miniDV to AVI

[brusno]

I'm backing up miniDV to AVI. I'm using Firewire because I believe that it is better than to use USB 2.0 (I have experienced the video to run less smooth with usb 2, lagging). When I'm keeping the footage as it is, I convert to MPEG2, 8Mbps, deinterlaced, on the fly. But when I have to edit it, I'm capturing it in DV, AVI type 1. I have read about audio-out-of-sync problems with type 2, so I guess that type 1 is a better choice?
I also have three options when capturing video for editing:
1. Using winDV with AVI type 1.
2. Using Nero Video with Lagarith codec (type 1), I think Lagarith only use type 1?
3. Using Nero Video with HuffYUV codec (type 1), I think HuffYUV only use type 1 too?

So the question is which of these are preferable to use?, or are they pretty much the same?
(After editing, I save the video to MPEG2, 8Mbps)

I don't see any difference in picture quality. But when I'm capturing in MPEG2 (8Mbps), editing it, and then saving it to MPEG2 again, THEN I can see that it is worse than the 3 options mentioned.

Another question is, if the MPEG2 capture is at the same bitrate as the AVI captures (24 Mbps) before editing, will then the resulting MPEG2 video (8Mbps) be about the same visual quality as in the 3 options above?

[pippas]

Using method 1 (WinDV via Firewire) is a direct digital copy of the data on your miniDV tape. No change . No quality loss at all.

[jagabo]

Another question is, if the MPEG2 capture is at the same bitrate as the AVI captures (24 Mbps) before editing, will then the resulting MPEG2 video (8Mbps) be about the same visual quality as in the 3 options above?

Capturing as DV AVI is simply a repackaging of the DV video and PCM audio that's recorded on the tape, an exact copy except for the AVI wrapper. Encoding to very high bitrate MPEG 2 would entail very little loss of quality but there's no reason to do so. You couldn't use it directly on DVD because DVD doesn't allow such high bitrates. Most editors work better with DV than with MPG.

[Cornucopia]

BTW, you started throwing in "Type1", etc. into the jargon: Type1 AVI (single "IVAS" stream) is only pertinent with DV-AVI files. Just about ALL other AVI files are standard Type2 ("VIDS"+"AUDS" streams). DV-AVI could be either (I have used both often). You shouldn't have to worry about Type1 vs. Type2 when you are capping via Firewire. Just go with whichever type is native to your capping app (though Type2 is more compatible with more 3rd party apps). Out-of-sync problems usually only occur in situations where you cap your audio separately from your video.

If you are indeed "backing up", you shouldn't be doing ANY change to the captured DV stream (whether Type1 AVI, Type2 AVI, or raw DV). Just save the file as-is. Why are you saving to MPEG2 after? If you plan on editing down the road, as jagabo has already stated, DV is much easier to work with and retains more of the original camera quality than MPEG2. MPEG2 would make sense if you plan on using 'em as source assets for DVD-Video. Not for archiving (quality = original = DV stream), not for small size/distribution (h.264 MP4/MKV would make more sense there).

Scott

[brusno]

You couldn't use it directly on DVD because DVD doesn't allow such high bitrates.

The high bitrates would only be when editing, and then save it to DVD at 8Mbps.

Most editors work better with DV than with MPG.

That is a good reason to edit in DV instead of MPEG2.

(though Type2 is more compatible with more 3rd party apps).

I think type 2 was the preferable in the late 90's and beginning of the millenium. Because less CPU power was needed to process this. If so, this could explain why type 2 is more compatible with more 3rd party apps. I have noticed that in VirtualDubMod, the type 2 is playing with sound, while playing type 1, the sound is absent.
Does this mean that if I'm editing a type 1 DV in VirtualDubMod, I will loose the sound when saving the edited file?. And what about type 2, will only video be edited while the audio is the same length? (Since they are separated).

If you are indeed "backing up", you shouldn't be doing ANY change to the captured DV stream (whether Type1 AVI, Type2 AVI, or raw DV). Just save the file as-is. Why are you saving to MPEG2 after?

I'm keeping the source file (DV-AVI) as it is, in addition to the edited MPEG2. I even keep the original tapes. This is because if alternative methods would come to mind, then I have the chance to fetch the DV video again, from the original source

I'm saving to MPEG2 after, because I usually show the results to family and friends on DVD. This is because everyone has players that can show DVD. While not everyone has blu-ray players yet, to show MP4.

But I will also encode it to MP4 (1.5Mbps). The advantage of MP4 is that it is taking less space while the quality is pretty much the same as mpeg2 (8Mbps). And you can get a lot of film on blu-ray discs if it is in MP4

I presume that the quality of the MP4 videos are getting better if I encode it directly from AVI (and not AVI-MPEG2-MP4)?

[2Bdecided]

1. Using winDV with AVI type 1.

That is a perfect lossless copy of the compressed video that's on the tape.

2. Using Nero Video with Lagarith codec (type 1), I think Lagarith only use type 1?

That takes the compressed video from the tape, uncompresses it (makes it larger; with luck it will not change the actual picture), and then uses Lagarith lossless compression to store it. So you make the file size much bigger, for no advantage at all.

3. Using Nero Video with HuffYUV codec (type 1), I think HuffYUV only use type 1 too?

That takes the compressed video from the tape, uncompresses it (makes it larger; with luck it will not change the actual picture), and then uses HuffYUV lossless compression to store it. So you make the file size much bigger, for no advantage at all.



The advantage to DV type 1 is that it's slightly smaller. The advantage to DV type 2 is that it has wider compatibility.

I always use type 2.

AVI files containing other codecs have a video stream and an audio stream inside. They do not have the two interleaved into one stream (like DV type 1), and they do not have two copies of the audio (like DV type 2).

Cheers,
David.

[jagabo]

I think type 2 was the preferable in the late 90's and beginning of the millenium. Because less CPU power was needed to process this.

The data arriving at the firewire port has audio and video multiplexed together. Type 1 DV AVI stores that muxed data as a single stream in the AVI file, marked as a video stream. Type 2 DV AVI makes a separate copy of the audio data and stores it as a second stream, marked as an audio stream. So there are two streams in a type 2 DV AVI, one marked video, containing the original video and audio streams, and another marked audio with a second copy of the audio stream.

Type 1 is less work to create and was popular when computers may not have been powerful enough to demux (DV) and mux (AVI) a second copy of the audio in real time. But programs have to be specially written to use type 1 DV AVI -- they have to know how to get the audio out of the single "video" stream, VFW does not handle it for you. Type 2 is easier to decode since no special knowledge of DV is necessary. Windows' VFW API takes care of it for you.

[brusno]

Then I will try type 2. I can use winDV with type 2 too. And I discovered that I can edit the video with avisynth instead of nero, with the trim function. I read the framerates in VirtualDubMod and put it in the trim(). And after that I encode it to mpeg2 with TMPGEnc, because it accepts .avs files as input. This is an easy and elegant way to do it
But then a new question arises; is it best to deinterlace in avisynth or in TMPGEnc? Or not at all, since it's going out to DVD format?

[Cornucopia]

You should know: If you use AVISynth to edit/process your footage, it will ALWAYS decode to uncompressed, so you are no longer keeping the original DV stream. That's fine if your ONLY edited version is supposed to be MPEG2 (where you can frameserve AVISynth to HCEnc, etc), but if you use VdubMod in between, it will need to save to an intermediate file (unless you are using Vdm only as an AVS script editor).

You should NOT be deinterlacing if your input is DV and your output is standard DVD-Video. Deinterlace only for output targets that only support progressive (youtube, etc.). And only on THOSE versions, not on the master.

Scott

[jagabo]

You should NOT be deinterlacing if your input is DV and your output is standard DVD-Video. Deinterlace only for output targets that only support progressive (youtube, etc.). And only on THOSE versions, not on the master.

The exception would be telecined film (for example, if you recorded a broadcast movie with your DV camcorder). In those cases you should inverse telecine to return to the original 23.976 fps film frames. Then encode progressive MPEG 2 at 23.976 fps with pulldown flags.

[Cornucopia]

Agreed. But I didn't think that scenario would come up with the OP, so I left it out. Bad Scott! Bad!!

Scott

[Richard_G]

Except the OP is in Norway so is using PAL and not NTSC so pulldown is not relevant.

[brusno]

but if you use VdubMod in between, it will need to save to an intermediate file (unless you are using Vdm only as an AVS script editor).

I'm using VdubMod to see the numbers of frames where I want to cut the video. I think if I save something from VdubMod, and then encode it to MPEG2 with avisynth in the background, then the color format is changed from YUV to RGB and then back to YUV again. And that would be bad wouldn't it?
But I may use VdubMod to demux the sound, to be used in the making of MP4. Yes I'm making both MPEG2 and MP4 out of DV.

Except the OP is in Norway so is using PAL and not NTSC so pulldown is not relevant.

Yes I'm in PAL-country, so it is PAL-DV which I'm working on. But I think it's a good thing to mention other scenarios which would apply for other users too, as Jagabo did.

[jagabo]

I think if I save something from VdubMod, and then encode it to MPEG2 with avisynth in the background, then the color format is changed from YUV to RGB and then back to YUV again. And that would be bad wouldn't it?

VirtualDub in Direct Stream Copy mode will give you frame accurate cuts with no reencoding of DV AVI. Ie, no loss of quality.

[brusno]

I think if I save something from VdubMod, and then encode it to MPEG2 with avisynth in the background, then the color format is changed from YUV to RGB and then back to YUV again. And that would be bad wouldn't it?

VirtualDub in Direct Stream Copy mode will give you frame accurate cuts with no reencoding of DV AVI. Ie, no loss of quality.

I tried Direct Stream Copy, and the avi-file which was originally 2.74 GB, turned in to 10.3 GB! Is it because it is recompressing the avi file? I think it's better to keep approximately quality and keep the filesize approximately as the original. But with all the other options in VDubMod, I think it will be color format convertion?

It may be the avisynth processing which makes the file bigger, because when using Avisynth then the AVI is uncompressed, as Cornucopia pointed out.

[2Bdecided]

I think you need to slow down, and do some more reading and testing - because at the moment you haven't got a clue what you're doing!

You can do lossless cuts-only edits in VirtualDUB (you don't need the old "mod" version), save directstreamcopy to DV-AVI (should be very fast, and of a size proportional to how much of the original you've kept - no more, no less), drop than AVI into TMPGEnc and encode using suitable settings for DVD. Job done. You don't need AVIsynth unless you're actually doing something to the video!

Cheers,
David.

[sanlyn]

I agree with 2Bdecided. Something is amiss here (and I don't know why you're using VDubMod, but it should work with DV-AVI). It would be a good idea to get a copy of MediaInfo (free) and post its report -- you can save MediaInfo's output as text, or copy it right off the display window. It would tell us a lot about what you're working with.

[jagabo]

And we're talking Video -> Direct Stream Copy, not Audio -> Direct Stream Copy (though you want to use the latter too). With Video -> Direct Stream Copy you can't perform any filtering. You can only cut and paste. With DV (or any other intra frame codec) these cuts and pastes can be frame accurate. With inter-frame codecs like Divx/Xvid you can only cut/paste on key frames.

[brusno]

It would be a good idea to get a copy of MediaInfo (free) and post its report -- you can save MediaInfo's output as text, or copy it right off the display window. It would tell us a lot about what you're working with.

Ok, I'll post mediainfo from 3 files. Number 1 is the original (captured in winDV, type 2), number 2 is Direct Stream Copy with avisynth trimming (clipping) which got huge. And number 3 is Direct Stream copy without avisynth.

NUMBER 1 (Original):
General
Complete name : C:\Prosjekt akvariet\Akvariet.avi
Format : AVI
Format/Info : Audio Video Interleave
Commercial name : DVCPRO
Format profile : OpenDML
File size : 2.74 GiB
Duration : 13mn 5s
Overall bit rate mode : Constant
Overall bit rate : 30.0 Mbps
Video
ID : 0
Format : DV
Commercial name : DVCPRO
Codec ID : dvsd
Codec ID/Hint : Sony
Duration : 13mn 5s
Bit rate mode : Constant
Bit rate : 24.4 Mbps
Encoded bit rate : 28.8 Mbps
Width : 720 pixels
Height : 576 pixels
Display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 25.000 fps
Standard : PAL
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Bottom Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 2.357
Time code of first frame : 00:04:15:08
Time code source : Subcode time code
Stream size : 2.63 GiB (96%)
Encoding settings : wb mode= / white balance= / fcm=auto focus
Audio
ID : 1
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Signed
Codec ID : 1
Duration : 13mn 2s
Bit rate mode : Constant
Bit rate : 1 024 Kbps
Channel(s) : 2 channels
Sampling rate : 32.0 KHz
Bit depth : 16 bits
Stream size : 95.6 MiB (3%)
Alignment : Aligned on interleaves
Interleave, duration : 40 ms (1.00 video frame)
Interleave, preload duration : 40 ms



NUMBER 2 (Direct Stream Copy with avisynth trimming (clipping):
General
Complete name : C:\Prosjekt akvariet\Akvariet 2.avi
Format : AVI
Format/Info : Audio Video Interleave
Format profile : OpenDML
File size : 10.4 GiB
Duration : 11mn 49s
Overall bit rate : 125 Mbps
Writing application : VirtualDubMod 1.5.10.2 (build 2542/release)
Writing library : VirtualDubMod build 2542/release
Video
ID : 0
Format : YUV
Codec ID : YV12
Codec ID/Info : ATI YVU12 4:2:0 Planar
Duration : 11mn 49s
Bit rate : 124 Mbps
Width : 720 pixels
Height : 576 pixels
Display aspect ratio : 5:4
Frame rate : 25.000 fps
Standard : PAL
Color space : YUV
Chroma subsampling : 4:2:0
Compression mode : Lossless
Bits/(Pixel*Frame) : 12.000
Stream size : 10.3 GiB (99%)
Audio
ID : 1
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Signed
Codec ID : 1
Duration : 11mn 49s
Bit rate mode : Constant
Bit rate : 1 024 Kbps
Channel(s) : 2 channels
Sampling rate : 32.0 KHz
Bit depth : 16 bits
Stream size : 86.6 MiB (1%)
Alignment : Aligned on interleaves
Interleave, duration : 40 ms (1.00 video frame)
Interleave, preload duration : 500 ms





NUMBER 3 (Direct Stream copy without avisynth trimming)
Complete name : C:\Prosjekt akvariet\Akvariet 3.avi
Format : AVI
Format/Info : Audio Video Interleave
Commercial name : DVCPRO
Format profile : OpenDML
File size : 2.73 GiB
Duration : 13mn 5s
Overall bit rate mode : Constant
Overall bit rate : 29.8 Mbps
Writing application : VirtualDubMod 1.5.10.2 (build 2542/release)
Writing library : VirtualDubMod build 2542/release
Video
ID : 0
Format : DV
Commercial name : DVCPRO
Codec ID : dvsd
Codec ID/Hint : Sony
Duration : 13mn 5s
Bit rate mode : Constant
Bit rate : 24.4 Mbps
Encoded bit rate : 28.8 Mbps
Width : 720 pixels
Height : 576 pixels
Display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 25.000 fps
Standard : PAL
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Bottom Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 2.357
Time code of first frame : 00:04:15:08
Time code source : Subcode time code
Stream size : 2.63 GiB (97%)
Encoding settings : wb mode= / white balance= / fcm=auto focus
Audio
ID : 1
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Signed
Codec ID : 1
Duration : 13mn 2s
Bit rate mode : Constant
Bit rate : 1 024 Kbps
Channel(s) : 2 channels
Sampling rate : 32.0 KHz
Bit depth : 16 bits
Stream size : 95.6 MiB (3%)
Alignment : Aligned on interleaves
Interleave, duration : 40 ms (1.00 video frame)
Interleave, preload duration : 500 ms
t Stream Copy without avisynth):General
Complete name : C:\Users\Tommy\Pictures\Prosjekt akvariet\Akvariet 3.avi
Format : AVI
Format/Info : Audio Video Interleave
Commercial name : DVCPRO
Format profile : OpenDML
File size : 2.73 GiB
Duration : 13mn 5s
Overall bit rate mode : Constant
Overall bit rate : 29.8 Mbps
Writing application : VirtualDubMod 1.5.10.2 (build 2542/release)
Writing library : VirtualDubMod build 2542/release
Video
ID : 0
Format : DV
Commercial name : DVCPRO
Codec ID : dvsd
Codec ID/Hint : Sony
Duration : 13mn 5s
Bit rate mode : Constant
Bit rate : 24.4 Mbps
Encoded bit rate : 28.8 Mbps
Width : 720 pixels
Height : 576 pixels
Display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 25.000 fps
Standard : PAL
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Bottom Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 2.357
Time code of first frame : 00:04:15:08
Time code source : Subcode time code
Stream size : 2.63 GiB (97%)
Encoding settings : wb mode= / white balance= / fcm=auto focus
Audio
ID : 1
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Signed
Codec ID : 1
Duration : 13mn 2s
Bit rate mode : Constant
Bit rate : 1 024 Kbps
Channel(s) : 2 channels
Sampling rate : 32.0 KHz
Bit depth : 16 bits
Stream size : 95.6 MiB (3%)
Alignment : Aligned on interleaves
Interleave, duration : 40 ms (1.00 video frame)
Interleave, preload duration : 500 ms




Seems like the Aspect Ratio is destroyed when using direct stream copy with avisynth. 5:4 instead of 4:3.


@2Bdecided:
I have been using VDubMod because of its ability to open MPEG2 files. I recall that this couldn't be done with VDub. But perhaps it is possible now?

[jagabo]

Did you open the DV AVI file directly with VirtualDub? Not indirectly via an AviSynth script? The latter would convert to uncomrpessed YV12 or YUY2. Video #2 shows it has been converted to uncompressed YV12.

VirtualDub can open MPEG 2 files using the MPEG 2 source plugin. Also with the DirectShow or ffdshow source plugins. The preferred method of opening MPEG 2 videos with any version of VirtualDub is to use AviSynth and Mpeg2Source() opening a index file built by DgIndex.

[poisondeathray]

I have been using VDubMod because of its ability to open MPEG2 files. I recall that this couldn't be done with VDub. But perhaps it is possible now?

vdub has a mpeg2 plugin for many years now (along with many other input plugins)
http://gral.y0.pl/~fcchandler/

[brusno]

Did you open the DV AVI file directly with VirtualDub? Not indirectly via an AviSynth script? The latter would convert to uncomrpessed YV12 or YUY2. Video #2 shows it has been converted to uncompressed YV12.

Number 2 is the original avi opened in VirtualDubMod via Avisynth script (using AviSource()), and then saved again with Direct Stream Copy. That's probably why it has been converted to YV12.
Number 3 is the original avi opened directly in VirtualDubMod, and then saved again with Direct Stream Copy. This is of approximately same file size as original.

It's probably better to use "cut" in VirtualDub or VirtualDubMod instead of using trim() in Avisynth. Because of the huge filesize which Avisynth is causing, and the color convertion. Because then I can save it to AVI again, very similar to the original. Only edited this time.
However, I used .avs (avisynh) in TMPEGEnc too when converting to MPEG2, and it worked well. The MPEG2 file is 700 MB (8 Mbps).
Mediainfo for this:

General
Complete name : C:\Prosjekt akvariet\Akvariet.mpg
Format : MPEG-PS
File size : 702 MiB
Duration : 11mn 49s
Overall bit rate mode : Variable
Overall bit rate : 8 299 Kbps
Writing library : encoded by TMPGEnc (ver. 2.525.64.184)
Video
ID : 224 (0xE0)
Format : MPEG Video
Format version : Version 2
Format profile : Main@Main
Format settings, BVOP : Yes
Format settings, Matrix : Custom
Format settings, GOP : M=3, N=13
Duration : 11mn 49s
Bit rate mode : Variable
Bit rate : 8 000 Kbps
Width : 720 pixels
Height : 576 pixels
Display aspect ratio : 4:3
Frame rate : 25.000 fps
Standard : PAL
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Bottom Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.772
Time code of first frame : 00:00:00:00
Time code source : Group of pictures header
Stream size : 672 MiB (96%)
Writing library : TMPGEnc 2.525.64.184
Color primaries : BT.601 PAL
Transfer characteristics : BT.470 System B, BT.470 System G
Matrix coefficients : BT.601
Audio
ID : 192 (0xC0)
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 2
Duration : 11mn 49s
Bit rate mode : Constant
Bit rate : 192 Kbps
Channel(s) : 2 channels
Sampling rate : 48.0 KHz
Compression mode : Lossy
Stream size : 16.2 MiB (2%)

You can do lossless cuts-only edits in VirtualDUB (you don't need the old "mod" version), save directstreamcopy to DV-AVI (should be very fast, and of a size proportional to how much of the original you've kept - no more, no less), drop than AVI into TMPGEnc and encode using suitable settings for DVD. Job done. You don't need AVIsynth unless you're actually doing something to the video!

Yes you're right, it's very easy to use the "cut" in VirtualDUB, and the advantage is that I'm getting an AVI-file with roughly the same file-size as the original. Depending on how much I'm cutting of course.
I've looked at both VirtualDUB and VirtualDubMod, and they are very similar. I liked the timeline in VirtualDUB better than in VDubMod though. But, what can VDUB do which VDubMod can't?

And TMPEGEnc is generating MPEG2 files. But what program is good to use when putting these files on a DVD (to make it DVD compatible with vob files)? And perhaps making a menu.
Nero is easy to use for this. But then I have to open the AVI file, edit it, and then convert it to DVD in Nero (all operations).
I can't use Nero to make DVD out of the MPEG2 files made in TMPEGEnc, because Nero will encode it one more time unnecessary.
Perhaps I can use ConvertXtoDVD for this. But this program will probably also re-encode the MPEG2 files from TMPEGEnc?

Update: I discovered that I can use the AVI file in ConvertXtoDVD directly. Then it will be only one encoding. But the question is which will do the best converting, TMPEGEnc or ConvertXtoDVD.
Another discovery is that when editing and processing in Nero, I can choose .ac3 audio. I don't have that option in TMPEGEnc. A possible workaround might be to demux the MPEG2 file and exchange the audio to .ac3 which can be encoded separately. I.e. in BeSweet.
ConvertXtoDVD can convert audio to .ac3, and also AVStoDVD.
I guess I have to do some testing, and comparing to find out what quality the programs are able to deliver.
I compared Nero and TMPEGEnc, and thought they were pretty much equal.

[brusno]

Nero-sample:
Nero_akvariet_2.demuxed.m2v

TMPEGEnc-sample:
Akvariet_TMPEGEnc.demuxed.m2v

[sanlyn]

It's probably better to use "cut" in VirtualDub or VirtualDubMod instead of using trim() in Avisynth. Because of the huge filesize which Avisynth is causing, and the color convertion. Because then I can save it to AVI again, very similar to the original.

You have it wrong. When you open a vid in VirtualDub without the benfit of opening in AviSynth, then make cuts, and save with directstreamcopy, there is no color conversion and no decompression.

If you open a file in Avisynth and use Trim(), and use Directstsreamcopy in VirtualDub to save, it works the same way as using "cut" in VirtualDub with directstreamcopy. There should be no difference. However, if you used Full Processing to save the AVisynth trimmed vid, then you converted to RGB32 and saved as uncompressed. That file will be several times larger than the others.

And TMPEGEnc is generating MPEG2 files. But what program is good to use when putting these files on a DVD (to make it DVD compatible with vob files)? And perhaps making a menu.
Nero is easy to use for this. But then I have to open the AVI file, edit it, and then convert it to DVD in Nero (all operations).
I can't use Nero to make DVD out of the MPEG2 files made in TMPEGEnc, because Nero will encode it one more time unnecessary.
Perhaps I can use ConvertXtoDVD for this. But this program will probably also re-encode the MPEG2 files from TMPEGEnc?

You need to do a little more reading. VOB is MPEG reorganized and renamed for DVD disc. There is no re-encode from MPEG to VOB.

I'd stay away from Nero's encoder.

Update: I discovered that I can use the AVI file in ConvertXtoDVD directly. Then it will be only one encoding. But the question is which will do the best converting, TMPEGEnc or ConvertXtoDVD.
Another discovery is that when editing and processing in Nero, I can choose .ac3 audio. I don't have that option in TMPEGEnc. A possible workaround might be to demux the MPEG2 file and exchange the audio to .ac3 which can be encoded separately. I.e. in BeSweet.
ConvertXtoDVD can convert audio to .ac3, and also AVStoDVD.

Don't know what version of TMPGEnc you're using. The older freebie TMPGEnc, as well as the old TMPGenc Plus 2.5, don't output Ac3 audio. TMPGenc Mastering Works, which is an HD and SD encoder, does. If you've done any intermediate processing with your source, your audio should be uncompressed PCM anyway and left that way until your final encode.

I compared Nero and TMPGEnc, and thought they were pretty much equal.

Depends on how you encoded. But generally speaking, old stand-by TMPGEnc would look better when properly used.

[poisondeathray]

If you open a file in Avisynth and use Trim(), and use Directstsreamcopy in VirtualDub to save, it works the same way as using "cut" in VirtualDub with directstreamcopy. There should be no difference. However, if you used Full Processing to save the AVisynth trimmed vid, then you converted to RGB32 and saved as uncompressed. That file will be several times larger than the others.

This isn't true, because avisynth frameserves uncompressed video. Using direct stream copy will give you uncompressed video if you used avisynth to open the file

Opening a video directly in vdub , then using vdub to cut in direct stream copy mode won't give you uncompressed video, you're left with original compression (eg. DV in his case)

[poisondeathray]

Nero-sample:

[Attachment 16994 - Click to enlarge]

TMPEGEnc-sample:

[Attachment 16995 - Click to enlarge]

I haven't looked at these closely yet, but the problem with comparing these is they use a different bitrate (~ 8Mbps vs. ~6.3 Mbps)

A more valid comparison would use similar bitrates

[Cornucopia]

There's a lot of other differences: BFF vs. TFF, VBR w/ Diff max BRs, M3N13 vs. M3N15 GopType,..

I'm surprised these both look as similar (at a glance) as they do.

Scott

[brusno]

You need to do a little more reading. VOB is MPEG reorganized and renamed for DVD disc. There is no re-encode from MPEG to VOB.

I'd stay away from Nero's encoder.

Yes I know that MPEG2 is DVD-format.
Re-encoding can occur even if it is from one format to the same, MPEG2 - MPEG2 or MPEG2 - VOB (I believe). It depends on the program I guess. But it may be that it isn't hurting the quality that much. It may also be, that the program isn't re-encoding even if it is telling me so. That it is only converting it from MPEG2 to VOB without quality loss. Hard to know, how the programs are working under the hood.
But if you have an avi file, and a mpeg2 file (let us say that they come from the same source, and that the avi file is the source), and are loading these two files into a program, i.e. Nero. And then converting (encoding(?)) these two into the same DVD. Would you feel confident that the two videos have the same quality?

Can you tell what it is that makes Nero's MPEG2 encoder worse than others? Is it creating more noise, blocks, etc.? Or isn't the video running as smoothly?
The reason I ask is that when comparing a Nero encoding with a TMPEGEnc encoding, I cannot see a big difference.

It is one issue with Nero that I know of. And that is the SmartEncoder. If this isn't deactivated, then the encoding will be crap. And unfortunately the SmartEncoder is ON by default. But if I deactivate this, then the encoding is good enough I think.

[jagabo]

Number 2 is the original avi opened in VirtualDubMod via Avisynth script (using AviSource()), and then saved again with Direct Stream Copy.

The output of AviSynth is always uncompressed video. In this case VirtualDub worked as expected -- AviSynth gave it YV12 frames, it saved YV12 frames.

[Cornucopia]

You need to do a little more reading. VOB is MPEG reorganized and renamed for DVD disc. There is no re-encode from MPEG to VOB.

I'd stay away from Nero's encoder.

Yes I know that MPEG2 is DVD-format.
Re-encoding can occur even if it is from one format to the same, MPEG2 - MPEG2 or MPEG2 - VOB (I believe). It depends on the program I guess. But it may be that it isn't hurting the quality that much. It may also be, that the program isn't re-encoding even if it is telling me so. That it is only converting it from MPEG2 to VOB without quality loss. Hard to know, how the programs are working under the hood.
But if you have an avi file, and a mpeg2 file (let us say that they come from the same source, and that the avi file is the source), and are loading these two files into a program, i.e. Nero. And then converting (encoding(?)) these two into the same DVD. Would you feel confident that the two videos have the same quality?

Can you tell what it is that makes Nero's MPEG2 encoder worse than others? Is it creating more noise, blocks, etc.? Or isn't the video running as smoothly?
The reason I ask is that when comparing a Nero encoding with a TMPEGEnc encoding, I cannot see a big difference.

It is one issue with Nero that I know of. And that is the SmartEncoder. If this isn't deactivated, then the encoding will be crap. And unfortunately the SmartEncoder is ON by default. But if I deactivate this, then the encoding is good enough I think.

To clarify: DVD is MPEG2 format (not necessarily vice-versa). There are plenty of MPEG2 encodes that ARE NOT DVD-Video compliant.
And for the umpteenth time: VOB is not the same as MPEG2-PS (or MPEG2-TS for that matter) and should not be treated as such. There are differences in the container, you can't just "rename" the extension. Luckily, assuming your footage is DVD-Video compliant, authoring SHOULD just remux from MPEG2-PS (or raw/ES video streams) to VOB rather than re-encode.

Yes, it does depend on the program, whether it re-encodes or not.

It IS possible to tell (by how long it takes to convert, and by doing an AVISynth Subtract() compare after the fact).

If the app is TELLING you it is re-encoding, it is.

Not that hard to know how they work, they kind of fall into particular categories: 1-Consumer apps that re-encode everything, 2-Prosumer apps that re-encode only non-compliant material (or an option to bypass), 3-Pro apps that don't re-encode and just baulk at non-compliant material (leaving you to do the re-encoding), 4-GUI frontends to FOSS CLI authoring & encoding apps that tell you if and when they encode (usually with the option to bypass).

re: Nero A/B compare - NO, they will NOT be the same. They might be CLOSE, but the source will ALWAYS be better quality.

Nero is not a CRAPPY encoder, but not that great in the big scheme of things. Nero's algorithms just aren't as smart, so yes, there is more noise/blocks, etc.
If your side by side with TMPGEnc is very similar, it could be you didn't optimize the TMPGEnc encoding very much. For MPG2, TMPGEnc isn't the BEST, but it's up there above Nero certainly.

Scott