone.
total rendering time: 4:21

Premiere 6. Ligos LSX mpeg encoder. Motion Estimation: 20 (highest setting). 5,000cbr. 224kbs audio.
total rendering time: 9:09

So, Vegas beat Premier 6 by 4:48

Of course I don't have Premiere Pro, but I'm downlaoding the new MC encoder tonight (have a dialup connection) so that I can compare those 2 speeds. Vegas RULEZ!

Forgot to say my system specs:
AMD Athlong 1800 XP
512mb DDR 266
Asus A7V8X-X MB
2x80gb HD's @ ATA 100

I also used VHS footage for my rendering tests, sothere was no difference between the way the mpeg's looked. But, at work I've used Vegas and Premiere 6 with DVCPro footage, and the Premiere 6 encoder makes the nice DVC stuff look like crap. The Vegas MC encoder makes it look as good as the DVC (well, 99% as good).

So, Vegas beat Premier 6 by 4:48

TheHappyFriar, this is what I was hoping to hear. I have a hard time believing that there would be such a big difference, because MPEG-2 encoding is a long way down the learning curve, and there shouldn't be that much difference between encoders.

SoFo, if you're watching this thread, perhaps you could comment? If the PC Mag numbers are bogus, you should be all over them demanding a "recount" (oh, oh, here come the Florida jokes).

Well, I don't like comparing apples and oranges, but because i don't have Premiere Pro, I can't give it a render test. But, the latest MC encoder should be the same as the one in PP, so then I'll be able to see how it adds up. I would to more extensive render tests (like overlays, transitions, etc.) but i don't like premiere and don't want to deal with it. :)

There are several different types who make rendering speed an important issue. Two are:

1. In video businesses where fast output in vital, e.g. same day operators like news gatherers. They ususlly have hardware assistance,.

2. People who like to think that they are important and therefore have to have the fastest rendering speed, not realising that there are 101 things to be getting on with while rendering occurs. I won't list all 101, but choices include:

Watching the render line move.
Watching the render line move whilst scratching the left elbow.
Sleep
Design and print video covers and labels
Work on the DVD or CD Rom version interface.
Go and see a client
Go and buy supplies.
Get on with a different project.
Take the dogs for walkies.

... add your own

I depend on Vegas for my living, and render speed has never been an issue. It's not irrelevant, but it's rarely as important as some reviewers make out.

I used the new Mainconcept encoder. Here's the updated results.

NTSC DV AVI (and all the propper specs) 2:04.25 All the renders were at 720x480 res and 16bit stereo audio.

Vegas 4.0. MC encoder built in. Rendering Quality: Good. Video Quality: 31. 5,000cbr. 224kbs audio. All the other mpeg-2 settings I left alone.
total rendering time: 4:21

Premiere 6. Ligos LSX mpeg encoder. Motion Estimation: 20 (highest setting). 5,000cbr. 224kbs audio.
total rendering time: 9:09

MC Encoder 1.04. Settings turned on the highest (search method 11, search range 31). 5,000cbr, 22kbs audio.
total rendering time: 3:10


MC Encoder Vegas Premiere 6
3:10 4:21 9:09


So, the new MC in Premiere Pro is faster then he Vegas encoder (if it's the same as the current MC encoder). Not by 2x or 3x, but by about 1.5x faster. That means a 2 day render in Vegas should take about 32 hours (1.5 days).
SoFo/Sony, did MC abandon you? This is a big time diference. Could we get a new encoder?
Of course Premiere Pro is going to be faster because it came out later (about 6-7 months). They also spend 3 years building it from the ground up (from what was said in DV magazine a couple months ago).

Of course the render time isn't big enough to turne me away from Vegas. It's still easy and FUN to use. :)

May as well jump in here.

The comparison was with Premiere Pro NOT Premiere 6. If you want to do that sort of test dig out VV 2 or 3 and compare. Premiere 6 with the ligos plug-in might compre to VV 3 with the ligos plug-in built in. Premiere 6.5 and VV 4 with the MC mpeg encoder built in would be a better test of speed. But not Premiere 6 and VV 4d.

BB asked "why compare?" and that part is easy - to give people an idea at least. Sure anyone could just simply come out and say that Premiere Pro or VV or Avid Xpress or whatever is the best NLE and leave it at that. But than people would say "yeah but how does this compare to what I am using now?" And beyond that "What I am using now works great for me but man I need faster rendering time does this have it?" without the comparision many people would be flyinh blind. [Well, many are already but that is another thread. ;) ]

VV users have a tendancy to 'want it all' because VV sort of wants people to feel they can 'do it all' with one piece of software. Is this bad? Not really, but if you get caught up in the marketing and the user hype you may forget why you use any product - does it work for what you need it for? If encoding to Mpeg is not a main priority of what you do than speed for rendering to it is not important. If it is important than those comparisons are important as well because you want to get the 'fastest' program to encode.

The specs of a review system may not be the same as your system. The programs running (beyond the actual NLE) may not be the same as what is running on your system. If I have a new hyper threading, clean test system with a gig of RAM and nothing but the OS on it and I install VV chances are I may get better results than an older system with no hyper threading and 256 meg of ram with a lot of programs installed. On the other hand if this new system is running tons or programs and I am multi tasking as I render and the NLE software does not support hyper threading than the render may be as fast, or slower, as the older system. See what I am saying?

Is the new MC plug faster? Probably it is. Is PPro faster? Chances are yes it is - especially compared to Premiere 6 and, I would think, Premiere 6.5. I just rendered out a Video CD from VV 4 and it took about 6 hours for an aproximate 70 minute piece. And this was from pre-rendered video. Is that slow? Yeah it is. Would I like something faster? Yeah. Do I have the latest, greatest PC with the most RAM that I can get? No. Is my primary use for an NLE rendering to Mpeg 1 or 2? No.

Wait a minute... I didn't ask why compare, rather I wondered out loud WHO was doing the comparing.

If some staff writer or worse some freelance guy hired by PC Magazine... give me a break. All modesty aside, what the hell does he know? I always get a kick out of the reviewers and rarely if ever do we learn the guy's/gal's credentials or even their name half the time. For we know his/her day job is a delivery person for Pizza Hut.

True, in the good old days (the 80's) when PC Magazine came out TWICE a month, ran about 500 pages, they did do extensive testing and also they would publish all the testing parameters. Those days are long gone. Now, all the so-called computer magazines are little more than a sham.

There's so many variables to consider when testing "rendering" speeds.

Just a few considerations:

a. Who's rendering engine Lycos, MC, some other, then what version?
b. What OS and what version?
c. What was the source of the "source" file?
d. What was the specs on the test box?
e. Was it rebooted prior to starting the test, other software running?

You get the idea...

The biggest downside with so-called 'lab testing' is how many out there make videos for a lab?

Raise you hand. That's what I thought.

That's exactly my though: "Who makes videos for a lab?" If we want to really compare stuff, you could capture/edit using Vegas with only a $400-500 computer. Then, use another computer with a dedicated hardware mpeg encoder (say, ATI AIW 8500DV). Then you could play back form your $500 computer to another $500 computer, and have an mpeg in just the amound of time it takes to playback your stuff. Then, for a total of about $2000 (500 per computer, 500 for vegas, and 500 misc, for price changes), you get REAL TIME mpeg encoding.

Now, we could buy the new Premiere Pro for $600, a computer that they used in the rendering test (maybe around $2500), and we have a total of $3100.

In that case your mpeg rendering time doesn't matter. To bad reviews don't tell you that!

For the cheap computers I was estimating on p3-667's. I used to use one to edit/capture and it worked perfectly. Only slow render.

filmy/BB, i also rendered with the latest MC standalon encoder, assuming that would be close to the PP one. The results are here:
http://www.sonicfoundry.com/forums/ShowMessage.asp?MessageID=213005&Page=1

It was only about 1.5x faster then the VV4 encoder. Of course with the price of Vegas being about 1/2 of PP, you could take that extra $300 and get a better processor, which would reduce the rendering time gap (strengthing my opinion that reviews should be bases on a set $ amount not on names).

Vegas is an RGB NLE. It will NEVER render as fast as a native YUV NLE. Nothing will fix that shy of a full ground up rebuild. Is it a big deal? Not for me. Why? Because where Vegas IS fast is in the editing interface. That means much more to me than rendering speed.

Craig H.