Transcode MPEG2 using DVD Rebuilder / DVD Shrink

[ggouweloos]

Hi,

I woulkd like to use DVD-Rebuiler (or DVD Shrink) to transcode a captured MPEG2 file that is on my Harddrive.
Only problem is that all those programs do not proces a single MPEG2 file. They need an authored DVD (VIDEO_TS) folder.
Any idea how I can trick those programs reading and converting my MPEG2 files? I tried renaming them to .VOB but no success there.

[mats.hogberg]

ReJig does mpg.

/Mats

[ggouweloos]

Thanks, I already tried that one.
But ReJig needs the audio to be demuxed first. Also read that DVD Rebuilder would do a better job.

[manono]

DVD-Rebuilder doesn't do any reencoding/transcoding. It's a front end for a bunch of programs that do reencode/transcode. If you want to reencode your MPG, then reencode it yourself. You can't "trick" DVD-RB into taking your MPG unless it's DVD compliant and you author it to a DVD.

What's so hard about demuxing?

[guns1inger]

DVD Rebuilder can transcode - it uses ReJig to do it. However it can only do it with a full DVD structure, as you have found. Rebuilder will only "do a btter job" is you use one of the encoders, and as Manano has pointed out, this you can do yourself.

[ggouweloos]

OK, thanks.

What's so hard about demuxing?

Demuxing (and Muxing) is not hard to do. It just takes a lot of extra time and handling especially when I record between 4-6 hours a day.

I was a bit disapointed with my Sony HD recorder in 2hours-mode. So I wanted to transcode a 1-hours-mode-Sony-recording back to 2hours-mode on the PC and see if that look a bit better in high-motion area's.
But since I am in a process of recording lots of videofootage lately it must be done quick, without (de)muxing would be nice.

[mats.hogberg]

disapointed with my Sony HD recorder in 2hours-mode. So I wanted to transcode a 1-hours-mode-Sony-recording back to 2hours-mode on the PC and see if that look a bit better

Encoding at a too high bitrate and then transcoding down, rarely (if ever) beats encoding at the correct bitrate to begin with.

/Mats

[ggouweloos]

Encoding at a too high bitrate and then transcoding down, rarely (if ever) beats encoding at the correct bitrate to begin with.

Are you sure about that?
Several standalone HD recorders have a specific mode for recording in 'Great' quality. This mostly gives only 30 minutes of video in 4.5GB. On the Sony HD recorder this is called HQ+.
It is NOT possible to record this to a DVD later on. It is always being transcoded or re-encoded because of the hight bitrate. I am not sure if it is using transcoding or re-encoding, but burning the trans/re-encoded version takes al long as the length of the video.

As far as I know those HQ+ modes are used to edit (mostly cut) the video and write it to a DVD afterwards in the desired (lower) quality.

[mats.hogberg]

Yes, I'm very sure of that.
Of course, if you have to do edits (where reencoding is unavoidable), your raw material should be as good as possible, so a "too high for DVD" bitrate makes sense. Even better would have been raw or DV AVI.
But if reencoding can be avoided, it should be avoided. There can never be anything gained by it, qualitywise, unless you have to do some clever filtering wizardry, which requires reencoding (which in turn implies your source has some problems).

/Mats

[ggouweloos]

Hi Mats,

I found this on another forum:

I cannot say if you really gain using HQ+ and HQ first, and then reencode to lower bitrate. As I know, Sony writes additional information about scenes complexity, so it can use it while reencoding to allocate more bits to more complex scenes and less bits to less complex scenes. It is called VBR encoding. Theoretically it should make a better quality. So, if you first record in HQ+ and reencode to eg. SP, it should be better than direct recording the same in SP.

Also I found that the HQ+ is a 15Mbps recording.
As I read the post right the direct SP (2 hours mode) recording is a 1-pass recording and the HQ+ to SP conversion is some sort of a 2-pass encoding. This would cause the endresult to be better than a straight SP recording.

[mats.hogberg]

This would cause the endresult to be better than a straight SP recording.

If you think your ability to encode surpasses Sonys.
Let's do some simple maths:
The signal you record is 100% quality.
You then encode it at 15000 kbps, to let's say 90% quality.
Then you encode it at 4000 kbps which is 50% quality.
You wind up with 50% of 90% of 100% = 45% quality.
Had you encoded to 4000 kbps (50% q) from the beginning, you'd gotten 50% of 100%=50%
No matter how good the HQ+ encode is, it will never reach 100%
If you think you can encode a second generation source and arriving at a better end result than yor hardware can, from a first genereation source, then you're right.
Now, note that transcoding is never as good as reencoding, so if you're really interested in quality (and you're certain you're better at encoding than your harware) you should reencode your HQ+ source down in size.

/Mats

[guns1inger]

Mpeg2 is a lossy encoding method. Every encode discards data from the image. Start with very high bitrates and encoding to lower bitrates discards more data because of the double encoding than encoding once at the correct bitrate.

If the difference visible or at least unacceptable ? That is up to you to determine for yourself. Some people believe that a 7GB movie transcoded by DVD shrink to 4.3GB is indistinguishable from the original.

One thing you should consider however - as you yourself pointed out - these recorders do not transocde to reduce file size. They re-encode. The two methods are different, and so is the quality of output obtained. They also differ in the amount of time and computing effort required.

If you want a quick solution, encode correctly to begin with. Sony also sell DVD based digital video cameras, so just because they do something doesn't make it the smartest or best thing to do. If you must encode at a higher bitrate because you believe it will give you the best outcome, then to transcode afterwards will simply undo all the effort, and produce a quality much lower than if you had simply encoded correctly in the first place.

[ggouweloos]

So the HD recorder uses re-encoding instead of transocoding.
Will the HQ+ re-encoding to SP mode be visibliy better than a direct SP recording.
Assuming the HQ+ to SP conversion is a 2-pass encoding (because the machine knows about the complexity of the scenes) instead of a 1-pass straight SP recording?

I guess I need to do some tests with the same videofootage and try both. I can compare frames on one of my PC's.

[guns1inger]

Remember that comparing still frames is a worthless exercise. because of the nature of mpeg compression, you need to compare scenes of like motion, if not the same scenes, to see what works.

The passage you quoted doesn't explicitly say that it does 2 pass encoding. It could equally apply to simgle pass encoding. So it is up to you to test it. record at HQ+ mode, which is outside DVD spec if it is 15mbps, record it SP mode on the recorder from the HQ+ recording, record it again using SP, this time from the source, and finally, re-encode the HQ+ recording on the PC. For the hell of it, why not transcode one as well.

One question - if you can only get 30 minutes to a disc in HQ+ mode, do you transfer it to your PC in chunks ? And you are worried about the time it all takes . . . . .