HD PVR 2 - AAC to AC3 Need Help

[DarrellS]

I am also in the VideoReDo TV Suite crowd. It is the only program that I tried that cut my files correctly.

As for converting 6 channel aac to 6 channel ac3, I found this program on doom9 that did the job for me where dBpoweramp and foobar2000 both failed.

http://forum.doom9.org/showthread.php?t=167263

AACtoAC3...

http://www.sendspace.com/file/tungt2

[VideoFanatic]

@ videofanatic, how long are your joins taking you ?

my path goes like this, using two new usb3 hdd's: [diskI] 1tb hdd usb2 -> [diskJ] 1tb hdd usb2 -> video.ts

I echo what Jagabo said - I have an 8 core 4.2 Ghz CPU so it's fairly quick to mux things - you can use the script . Also I never use USB hard drives for anything. Windows 8 for example has buggy usb hard drive support. I find them slow and unreliable compared to internal Sata III hard drives. When I want to store files outside my computer for archival, I just copy stuff from my internal drives onto eSata drives which function as if they were internal hard drives.

Easy way to mux the split files into 1 file would be by using this script that Jagabo gave me:

echo MUXOPT --no-pcr-on-video-pid --new-audio-pes --vbr --vbv-len=500 >built.meta
echo V_MPEG4/ISO/AVC, %1+%2+%3+%4+%5+%6+%7+%8+%9, insertSEI, contSPS, track=100 >>built.meta
echo A_AAC, %1+%2+n3+%4+%5+%6+%7+%8+%9, timeshift=-466ms, track=101, lang=USA >>built.meta
"X:\Portable Installations\tsMuxeR_2.3.2(b)\tsMuxeR.exe" built.meta "%~d1%~p1appended.ts"
pause

On another subject I tried the Elgato on my brand new Quad Core 2.14 Ghz laptop with 8GB ram & Windows 8.1 (min requirements were Dual Core 2Ghz). There was nothing running on it except the Elgato software. The recordings looked OK at first but on closer inspection on some portions of the video there were glitches in the recording that weren't in the source such as where the picture would jump up every several seconds or so for a split second. Example: https://www.youtube.com/watch?v=sa7_cX7ST9Y

Today I've bought the HD PVR 2 Gaming Edition Plus as that seems to be a full hardware encoder unlike the Elgato which uses your CPU to encode the video. I'm hoping I'll have more luck with that. While I'm waiting for it to arrive I'm back to using the Elgato on my desktop.

Was wondering, can the HD PVR 2 Gaming Edition Plus support 5.1 surround via HDMI input?

[vhelp]

that script is getting me nowhere, its not working. plus, i'm not building or doing anything with meta files. i am just trying to test the speed of JOINing two (or more files) based on jagabo's comment earlier, using the txMuxR app in command line mode. my version is:

Network Optix tsMuxeR. Version 2.5.5(b). www.networkoptix.com

tsmuxer.exe "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0001.ts" + "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0002.ts" i:\test2.ts

tsmuxer.exe MUXOPT "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0001.ts" + "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0002.ts" i:\test2.ts

tsmuxer.exe "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0001.ts" "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0002.ts" i:\test2.ts

all these tests scripts did not work for me.

[vhelp]

so i've been googling around for parameter examples to no avail. no examples. i've tried adding searching terms like: join, append, add, new, concat, etc, but don't see any examples, yet. this is frustrating. just want to test -i/-add/-new/-cat/-concat/-join/-app/-append/ functions but they don't exist as parameter terms in tsmuxer.

edit -- later on..

after more searching, i did not find any example for command line usage for JOINing. so i turned to the gui and found that it seems to include the parameters used for the join operation. however, it did not work as planned:

Network Optix tsMuxeR.  Version 2.5.5(b). www.networkoptix.com
Can't detect stream type

is the result. so i decided to let it do it through the gui, because i just want to test the join speed. anyway, its been joining two 4096g files, and it just completed in 12min, 38sec. so command line or not, it takes the same amount of time, and although i can't get the command line method to work, my custom filejoiner app worked better: filejoiner vid1 vid2 finalvid.ts i think i will just stick with that for my files.

[jagabo]

my custom filejoiner app worked better: filejoiner vid1 vid2 finalvid.ts i think i will just stick with that for my files.

Is that the same as

copy /b vid1.ts+vid2.ts finalvid.ts

using the command line copy command? Or do you actually parse and patch the TS stream?

[VideoFanatic]

The script I gave you works for me with tsMuxeR_2.3.2(b), just change the tsmuxer directory to where ever you have TS Muxer located:

echo MUXOPT --no-pcr-on-video-pid --new-audio-pes --vbr --vbv-len=500 >built.meta
echo V_MPEG4/ISO/AVC, %1+%2+%3+%4+%5+%6+%7+%8+%9, insertSEI, contSPS, track=100 >>built.meta
echo A_AAC, %1+%2+n3+%4+%5+%6+%7+%8+%9, timeshift=-466ms, track=101, lang=USA >>built.meta
"X:\Portable Installations\tsMuxeR_2.3.2(b)\tsMuxeR.exe" built.meta "%~d1%~p1appended.ts"
pause

I don't use later TSMuxer versions as they have bugs. The script doesn't do anything with meta files, it merely combines several files into 1. First of all go to: C:\Users\Dave\AppData\Roaming\Microsoft\Windows\Se ndTo

Open Notepad. Save As and choose save as type "All Files" and save As "Ts Muxer.bat". Move the bat file into the SendTo directory as mentioned above. Paste the script into the bat file.

To run the batch file, select all the files you want to join then right-click on the 1st one and choose "Send To" and select the TS Muxer.bat file. The files will then be joined in alphabetical order.

[jagabo]

The script doesn't do anything with meta files, it merely combines several files into 1.

Actually, it first builds a meta file ("built.meta") then passes that file to tsmuxer -- the same thing that tsmuxerGUI does (though tsmuxerGUI uses tsmuxer to get specific information from the source files to customize the meta file). If you want you can delete the meta file after calling tsmuxer, by adding "del built.meta" to the batch file. I didn't delete the meta file so I could examine it to verify I built it correctly.

[vhelp]

Is that the same as

copy /b vid1.ts+vid2.ts finalvid.ts

using the command line copy command? Or do you actually parse and patch the TS stream?

yes, it is a binary copy. i made it to copy the partial "test" hevc encodes i would make, instead of re-encoding the same sections again, i would continue from the last frame onward, another n-number of frames. then i would just join them again via the filejoiner tool. part of the reason for creating the tool was because in my xp system, i seem to have to always include a cmd /c and that is just one more addition i didn't want to do. all those joins with the tool worked succesuflly. been using since, when i do those type of projects. however, with this latest set of projects with joining elgato encoded .ts files, it seems the tool might be better for me, less hassling about what parameters to use and get working inother front end tools. however, i have not thoroughly tested it on .ts files. i will have to run some short, stop and go captures, and see how the filejoiner will do in those tests. and, maybe another test for short encoded avc clips from x264 and see how that goes. anyway. sorry for the long explanation.

[jagabo]

in my xp system, i seem to have to always include a cmd /c and that is just one more addition i didn't want to do.

That's a bizarre statement. What's so hard about adding CMD/C to a batch file?

Copy/B should work with the sequential files from the ElGato (it worked with the sequential test files I have). There won't be a break in time codes from segment to segment. But with random ts files most player will only play the first segment.

The bigger problem is determining the number of files on the command line and building the proper command line. I suppose you could just have multiple batch files, copy2.bat, copy3.bat, copy4.bat and just select the one you want.

[vhelp]

@ VideoFanatic, i hope the hdpvr2 does better for you, quality'wise. i don't like how the pvr handles dark and black scenes. lots of chisel chod. but the elgato handles those pretty well, using the same bitrate. unfortunately, w/ the hdpvr, the CBR mode is broken. i always record in 13MBps bitrate. but it does not keep that bitrate setting continuously. it still encodes between 6-11 MBps, and usually at lowest birtate. that should explain why the dark scenes look awful, and pretty much everything else. but the only negative issue i have at this time w/ the elgato is the component connection is not recording my 480i properly. it is squeeshing 720 inside 640x480 dimensions and adding 80 pixels to fill the remaining 720 dimension when i go from hdpvr -> component_out -> elgato component. it doesn't seem to use a lot of resources as i thought it would, only +/- 21% about average on my dell i3 laptop w/ win 7 prem. maybe that is considered high for most people.

[jagabo]

elgato... doesn't seem to use a lot of resources as i thought it would, only +/- 21% about average on my dell i3 laptop w/ win 7 prem. maybe that is considered high for most people.

The Hauppauge HD PVR 2 consumes less than 1 percent when not displaying the video. So it can easily be used to record the output of the computer it's running on. Not so with the ElGato -- it will slow down whatever else your doing and drop frames.

[vhelp]

Copy/B should work with the sequential files from the ElGato (it worked with the sequential test files I have). There won't be a break in time codes from segment to segment. But with random ts files most player will only play the first segment.

i thought it didn't work for some people.

[vhelp]

The Hauppauge HD PVR 2 consumes less than 1 percent when not displaying the video. So it can easily be used to record the output of the computer it's running on. Not so with the ElGato -- it will slow down whatever else your doing and drop frames.

i have not seen this with my recording projects. but then again, i never tested this while running other tasks on the laptop.

[vhelp]

i thought i would report that i managed to get tsmuxer working in command line.

1. i copy all files and settings (get from tsmuxerGUI's meta file section) and pasted into a file. i simply used files.txt
2. in dos console window i entered: tsmuxer files.txt newfile.ts
3. then, wait a few minutes or more, depending on how many files, (in my case is a long while)
4. finished

files.txt

MUXOPT --no-pcr-on-video-pid --new-audio-pes --vbr  --vbv-len=500
V_MPEG4/ISO/AVC, "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0001.ts", insertSEI, contSPS, track=100
A_AAC, "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0001.ts", timeshift=-284ms, track=101, lang=eng
V_MPEG4/ISO/AVC, "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0002.ts", insertSEI, contSPS, track=100
A_AAC, "J:\EGC_Library\- - - - 2014-02-15 08-00-33\Recording_2014-02-15_20-00-34_0002.ts", timeshift=-284ms, track=101, lang=USA

[jagabo]

That's basically what my batch file did -- it built the meta file and spawned tsmuxer for you. Of course, you may have to modify the batch file to suit your ts files' properties -- audio delay, etc.. The tsmuxer instructions say the meta file must have the extension .meta. But I guess your .txt file works.

[VideoFanatic]

Jagabo I was given this script to amplify the audio to 0dB. I uses FFMPEG and AAC Gain. I was told if the audio is already 0dB then it will have no effect otherwise it will increase the volume to 0dB without clipping.

ffmpeg -i $args[0] -c copy -vn gain.m4a
aacgain -r -k -m 10 gain.m4a
ffmpeg -i $args[0] -i gain.m4a -map 0:v -map 1 -c:v copy -c:a ac3 -ac 2 `
  -ab 128k outfile.ts

However I'm not sure how to get it working correctly. Basically I have a h264 .ts file with AAC audio that I want to mux the video back to .ts but re-encode the audio to AC3. You gave me the below script so that when I Right-click on a file and select "Send To" the script, a file called "original-file-name.AC3.ts" is created in the same directory as the file I right-clicked on. Basically I want to use the below script but also do the above amplification on the AC3 file created. Do you know what script I should use to do that please?

"X:\Portable Installations\ffmpeg-2014 May 14-git-72dcd48-win64-static\bin\ffmpeg.exe" -i %1 -acodec copy "%~d1%~p1%~n1.aac"
"X:\Portable Installations\ffmpeg-2014 May 14-git-72dcd48-win64-static\bin\ffmpeg.exe" -i %1 -vcodec copy -acodec ac3 -ac 2 -ab 128k "%~d1%~p1%~n1.AC3.ts"

The top line demuxes the AAC into an aac file.
The bottom line converts the AAC into an AC3 file and muxes with the original TS file to create a file called "original file name.AC3.ts"

[jagabo]

I think you want something like this:

"G:\Program files\ffmpeg\bin\ffmpeg" -i %1 -c copy -absf aac_adtstoasc -vn gain.m4a
"G:\Program files\aacgain\aacgain" -r -k -m 10 gain.m4a
"G:\Program files\ffmpeg\bin\ffmpeg" -i %1 -i gain.m4a -map 0:v -map 1 -c:v copy -c:a ac3 -ab 2 -ab 128k "%~d1%~p1%~n1.AC3.ts"
"G:\Program files\ffmpeg\bin\ffmpeg" -i gain.m4a -acodec copy "%~d1%~p1%~n1.aac"
del gain.m4a

That will give you a normalized AAC file with the base name of the source, and a TS file with normalized AC3 audio with the base name of the source. You many need other switches depending on your sources.

But... Commercial releases usually have AC3 audio encoded at about -10 dB. So you're videos will have abnormally loud audio. Players that compensate for that by applying a 10 dB gain will distort when playing your videos. And 128 kbps AC3 is pretty crappy.

[VideoFanatic]

Thanks. That works perfectly. When dumixing the original AAC into a separate AAC file, is that being re-encoded due to AAC Gain being used to increase the volume?

I'm only using that script for stuff recorded off TV so I won't get that problem you mentioned will I?

Just wondering about what you said about commercial releases. -10 dB is louder that 0dB isn't it? Why would they release stuff like that? I thought 0dB is the loudest you should go because any louder and you get clipping?

I was told on the FFMpeg forum that the script is supposed to raise or lower the volume to 0dB. So was just wondering why after using the script that a video player would add a 10 dB gain and why would that distort the audio if it was originally -10 dB anyway?

Any idea why AAC Gain has such a low rating on VideoHelp.com? It works perfectly in that script you gave me.

[jagabo]

Minus 10 dB is quieter, not louder.

[VideoFanatic]

So are you saying I should encode the audio at -10dB then to prevent any problems?

If so, what part of the script should I change?

Do you know if demuxing and amplifying the AAC in the script is re-encoding the audio due to the amplification or does AAC Gain to lossless normalization?

[jagabo]

I don't know anything about AAC Gain. I simply used your command line.

[VideoFanatic]

OK I understand but could you clarify something for me please (You don't need to know anything about AAC Gain to help with this). Rather than encoding to 0dB, if I encode to -10dB will that prevent the problems you mentioned as follows?:

But... Commercial releases usually have AC3 audio encoded at about -10 dB. So you're videos will have abnormally loud audio. Players that compensate for that by applying a 10 dB gain will distort when playing your videos. And 128 kbps AC3 is pretty crappy.

[jagabo]

OK I understand but could you clarify something for me please (You don't need to know anything about AAC Gain to help with this). Rather than encoding to 0dB, if I encode to -10dB will that prevent the problems you mentioned as follows?:

But... Commercial releases usually have AC3 audio encoded at about -10 dB. So you're videos will have abnormally loud audio. Players that compensate for that by applying a 10 dB gain will distort when playing your videos. And 128 kbps AC3 is pretty crappy.

Yes.

[VideoFanatic]

Commercial releases usually have AC3 audio encoded at about -10 dB. So you're videos will have abnormally loud audio. Players that compensate for that by applying a 10 dB gain will distort when playing your videos. And 128 kbps AC3 is pretty crappy.

Isn't there a program that can tell me what the actual dB is of an audio or video file? Because it's confusing. For example I used my script and changed the value in this line from 10 to 0. Because 10 was too loud I think based on what Jagabo said.

X:\Portable Installations\AAC Gain 1.9\aacgain" -r -k -m 10 gain.m4a

I then opened the audio file in an audio editor and judging by this screenshot I took, the db is -6. I also opened a file from my satellite box and that was also -6. So judging by the fact that my box records in that volume I would say that -6dB is a safe volume to convert my Elgato recordings to and that I won't get any problems when playing them in any Bluray player with regards to distortions. What do you guys think?

What's confusing is that when I converted the AAC file to MP3 so I could open it in MP3Gain, MP3Gain said the volume was 89dB. I can't even see that dB number in my audio editor.

How can I find out the true dB volume of a file?


[Attachment 25628 - Click to enlarge]

[VideoFanatic]

Jagabo - I'm currently using this script to mux a h264.ts file with AAC to a TS file with AC3 (AAC is re-encoded to AC3) and demux the original AAC audio into an AAC file.

"X:\Portable Installations\ffmpeg-2014 May 14-git-72dcd48-win64-static\bin\ffmpeg.exe" -i %1 -c copy -absf aac_adtstoasc -vn gain.m4a
"X:\Portable Installations\AAC Gain 1.9\aacgain" -r -k -m 0 gain.m4a
"X:\Portable Installations\ffmpeg-2014 May 14-git-72dcd48-win64-static\bin\ffmpeg.exe" -i gain.m4a -acodec copy "%~d1%~p1%~n1.aac"
"X:\Portable Installations\ffmpeg-2014 May 14-git-72dcd48-win64-static\bin\ffmpeg.exe" -i %1 -i gain.m4a -map 0:v -map 1 -c:v copy -c:a ac3 -ab 2 -ab 128k "%~d1%~p1%~n1.AC3.ts"
del gain.m4a

There's a bug in FFMPEG that causes the resulting TS file to be 5% larger than the original. I've reported the bug to them but I've not heard of a fix yet. Hopefully the bug isn't in MKV conversions.

Any idea how I can do the above script but mux the TS to MKV instead of to a TS file?

[jagabo]

Just change the output file extension from .TS to .MKV.

[VideoFanatic]

I tried that but it gives an error message. Something about incorrect headers.

[jagabo]

Worked fine here.

[gll99]

I tried that but it gives an error message. Something about incorrect headers.

Works here too but you have to change the ext. in the script and not after the fact. I didn't read too far back so maybe the version of ffmpeg is different.

[VideoFanatic]

I'm using the latest 2014-07-27 64-bit FFMPEG version from here: http://ffmpeg.zeranoe.com/builds/

This is my script:

"X:\Portable Installations\ffmpeg\bin\ffmpeg.exe" -i %1 -c copy -absf aac_adtstoasc -vn gain.m4a
"X:\Portable Installations\AAC Gain 1.9\aacgain" -r -k -m 0 gain.m4a
"X:\Portable Installations\ffmpeg\bin\ffmpeg.exe" -i gain.m4a -acodec copy "%~d1%~p1%~n1.aac"
"X:\Portable Installations\ffmpeg\bin\ffmpeg.exe" -i %1 -i gain.m4a -map 0:v -map 1 -c:v copy -c:a ac3 -ab 2 -ab 128k "%~d1%~p1%~n1.AC3.ts"
del gain.m4a
"X:\Daves Folder\Sounds\VideoRedo Completed Sound Short.WAV"
pause

All I did was change the underlined text to "MKV" like you said. When I run the batch file when it starts to convert to MKV it just fails and says this "Could not write header for output file #0 (incorrect codec parameters ?): Invalid data found when processing input".

Having a source file as MKV is no good as FFMPEG adds a 1 minute 5 second delay to the file. So I'd like to try converting from TS to MKV as that only has a -5ms delay which doesn't effect the sync as it's so small it makes no difference.

I would like to convert from TS to MKV. Any idea how to get the script working?