smartLabs tsMuxeR: Transport Stream muxer

[BZeeme]

Roman (or other)

Has development of tsMuxeR stopped? :( No current info at smlabs.net. I seldom use it as stand-alone but other programs I use frequently do use it. I'm considering a donation but don't want to put money on it if it is a dead project.

[73ChargerFan]

It's not dead, it is freeware developed by a commercial company. About once a year, they revisit it and update or add features as required. It is very useful & stable as is.

[jj666]

Yes, I'm sure only about 3 people donated also, which probably doesn't help development.

-jj-

[BZeeme]

That's a shame as tsMuxeR is a critical component of many front-end tools doing BD manipulation.

[G_M_C]

They could also choose to open up the source, to an more Open Source like license format. That's not unusual, even MS does it now and again.

It saves them the trouble of developing it further, and gives them a better developed product in return. They can then use that better tsMuxeR with their own producs, like they do now (it's a part of a suite as i understand).

They might save spending bandwidth, as we can expect the Open Source version to "pop up" in sorceforge or such ;)

So it seems you can make this change benificial for all parties, making that there are several arguments in favour of such a change :)

[73ChargerFan]

G_M_C

Except there is no other tool as functional as this one. IMO, it has a high value to them as a closed source product, that can be sold to a larger company later if they have a cash crisis. But, never say never.

Could you convince madshi about the benefits of open source? :D

[G_M_C]

G_M_C

Except there is no other tool as functional as this one. IMO, it has a high value to them as a closed source product, that can be sold to a larger company later if they have a cash crisis. But, never say never.



You said the crucial thing yourself; "Except there is no other tool as functional as this one".

If that is the case, why havent they capitalized on it already ?

[Pl4yit]

Just wanted to mention a problem that I have come across when feeding tsMuxeR with a PAL mpeg2 file and a SRT subtitle file.

Although the command line parameters indicate to render the subtitles for a video size of 720x576 (PAL), the rendered subtitle stream is in 720x480 (see tsMuxeR log here below copied from the GUI). The result is that the subtitles are not visible during playback.

Command Line:
MUXOPT --no-pcr-on-video-pid --new-audio-pes --avchd --vbr --vbv-len=500
V_MPEG-2, "D:\video.mpeg2", fps=25
A_AC3, "D:\audio.ac3", lang=eng
S_TEXT/UTF8, "D:\English.srt",font-name="Arial",font-size=28,font-color=0x00ffffff,
bottom-offset=24,font-border=2,text-align=center,video-width=720,video-height=576,fps=25, lang=eng


Log copied from GUI:
SmartLabs tsMuxeR. Version 1.10.6 http://www.smlabs.net
MPEG-2 stream does not contain fps field. Muxing fps=25
Decoding MPEG2 stream (track 1): Profile: Main@8. Resolution: 720:576p. Frame rate: 25
Decoding AC3 stream (track 2): Bitrate: 448Kbps Sample Rate: 48KHz Channels: 6
Decoding PGS stream (track 3): Resolution: 720:480 Frame rate: 25
Processed 62820 video frames
Creation of AVCHD playlist
Creation of AVCHD stream info and seek index
Mux successful complete.
Muxing time: 5 sec

[deank]

This bug was reported few months ago. That was one of the reasons to add advanced processing of subtitles in multiAVCHD to overcome certain bugs in tsMuxeR.

[Pl4yit]

This bug was reported few months ago. That was one of the reasons to add advanced processing of subtitles in multiAVCHD to overcome certain bugs in tsMuxeR.

OK.
Thanks, Dean.

[deank]

Another option is to prepare your subtitles first and then use tsMuxeR. You can convert your SRT to SUP/PGS with easySUP (http://forum.doom9.org/showthread.php?t=149160) and then tsMuxeR will mux them just fine and they will be visible.

[Pl4yit]

Another option is to prepare your subtitles first and then use tsMuxeR. You can convert your SRT to SUP/PGS with easySUP (http://forum.doom9.org/showthread.php?t=149160) and then tsMuxeR will mux them just fine and they will be visible.

Great. I will have a look at EasySUP. Thanks for the info.

[Ryu77]

It's not dead, it is freeware developed by a commercial company. About once a year, they revisit it and update or add features as required. It is very useful & stable as is.

Once a year? The first release was in Jan 2008 and there have been close to 20 revisions since then (I could be wrong here as that was just a guess from the top of my head).

I think the reason why there was a delay of a few months for the most recent release is that tsMuxeR (as you said) has got to quite a stable point now. There might be a few little bugs still, but I haven't come across any.

[73ChargerFan]

Once a year? The first release was in Jan 2008 and there have been close to 20 revisions since then
Yeah, 8 at the beginning, then none for almost a year, then 12 more. I'm happy though.

[Boulder]

I'm trying to mux a 5.1-channel AAC track with a video file, but tsMuxerR GUI doesn't recognize the file :( The audio file is created out of a regular dts file using eac3to. Is there anything to check before I start looking for a different transport stream muxer?

[deank]

No, just convert your DTS to AC3 5.1 (not to AAC). AAC is not supported in Blu-ray/AVCHD anyway.

[Boulder]

I'd like to use the transport stream in my Popcorn Hour A-110, which decodes TS in hardware. Thus no need for Blu-ray compatibility :)

[deank]

Okay, but still you decided to reencode audio from DTS to AAC. If you want to use tsMuxeR, reencode it to AC3 instead and it will be okay.

[Beastie Boy]

I'm trying to mux a 5.1-channel AAC track with a video file, but tsMuxerR GUI doesn't recognize the file :( The audio file is created out of a regular dts file using eac3to. Is there anything to check before I start looking for a different transport stream muxer?

tsMuxeR will accept AAC audio, but it must be a raw file, not in a container. MP4 or m4a files won't work. I tried this some time ago and used YAMB with mp4Box to demux the .aac stream from the container.
I did find that I had errors if the bitrate was too high, ie quailty greater than .90 or something with the Nero encoder.

Hope this helps. Cheers, Beastie.

[Boulder]

Thanks, that info will help me a lot :)

[M-Blaster]

The main movie of some new released blu-ray consists of several small and big .m2ts files.

1. Is there a comfortable function to identify the .m2ts files that are containing the main movie?
2. Is there an easy way to merge them ( f. g. 00001.m2ts till 00023.m2ts) into one big .m2ts file?

[crl2007]

You identify them with BDInfo. And you merge them with eac3to. Or with tsMuxer.

[asc28]

Can't seem to mux with negative timeshifts for an ac3 track. Tried tsMuxer and GUI 1.10.6 on multiple ac3's. Always the same result: mux is fine, but delay is wrong in the output file
tsmuxer 01.meta 01.tsMUXOPT --no-pcr-on-video-pid --new-audio-pes --vbr --vbv-len=500
V_MPEG-2, "D:\01.m2v", fps=29.970
A_AC3, "D:\01.ac3", timeshift=-100mstsmuxer 01.tsSmartLabs tsMuxeR. Version 1.10.6 http://www.smlabs.net
Track ID: 4113
Stream type: MPEG-2
Stream ID: V_MPEG-2
Stream info: Profile: Main@8. Resolution: 720:480i. Frame rate: 29.97
Stream lang:
Track ID: 4352
Stream type: AC3
Stream ID: A_AC3
Stream info: Bitrate: 224Kbps Sample Rate: 48KHz Channels: 2
Stream lang: eng
Stream delay: 28

[tebasuna51]

Can't seem to mux with negative timeshifts for an ac3 track. Tried tsMuxer and GUI 1.10.6 on multiple ac3's. Always the same result: mux is fine, but delay is wrong in the output file

Bad idea use audio timeshifts/delay of containers.
Always is best use DelayCut to cut or delay an audio to sync with video before mux.

[asc28]

Bad idea use audio timeshifts/delay of containers.
Always is best use DelayCut to cut or delay an audio to sync with video before mux.Thought non-destructive timeshifts was the whole point of having this feature in a container. It's also advertised in the example shown in the readme. Can anyone confirm that negative timeshifts work?

[Inspector.Gadget]

Negative delays work here using TSMuxer / TSMuxerGUI 1.10.6. Demuxed DTS is bit-identical to original imported from MKV container.

[jdobbs]

Thought non-destructive timeshifts was the whole point of having this feature in a container. It's also advertised in the example shown in the readme. Can anyone confirm that negative timeshifts work?They work for me (v1.10.6).

[asc28]

Thanks. I wonder what's wrong for me, anyone noticed any mistakes in the meta or command-line?

[laserfan]

Thanks. I wonder what's wrong for me, anyone noticed any mistakes in the meta or command-line?The meta looks fine to me, but what do you mean when you say "delay is wrong in the output file"? Audio out-of-sync? Have you tried other values?

The (very) few times I have experienced out-of-sync audio it's taken me quite a few tries (at re-muxing with different timeshift values) to get it perfect. And I find I'm always guessing wrong at plus or minus timeshift values. :o

[Capsbackup]

I too have found that negative time shifts do not seem to work with tsMuxeR.
At least, if you set the delay to a negative number, remux that file, and then check the results with tsMuxeR GUI, the .m2ts will not have the negative delay you set before the mux when you highlight the audio track.
I have reported this before, but this is the first I have read that someone else has experienced the same issue.

EDIT: previous report:
http://85.230.118.136/showthread.php?p=1325855#post1325855

[asc28]

The meta looks fine to me, but what do you mean when you say "delay is wrong in the output file"? Audio out-of-sync? Have you tried other values?I mean that when you analyze the ts with tsmuxer or eac3to, the delay isn't what you set it to.

[Capsbackup]

I mean that when you analyze the ts with tsmuxer or eac3to, the delay isn't what you set it to.

I agree, this is what I experienced too!
I have not found any other original BluRays that have a negative audio delay, luckily. Pirates of the Caribbean 3 At Worlds End is the only one. Since a -12ms is probably too small to make a noticeable async, I have not followed up with this concern.

[laserfan]

I mean that when you analyze the ts with tsmuxer or eac3to, the delay isn't what you set it to.Ok, but has the setting been incorporated into the mux? I have only had sync errors in a few instances, but I've never *not been able* to use timeshift to fix them (though it has involved for me many attempts trial-and-error to get sync perfect).

But no, I've also never used tsmuxer to try to read the file again and see the changes I made.

[Capsbackup]

Ok, but has the setting been incorporated into the mux? I have only had sync errors in a few instances, but I've never *not been able* to use timeshift to fix them (though it has involved for me many attempts trial-and-error to get sync perfect).

But no, I've also never used tsmuxer to try to read the file again and see the changes I made.

For me, the timeshift value inserted by either the command line thru BD-RB or tsMuxeR GUI is not the same value that the remuxed file has when read again thru tsMuxeR. My example had such a minor delay, (-12ms) that it was not going to be noticeable.
My only point is that tsMuxer does not seem to honor the negative delay amount inserted, so if a file had a delay that would be apparent, the timeshift option in tsMuxeR would not correct this.
This, to me, is more of a problem when using BD-RB in full movie backup mode, especially with PiP, since manually correcting such a delay is not an option outside of BD-RB and then being able to substitute back into BD-RB without breaking the full movie playback.
For a movie only backup, there are other options to correct the delay, at least for AC3 audio. I'm not sure if there is an option for a DTS HD Master or Dolby True HD audio file though.

[deank]

Are you sure that the negative delay is not properly applied?

Do not get confused by the value you set and the one you see after a 're-check'.

I think m2ts container is not as flexible as mkv and achieving a larger negative delay may sometimes require cutting the audio stream. Me - I never had any troubles with -500ms or -600ms delay. I played a lot and +/- was always properly processed (at least with h.264/avc video).

Dean

[Capsbackup]

@deank;
I have only had one movie that had a negative audio delay. I did try this one several times though, with the same results.
Perhaps the value( -12ms) is too small, but I doubt it.
I have tested other positive audio delays, and after a recheck, the .m2ts file does have the same positive delay as the original.
This was why I believed there was an issue with negative delays, and responded to asc28.
However, it is encouraging to see your results differ and you have had many more attempts to try than me! :)
Yours is surely a better, more favorable result than ours, especially with the lack of updates and responses from the tsMuxeR developers.

[laserfan]

It always works for me too. My assumption would be that your -12ms change "took" (worked) but it's VERY hard (impossible?) to judge "by eye". I usually trial/error with 20 or 50ms and before looking again at the result.

[Capsbackup]

Like I said, it was hard to tell if the audio was async with only -12ms to begin with. My only point is that if the original displays a negative audio delay when viewing with tsMuxeR GUI, the resulting remuxed .m2ts was not displaying that same delay after setting the negative delay within tsMuxeR and viewing the remuxed file again, with tsMuxeR GUI.
The original and backup both appear to have audio sync.
If the software displays an audio delay with the original, be it negative or positive, then should the remuxed file also display that same audio delay when viewed after remux with tsMuxeR GUI?
If the software corrects the audio delay, as programs like delaycut 1.3.0.0 does, then I would not expect to see an audio delay reported on the remuxed file when viewed with tsMuxeR GUI.
EDIT:
I don't believe this is the way tsMuxeR works though. I experienced an 83ms audio delay on a movie that was noticeable, and after applying the 83ms delay, the remuxed file, when viewed thru tsMuxeR GUI, displayed this same audio delay as the original.

[laserfan]

If the software displays an audio delay with the original, be it negative or positive, then should the remuxed file also display that same audio delay when viewed after remux with tsMuxeR GUI?This is of course very different from the question raised "does negative timeshift work?" cuz it clearly does work.

[asc28]

It works for practical purposes, but it doesn't do it by writing a negative delay metadata. Instead, it cuts the audio destructively (by more than the delay you wanted if it's between a specific AC3 packet frame size), then it applies positive delay to compensate for any overcutting.

[Capsbackup]

It works for practical purposes, but it doesn't do it by writing a negative delay metadata. Instead, it cuts the audio destructively (by more than the delay you wanted if it's between a specific AC3 packet frame size), then it applies positive delay to compensate for any overcutting.

If this is true, it might explain why my -12ms delay, after running it thru tsMuxeR, produced a 20ms(positive) delay on recheck with tsMuxeR. :confused:
Of course, I did not ever experience a noticeable delay regardless, but understanding the way the software treats the delay is good to know.

[jdobbs]

It works for practical purposes, but it doesn't do it by writing a negative delay metadata. Instead, it cuts the audio destructively (by more than the delay you wanted if it's between a specific AC3 packet frame size), then it applies positive delay to compensate for any overcutting. That's the logical way to do it. You have to start on a packet boundary, and you can't set a negative PTS. The alternative would be to delay the video by an equivalent amount -- but then you are left having to deal with every other stream's time (like PGSs, other audio, etc).

[SUPERBIF]

I'm new to this, but I'm trying to convert my HD DVD to Blu-Rays.

There's some strange things I found out.
I use evodemux to strip and only have 1 video and 1 audio track.

When I have stripped and take it into tsmuxer. I can see it says 1080i and 29.97fps.

I then press remove pulldown so it says 24000/1001.


My questions are:

1. Why does it not show 1080p and 23.976fps in the first place?
2. Is it okay to remove pulldown to force it to 23.976?
3. When I use mediainfo on the original evo file it shows 24.8 Mbit bitrate, but when I take the converted Blu-Ray it shows 18 Mbit. Shouldn't the bitrate be the same when it's not compressed?
4. Also I can see mediainfo shows audiodelay sometimes were tsMuxer doesn't show any. What should I count on?
5. Does it matter that the original in mediainfo shows (scan order 3:2 pulldown) and in the convertet shows (scan order button field first)?

[whitemandingo]

It looks like I found the right thread. I too have been struggling with audio sync on a couple of titles. I've remuxed many times with different delay values, but can never get the sync just right. I know there are some smart folks on this thread so maybe someone here can help.

From a practical standpoint, when playing the BDMV folder with VLC I can use the track synchronization option in VLC to get the audio/video sync perfect. The setting is: advance of audio over video= -3.000s Assuming this is a valid point of reference, could a formula be applied to determine the correct tsumuxer delay value so the output streams are in sync? Or, does ac3 packet boundary compensation make this unfeasible?

[dvdboy]

I don't know if I'm doing something stupid, but when I try to playback the output I create below in VLC, I get three video tracks, no audio tracks:

MUXOPT --no-pcr-on-video-pid --new-audio-pes --vbr --vbv-len=500
V_MPEG-2, "\\192.168.3.6\TEST\{stitched sources}_MPEG.m2v", fps=25
A_LPCM, "\\192.168.3.6\TEST\test 28 01_02 25fps.wav", lang=jpn
A_LPCM, "\\192.168.3.6\TEST\test 28 03_04 25fps.wav", lang=eng

Is this a VLC bug, or am I doing something stupid.

[laserfan]

I too have been struggling with audio sync... From a practical standpoint, when playing the BDMV folder with VLC I can use the track synchronization option in VLC to get the audio/video sync perfect. The setting is: advance of audio over video= -3.000s Assuming this is a valid point of reference, could a formula be applied to determine the correct tsumuxer delay value so the output streams are in sync?Wow, 3 seconds is a lot; I've never had an effort anywhere close to 3 seconds out.

Sorry, I don't know any "magic formula" for getting this right myself, though I've never used VLC for this (will try it next time). I have only done trial & error with tsMuxeR, which fortunately is easy to do and only takes a few minutes to re-mux for every trial.

Of course, the other wild card is the PC playback mechanism i.e. once you've re-muxed the thing, playing it back and observing the sync. I've used PowerDVD for this though it's not perfect itself.

[whitemandingo]

It's not actually 3 seconds so it may not be a valid point of reference for tsmuxer, but that's the correct VLC setting to make it sync perfectly in VLC and it's off the same amount no matter which computer it's played on so system specs are not a variable. I'm not sure what the actual delay is, but it's enough to be annoying if not corrected and too much to commit it to DVDR+DL for playback on set-top player or PS3. I was hoping the VLC reference might be enough to translate into a correct tsmuxer setting if someone else has experienced this.

[asc28]

and you can't set a negative PTS.Why then do we see negative delays in commercial transport streams (i.e. m2ts on Blu-rays)?

[rapscallion]

It's not actually 3 seconds so it may not be a valid point of reference for tsmuxer, but that's the correct VLC setting to make it sync perfectly in VLC and it's off the same amount no matter which computer it's played on so system specs are not a variable. I'm not sure what the actual delay is, but it's enough to be annoying if not corrected and too much to commit it to DVDR+DL for playback on set-top player or PS3. I was hoping the VLC reference might be enough to translate into a correct tsmuxer setting if someone else has experienced this.
I have a movie that is over 4 sec out of sync in VLC.
When I enter that #, a neg btw (-4100ms), it syncs perfectly when I create an MKV or run through tsmuxer.(mediainfo shows a positve 28ms though) Yet, if I enter 28ms and Mediaifo shows 28ms, it is completely out of sync. Go figure.

So, to answer your question, the VLC sync # is valid for other apps.

[jdobbs]

I have a movie that is over 4 sec out of sync in VLC.
When I enter that #, a neg btw (-4100ms), it syncs perfectly when I create an MKV or run through tsmuxer.(mediainfo shows a positve 28ms though) Yet, if I enter 28ms and Mediaifo shows 28ms, it is completely out of sync. Go figure.

So, to answer your question, the VLC sync # is valid for other apps. That's probably because after removing the -4100ms the 28ms point was the nearest boundary. So the +28ms correction was made to put it in sync. Just a quess.