How to reduce rendering time?

Videojohn wrote on 3/31/2004, 7:49 AM
I ahe a Pentium 4, a processor of 1,7 Ghertz and 512 MB RAM... and the rendering time and preparing time are very long (and i have to rerender eberything even if I change only a small thing in my menu).
So what can I do to have a render time much lower?
For exemple for a film of about 17 minutes it takes between 1h30 to 2 hours, to transfrom in MPEG2. What can I do to render "real time", like 17 minutes of film equal 17 minutes to encode in MPEG 2.

Thank you

Videojohn

Comments

johnmeyer wrote on 3/31/2004, 9:33 AM
The usual advice is to render the MPEG2 and AC3 files in Vegas. This way the only thing that needs to be rendered when you change menus in DVDA is the menu. This happens in a few seconds. The actual preparation time then consists merely of combining the audio and video files and creating IFO (navigation files).

Once you follow this advice, you may next want advice on how to speed up the preparation time. You can get close to a 2x improvement by always putting your prepared files on a physically separate hard disk from the disk that holds your MPEG and AC3 files.

Hope that helps!
Videojohn wrote on 4/1/2004, 7:15 AM
Thanks John,

But even in Vegas 4, to render a 17 minutes film in MPEG2 (as MPEG 2 for DVDA, without audio) it takes more than 1h30!!!!
It effectively take few seconds, but it is for the sound only.
Can you tell me more

Thank you

Matthieu
johnmeyer wrote on 4/1/2004, 12:20 PM
Rendering time is the time it takes to convert an existing AVI file to MPEG2. If you use all the default settings, and put an AVI file onto the Vegas timeline, AND IF YOU DON'T DO ANYTHING TO THAT AVI FILE (i.e., you don't add fX, or change the video buss track, etc.) then on your speed computer, it should take about 34 minutes to do a 17 minute clip.

However, if you start adding effects, titles, etc., then you must add the time it takes Vegas to create those effects to the time it takes Vegas to convert the AVI file. Some effects (such as simple dissolves between scenes) can be created quickly. Others can take hours, even for just a few minutes of video.

The only thing that can make all this go faster is to purchase a faster computer.
cheroxy wrote on 4/1/2004, 1:41 PM
What do you mean when you say "prepared files?" When I think of that I imagine the mpeg or ac3 files, but you say to have them separate from the prepared files. I'm intersted in doing what you say to have an improvement. Could you explain it further?
thanks,
Carson Calderwood
johnmeyer wrote on 4/1/2004, 7:47 PM
You create your masterpiece in Vegas. You then render those files, either in Vegas or in DVDA, into a video file (MPEG2) and an audio file AC3.

In DVDA, you then take one or more MPEG2 and AC3 files, and develop navigation (chapter points, menus, etc.).

You then use DVDA to prepare the files that will actually be burned onto your DVD. These files have VOB, IFO, and BUP extensions.

What is sometimes confusing is that if, in Vegas, you decide to render to an AVI file instead of MPEG2 and AC3 files, then when you bring that AVI file into DVDA, it must both render (into MPEG2 and AC3) and prepare (into VOB, IFO, and BUP files) all in one go.

The preferred way (for a whole host of reasons) to handle your "workflow" is to render in Vegas, and then import the resulting MPEG2 and AC3 files into DVDA. In this way of doing things, DVDA merely has to "prepare" the VOB files. This preparation process is actually quite simple and merely involves "multiplexing" the audio and video files together (into the VOB files) and creating the navigation files (IFO and the backup of the IFO which is called the BUP file). If DVDA was a full-fledged authoring program (which it should be, given its price) it would also let you combine multiple audio channels and various subtitles during this preparation process, as well as let you create complex navigational logic (e.g., if I have been to this menu before, then do something different this time).

If you are only preparing files, then you can get a dramatic (close to 2x) improvement in preparation time by putting the prepared files in a different physical disk from where you stored the MPEG2 and AC3 files. This is due simply to the way a computer disk works: If it has to read and write from the same physical disk at the same time, it will perform at roughly half the speed compared to just reading, or only writing.
cheroxy wrote on 4/2/2004, 6:07 PM
Thanks John, that is pretty logical. I don't know why I didn't ever think of that.
cheroxy wrote on 4/2/2004, 6:20 PM
Does this also apply to rendering from vegas? Should I keep my imported dv files on a separate drive than the one that I render my mpeg2 and ac3 files to?
thanks,
Cheroxy
pb wrote on 4/3/2004, 6:45 AM
A great time saver is to render your timeline to a completed AVI file. Once you have done that you can use either the built in Main Concept DVDer or the standalone 1.4 version. Separate drives are a must, no argumnent there. I've achieved 0.65:1 rendering times with MC 1.4. Another reason to do the render to AVI then go 1.4 is 1.4 is a bit more tweakable. I use anything from DVD-A to (hehehehehehe) DVD Complete to Adobe Encore for authoring so I'm not all that hung up on needing to export the markers into the DVD compliant MPEG2. AC3? You bet! Get thee hence, bloated files.

Peter
johnmeyer wrote on 4/3/2004, 7:38 AM
Does this also apply to rendering from vegas? Should I keep my imported dv files on a separate drive than the one that I render my mpeg2 and ac3 files to?

Yes. When you render from Vegas (to an AVI, MPEG2, AC3, or anything else) it will go faster if you render to a different physical drive from the one that holds the files used in your Vegas project. The difference will be particularly noticeable in a "cuts-only" project, where Vegas has to do very little rendering. If it truly is "cuts-only," Vegas doesn't touch the bits in the file at all, and most of the "rendering" time is actually just the time it takes your computer to copy from one file to another. When the source video resides on a physically different drive (as opposed to just another drive letter on the same physical drive) from the the destination video, then that render/copy operation can be done using DMA (you DO have DMA enabled, don't you?). DMA lets disks transfer data between themselves without any involvement from the CPU. The disk write speed of the destination disk becomes the main limiting factor. If you have a destination disk that can write at superfast speeds, then this process can really fly.

As for the standalone MainConcept MPEG2 encoder, it is definitely faster than the one built into Vegas, but some of this speed is due to its initial settings. I started a thread on this a month ago:

Mainconcept Standalone Encoder

As you will see, my thinking evolved as I did more tests, and got feedback from other users. My conclusion was that the standalone is a faster, but if you set it to provide roughly equal quality to the Vegas encoder, the speed differences aren't quite as great. What IS different is the quality you can get if you tweak all the advanced settings. The standalone encoder can definitely produce better quality. You can also frameserve (using Staish's plugin) from Vegas to the standalone. This provides an easier workflow, and also avoids the color compression to the DV 4:1:1, if your source footage was not originally DV. Be sure to enable the RGB 16-235 flag if you frameserve, or your results will look "washed out."
cheroxy wrote on 4/3/2004, 9:20 AM
Thanks again John. I always appreciate your well thought, tried and tested answers.
later,
Cheroxy