using x264 encoder with audio/video in one step possible ?

[vhelp]

x264 encoder,

what is the secret being used to encode audio/video in one step via command line ?

my current process:

1. capture analog video, directv
2. edit via timeline in virtualdub via avisynth script, it includes both audio/video
3. encode to video-only, mkv via x264 command line, since we can't include audio in this step
4. from step 2, with virtualdub, save audio out to wav file
5. Tools\faac\faac.exe --- (FAAC 1.24.1 (May 17 2005) UNSTABLE)
5. "audio.wav" -q 128k -o "audio.aac" --- (mkvmerge v2.0.2 built on Feb 21 2007 23:40:43)
6. Tools\mkvmerge\mkvmerge.exe
6. "audio.aac" -A "d:\video_only.mkv" -o "video_audio.mkv"
7. finished

this has turned into a laborous task since i tend to break encodes into sections (cut commerails) and encode each piece separately. even if i mux all pieces together, its still a lot of work.

specs include:
amd 2 core, windows xp, 2gig ram, built-in video* card, 8 TB usb drives

thank you in advance!

[poisondeathray]

Don't join them after encoding, you're asking for trouble and the possiblity of non union. Use trim() and aligned splice in the avs script



You can approach it with multiple steps , or with multi-batch file, but using a binary that has audio support is a lot faster

Taro compiles them with audio support (either qtaac, or faac, or lame mp3 audio options) , plus a bunch of other patches along with 10bit/8bit versions. These "tmod" builds are basically "everything under the sun" builds. If you don't want the dozen other patches you'll have to compile it yourself with audio

Read the --fullhelp for full instructions, but basically you can do everything in 1 line (audio/video/muxing)

https://astrataro.wordpress.com/category/encode/x264/

[El Heggunte]

[edited after reading PDR's reply]

Hi there again.

If I am not misunderstanding you , you want to use a recent build of ffmpeg ---

--- which includes "libvo-aacenc", and you may have to recompile the source-code in order to include "libfaac".

Besides...

--- upgrade your FAAC to version 1.28

---upgrade your MKVtoolnix to 5.x.x (or to 4.9.1 at least)

[poisondeathray]

yes, ffmpeg/ffmbc is another option, but the switches/syntax is slightly different than x264cli , and the audio is definitely worse than qt

[vhelp]

pdr--it works.. it encodes aud/vid in one step. mucho thank you!

now the aac i'm worried about because mediainfo says its aac but it does not give me the bitrate.

the one negative thing about this x264_32_tMod+MixAQ-8bit-420.exe build is the stats or progress it reports. first, you have to add it to your param or you get nothing. but its not the same report that the actual x264.r2184.exe gives, i.e., in the mod, the frame number will sometimes show and other times not. its very confusing. so i can't lock onto it to display the current frame picture being encoded in my tool suite. also, it does not give you the length of time it will take. so stairing at the progress can be frustrating.

re the trim....i do use the trim in the script, just not all the time. when i do use'em, the method i use is:

#a=trim( 00154,0).ivtc
#b=trim( 24924,0).ivtc
#c=trim( 40914,0).ivtc
#return a+b+c # full episode, 2012-05-16 wed

it was hard finding a decent aac cmd line encoder. i read many different suggestions of best tool but goolge'ing for them turned up mostly virus webpages or pages full of virus/trojan downloads not to mention the people suggesting them did not include a d/l link. i don't remember where i d/l'ed the faac....same with mkvtoolnix aka mkvmerge. these were all trial and error, finds.

audio--i encode to 128 because i ready somewhere that it is virtually transparent. but i haven't tested that for myself. i am just taking it in on good faith. with the x264 mod, i added these to the param, i assume i am encoding to aac and 128 Kbit/s:

--acodec aac --abitrate 128 --verbose

and last, i am still determinging the best param setup for maximum picture quality. i want as transparent to my source (directv) as possible. at this point in my video endeavors, i feel that adding avisynth script "filtering" (cleaning) is an unecessary step that adds extra time to the encode. and lets face it, i am not getting any younger. at least i am very close to setteling on a few param setups.

i assume stereo, media info:

Audio
Count                            : 157
Count of stream of this kind     : 1
Kind of stream                   : Audio
Kind of stream                   : Audio
Stream identifier                : 0
ID                               : 2
ID                               : 2
Unique ID                        : 2
Format                           : AAC
Format/Info                      : Advanced Audio Codec
Commercial name                  : AAC
Format profile                   : LC
Codec ID                         : A_AAC
Codec                            : AAC LC
Codec                            : AAC LC
Codec/Family                     : AAC
Duration                         : 59768
Duration                         : 59s 768ms
Duration                         : 59s 768ms
Duration                         : 59s 768ms
Duration                         : 00:00:59.768
Channel(s)                       : 2
Channel(s)                       : 2 channels
Channel positions                : Front: L R
Channel positions                : 2/0/0
Sampling rate                    : 48000
Sampling rate                    : 48.0 KHz
Samples count                    : 2868864
Compression mode                 : Lossy
Compression mode                 : Lossy
Delay                            : 0
Delay                            : 00:00:00.000
Delay, origin                    : Container
Delay, origin                    : Container
Delay relative to video          : 0
Video0 delay                     : 0
Language                         : en
Language                         : English
Language                         : English
Language                         : en
Language                         : eng
Language                         : en

media info from faac v1.24:

** edit: i re-encoded w/ x264.r2184.exe, and wave2aac, and merged with mkvmerge.exe **

Audio
Count                            : 157
Count of stream of this kind     : 1
Kind of stream                   : Audio
Kind of stream                   : Audio
Stream identifier                : 0
ID                               : 1
ID                               : 1
Unique ID                        : 1704685750
Format                           : AAC
Format/Info                      : Advanced Audio Codec
Commercial name                  : AAC
Format profile                   : LC
Codec ID                         : A_AAC
Codec                            : AAC LC
Codec                            : AAC LC
Codec/Family                     : AAC
Duration                         : 59797
Duration                         : 59s 797ms
Duration                         : 59s 797ms
Duration                         : 59s 797ms
Duration                         : 00:00:59.797
Channel(s)                       : 2
Channel(s)                       : 2 channels
Channel positions                : Front: L R
Channel positions                : 2/0/0
Sampling rate                    : 48000
Sampling rate                    : 48.0 KHz
Samples count                    : 2870256
Compression mode                 : Lossy
Compression mode                 : Lossy
Delay                            : 0
Delay                            : 00:00:00.000
Delay, origin                    : Container
Delay, origin                    : Container
Delay relative to video          : 0
Video0 delay                     : 0

[poisondeathray]

the one negative thing about this x264_32_tMod+MixAQ-8bit-420.exe build is the stats or progress it reports. first, you have to add it to your param or you get nothing. but its not the same report that the actual x264.r2184.exe gives, i.e., in the mod, the frame number will sometimes show and other times not. its very confusing. so i can't lock onto it to display the current frame picture being encoded in my tool suite. also, it does not give you the length of time it will take. so stairing at the progress can be frustrating.

I'm not sure what your're talking about ? It gives more information realtime than the vanilla builds in the cmd box , it tells you eta. If you mean a printed report you can use --log-level and specify a printed text file e.g. 2>log.txt . This is the same as the vanilla builds

re the trim....i do use the trim in the script, just not all the time. when i do use'em, the method i use is:

#a=trim( 00154,0).ivtc
#b=trim( 24924,0).ivtc
#c=trim( 40914,0).ivtc
#return a+b+c # full episode, 2012-05-16 wed

The danger here is you should IVTC before you trim(), not after - otherwise you risk combed transitions

And as mentioned before, joining streams after encoding is more prone to error and problems than doing it as 1 encode

it was hard finding a decent aac cmd line encoder. i read many different suggestions of best tool but goolge'ing for them turned up mostly virus webpages or pages full of virus/trojan downloads not to mention the people suggesting them did not include a d/l link. i don't remember where i d/l'ed the faac....same with mkvtoolnix aka mkvmerge. these were all trial and error, finds.

WHHAAAT ??? not difficult to find at all. neroaac has been around for years and one of the most popular

neroaac, qaac, qtaac .... these should be the 1st few links on google

audio--i encode to 128 because i ready somewhere that it is virtually transparent. but i haven't tested that for myself. i am just taking it in on good faith. with the x264 mod, i added these to the param, i assume i am encoding to aac and 128 Kbit/s:

--acodec aac --abitrate 128 --verbose

and last, i am still determinging the best param setup for maximum picture quality. i want as transparent to my source (directv) as possible. at this point in my video endeavors, i feel that adding avisynth script "filtering" (cleaning) is an unecessary step that adds extra time to the encode. and lets face it, i am not getting any younger. at least i am very close to setteling on a few param setups.

Read the help file, you should use --acodec qtaac to active qt . It really is noticably better than faac (neroaac and qt might be similar, but both are noticably better than faac I think most will agree). TVBR mode is better than ABR (it's like CRF mode for video encoding vs. ABR... nobody uses ABR with x264). Similarly only if you need a set bitrate should you use birate mode. --aquality ranges from 0-127 for VBR mode, higher values give higher quality and bitrate (opposite to quantizer value in video). Read the --fullhelp for the settings and see the reviews and discussion and tests at hydrogenaudio if you want more information on qtaac or qaac .

If you use qtaac, mediainfo reports everything like bitrate in the final file

[poisondeathray]

It might be that you are using mkv container- audio bitrate isn't always reported by mediainfo in mkv container

(And you should ditch the fixation on using set bitrates - it's a bad approach to video or audio if you want "ideal" quality)

So you can do all the basic stuff with the tmod builds (they even have cropping, padding, yadif, lavf) so you don't even need avisynth or splitters , decoders

[El Heggunte]

FAAC 1.28:

http://www.rarewares.org/aac-encoders.php

fhgaacenc:

https://github.com/tmkk/fhgaacenc

qaac:

http://sites.google.com/site/qaacpage/

Nero "is a name known to all" , and

enc_aacplus is regarded as "outdated stuff" now

P.S.: Last but not least, MKVtoolnix @

http://www.bunkus.org/videotools/mkvtoolnix/win32/

[poisondeathray]

So you can do all the basic stuff with the tmod builds (they even have cropping, padding, yadif, lavf) so you don't even need avisynth or splitters , decoders

Actually I take that back. The 1 big thing that should be considered "basic" that is missing in ffmpeg/mplayer/libavcodec/libavfilter etc.... is a working IVTC filter.

So you still need avisynth for some basic things (and of course more advanced filtering and processing)