Source files drive and dvd folder drive question.

[drti2]

Does the source files drive and dvd folder drive have to be the same?

i.e. is it faster if I put the dvd folder on the same folder/drive that the source files locate? or put the dvd folder in differrent drive?

[iooi]

Nope source and destination dont have to be the same.
To be honest its a lot easier to keep track if they arnt in the same place. Havent noticed any speed diffrence if they go to diffrent drives and have done it both ways.

[Callan]

To be honest its a lot easier to keep track if they arnt in the same place. Havent noticed any speed diffrence if they go to diffrent drives and have done it both ways.

It does not, or little, if either the drives are on the same physical harddisk, or, if, like it is now, most of the work of demuxing, frequency change, remuxing and adding 'Nav Packs' is done on the drive where the original files are. It would make a difference if the original files where on any other physical drive, and the work of producing both the 48K ._DVD files and the VOBs on another but the speediest partition (the first) of the quickest harddisk.

The best would be, of course, if the originals were on, say, R:\ which in my setup is a sort of archive or attic where nothing is done except storage (the other partition of the separate physical disk for capturing video (Q:\)), the demuxing etc. where done in G:\Temp on the first physical drive, the _DVD files then were written to, say, Q:\, and the finished VOBs to F:\DVD.

Or else, what is the use of a RAID 0 subsystem, except the doubled speed? What use is a separate physical drive for video capture? People say that it is not good to have the swapfile and the growing capture file on the same harddisk. Or why does e.g. Norton Utilities go to the trouble of situating a Windows swapfile (of fixed length) in front of everything else on the first partition of the first physical drive? Speed, me dear.

But what are we talking? Chrissyboy has indicated that he does not want to "go down that road again". Too complicated, he says. I don't quite buy that: if you can write the VOBs to another drive/harddisk, why not other things as well? But then, only those would profit in a worthwhile way who have at least two harddisks in their system. That's not your typical, off-the-shelf setup. What percentage of PC users would have that?

Callan

I just wanted to know now. I only had 1.03 Gb MPEG-1 available, so I tried it out on that.

First, I let SVCD2DVD do its job without Authoring DVD Folders. For all of that happens on the location of the original files anyway.

Then I removed the originals, leaving only the _DVD files in the Q:\!Source folder, and directed SVCD2DVD to put its single VTS into F:\DVD. That took 2 minutes and 20 secs.

Then I did this again, but let it send its produce to Q:\!Source. This time, it took 4 mins and 4 seconds.

Remember: the originals amounted to only a fourth of typical DVD content. Less even. That would come to, say, 9 mins for DVD folders another drive, and 16 mins for the same drive. Quite significant, I would say.

And that is only the making of DVD Folders, VTS(s) and VOBs. The demux/remux business might show similar, or even bigger, gains in speed.

Having an AV@MPG2 capture card which, used as it is in Stream Machine, makes very compact high-bitrate MPEGs, I noticed that somehow they often went out of sync with this demux/remux business. Even if I did it myself.

Now I have bought TMPGEncAuthor. I do not know how exactly it works. I guess 'on the fly' changing the 44k1 to 48K while it builds the VOBs (on another drive, too). I have a feeling that it takes longer, but I have not been able, or willing, to test this. It does have not that great a Menu making department, but on the other hand lets you edit out bad spots, or anything else, like commercials, before it starts. And I don't have Sync problems any more. On the downside, they ask $70 for the program.

Oh, it takes AV@MPG2 files immediately. The ATI AIW produce on the other PC has to be put through TMpgEnc for putting in GOP headers changing the GOP sequence to 15 frames max.

It also takes (regular) VCD, SVCD, MPEG-2 half-D1 and VOB files. If the VCD's are not too irregular, you get a warning which you may ignore.

I do not have it that long, yet. Maybe I have not had enough time yet for discovering its more freaky reactions. If there are any..

Callan

[ChrissyBoy]

I have 3 drives.

I wrote a version of SVCD2DVD which used a "second" drive whereever possible. Meaning for each process the output was always to a drive other than the source. The complexity - which you are not buying - is due to the fact that there are many routes through the program: PS, ES DVD output, AC3, MPA, skip etc. Some of these are determined at runtime (skip audio resample if 48K for example). So you can have the audio stream for a mpg on one drive, the video on another or maybe on the same (bit like quantum physics :P ).

Anyway as with most problems there can be elegent solutions & i think i came up with a rather nice one. The complexity thus was now in the testing/maintenence. Now as I've said before SVCD2DVD is an application which has to work on 98, NT, 2K & XP (and all derivetives), it also has to work in all countries. Then it has to work on all manner of mpegs. Its stability is somewhat of a marvel at times. So i didn't want to introduce a testing nighmare too. Plus it only shaved off 10 seconds for me. It simply wasn't worth it. End of story.