Dave Friend
Forum Replies Created
-
An excellent explanation of black level for digital production can be found at
https://pro.jvc.com/pro/attributes/prodv/clips/blacksetup/JVC_DEMO.swf -
kehn,
Thanks for the reply. So I can possibly avoid this problem which NLE are you talking about?
Dave
-
[kenh] “One other thing that we found that caused menus to hang was mixing of DF and NDF footage.”
DF and NDF, as in drop frame and non-drop frame? There is no difference in the footage between these two. DF and NDF only refer to how timecode is handled.
What exactly did you change in regards to DF vs NDF?
-
Chris,
Two other possible problem areas besides replication occur to me. First, it may be an authoring issue. Second it could be the client’s player.
On the authoring side it’s not improbable that an error was made in program flow control logic. I do not know if DVD Producer allows access to PGC/cell pre and post commands and/or GPRMs the way some Sonic products do. If it does, it’s not hard to create bugs that mess up the program flow. (Done it many times myself.) Short of talking with the people who actually did the authoring there aren’t a lot of options for troubleshooting this.
One thing you can look for using a set-top player is if things go wrong after the resume play function has been used. If you run into problems after using this feature then buggy authoring is highly likely.
A couple of software tools that allow you to examine the structure of a DVD, including navigational commands, are ifoedit (https://www.ifoedit.com/) and PgcEdit (https://home.tiscali.be/debie.roland/pgcedit/). These programs are really designed to create DVD copies, ostensibly to create a ‘backup’. PgcEdit is particularly adept at spotting illegal or invalid commands. These require quite a lot of understanding about the technical details of a DVD’s structure to be of value for your needs. But the price is right (free) and perhaps you will learn new things. (I don
-
[Roadkill] “Sonic has recently dropped to price from $899 to $599 to… $199. 🙂 “
When it’s 19.95 my wallet opens. ;^>
-
“The lower the number the higher the quality, but also bigger
file sizes. ideally you would want the smallest GOP possible.”Not necessarily. Small GOP (down to an all I frame structure) only looks good when you have a high enough bit rate to make decent pictures. It’s a bit rate higher than allowed by the DVD spec to be certain. Also, picture quality will will often improve with a longer GOP at the data rates that are typical of a long-form (feature length) DVD. Even if you are not filling the disc and can run at or near the max rate for a DVD there are advantages to using at least a moderate sized GOP.
I
-
A seldomseparate used feature of the DVD spec. Somewhat like alternate audio tracks for different languages there can be separate video tracks with other angles of the video (think multi-cam).
If you dig into the controls of your DVD player you should find the function that allows you to choose which ‘angle’ (a.k.a. video track) to view. Of course you only have the option if an alternate angle is present on the disc. I’ve only ever seen one – a training DVD for hair stylist.
There are some technical requirements for a multi-angle compliant mpeg2 file. I think it mostly pertains to having the same and a consistent GOP structure on each angle. And of course your authoring application must allow for multi-angle DVD creation.
Dave
-
“1) what were the settings used for your encodes? “
2-Pass CBR @ 6.5Mbits. Relatively small GOP attempting to create a structure that looked like /IBPBPBPBPB. Not every encoder allows tight control over the GOP structure or the setting is masked by a ‘user friendly’ interface. But I got as close as I could.
The choice of 6.5Mb/s was made because many experts in the science of MPEG technology (folks who understand the math of Discreet Cosine Transformations the way I understand the math of addition) state that 6.5 is the ‘sweet spot’ for mpeg2. Personally, I just wanted a bit rate that might show some artifacting so I wouldn’t have to use the high-power part of my bifocals to see what was happening.
“2) what type of monitor was used to view the results (9″ TV, or 46” TV, or computer monitor, or ???) “
A Sony PVM-14M2U by way of a Y/C connection out of an off-the-shelf Sony DVD player.
“3) what were the “judging” criteria — quality only, or was speed considered? “
Mostly looking for obvious mpeg2 artifacts like mosquito and contouring noise. The source video isn’t torture test material by any stretch of the imagination. Actually, there aren’t any really complex motions happening so couldn’t really evaluate that aspect. Instead interest was high in seeing how text and other graphic elements were showing up. So much of the work we do is high in that type of material and clients really notice jaggies on their ‘important text messages’.
Speed was not a consideration. The input video was only :30 and so no obvious speed differences were noticed. Then again, I didn’t pay attention. (Sounds like I need to devise another test bench.)
“4) afaik, I thought DVD Workshop used the Mainconcept Encoder (or a version of it)??? “
That may well be. I do not know for sure. The encode settings controls in DVDWS don’t have the Main Concept look and feel that I’ve come to expect. Also, there is no copyright citation for Main Concept technology in the ULead literature, help files or “About” box.
Other questions?
Dave
-
Mike,
I created my own little shoot-out last week using software encoders. I encoded a :30 tv spot using different pieces of software that are either part of DVD Authoring Packages or stand-alone encoders and, in one case, integrated into an NLE.
The spot was edited on an NLE using uncompressed files/footage and a lot of computer generated material too. Four of the encodes used an uncompressed AVI (Pinnacle YUV codec) that was a “mixdown” of the NLE timeline. The fifth was encoded directly from the NLE’s timeline. The same bit-rate was used for each encode and, as much as the various application’s options allowed, all other settings were duplicated.
The five contenders:
1. Ligos (direct from the NLE timeline)
2. TMPGEnc 3.0 XPress
3. Cinemacraft Basic (the $58 version)
4. Main Concept (via Adobe Encore, Main Concept provides the encoding engine to Adobe, and others)
5. Ulead (via ULead DVD Workshop v2.0. They seem to have rolled their own encoder.)The easy winner was CinemaCraft. Multiple sets of eyeballs agreed.
My only downside to CinemaCraft is the lack of support for ac3 audio. ULead, TMPGEnc and Encore can create ac3 audio output.
(Weiss, email me your ftp coordinates and I’ll give you the uncompressed file so you can run it through ProCoder. I’ll also give you the CinemaCraft mpeg file for your evaluation.)
Dave
-
[David Garrett] “How do you mean “control the render settings for motion menus”, “
I mean the bitrate, gop structure… all the parameters that ‘transcode settings’ provide. Seems to me this might be important to folk who want to include moving video in their menus.
(Is it me ? Or does the ‘quality’ of encore’s menu output kinda suck.)
Dave