CCE Version 2.67.00.23 + MPA Audio

[GParent]

I just installed CCE Version 2.67.00.23.

Up until now, I have always stripped audio from my AVI files using VIRTUAL DUB. I always wondered why an encoder like TMPGEnc seems to create a good audio file during encoding. Yet, I always followed the conventional wisdom about staying clear of CCE's audio file generation.

Now, purely because I installed a new CCE version and wanted to experiment, I decided to try a sample clip from my AVI capture (using Canopus ADVC-100) and generate a MPA file from CCE, which I used to author a DVD, and it sounds fine to these ears.

TWO QUESTIONS?

(1) What were the original reasons and circumstanes that caused users to recommend not using CCE to generate audio output?

(2) Is CCE Version 2.67.00.23 improved enough to recommend it for generating MPA output?

Thanks,

Gary

[bel]

Almost all do not like mpa created by cce.
For you and me it's OK.
Best of all, for me, I never never never run into a/v sync problem.

[GParent]

I'm not saying its as good as uncompressed PCM audio playing stereo music, but if you're generating a DVD from a Videotape with mostly dialogue and minimal music, or just background music, it seems a convenience to gen the compressed MPA audio and be done with it. Previously, I have to strip off a WAV file using Virtual Dub and whip up an AC3 Dolby 2/0 file using BeSweet. With the MPA, its good enough, IMHO.

Also, thanks for the feedback about not having sync problems!

[dumb13]

if you're generating a DVD from a Videotape with mostly dialogue and minimal music, or just background music, it seems a convenience to gen the compressed MPA audio and be done with it

I agree, this is a convenience - however, I also experienced clicks and pops in mpa audio encoded by CCE, which is rather annoying, even in simple cases you described.

What I am doing is I use TMPGEnc to compress audio in a separate run (after CCE is done with the MPEG2) - using DVD compliant audio settings, and output audio stream only. It is really fast, and produces high quality. This way there is no need to extract audio to wav.

So far, I have not experienced any sync problem - I mostly work with DV (AVI) footage, or archived VHS footage digitized through my DV-cam's A/D converter.

Regards
d

[Matthew]

I wouldn't trust TMPGEnc's internal routines...won't toolame do this? You can link it in TMPGEnc's options too.

[GParent]

>>I agree, this is a convenience - however, I also experienced clicks and pops in mpa audio encoded by CCE, which is rather annoying, even in simple cases you described.<<

I will be extra aware of the pops and clicks you mentioned. However, I have not had a single problem since I began generating MPAs from CCE. Is it possible that because I'm using the mist current CCE Version: 2.67.00.23 that Cinemacraft could have improved their audio extraction? Just a thought.

Regards your using TMPGEnc for stripping audio WAV files, I have also used that method and it works! But I have also gotten incredibly good results rendering a WAV file out of VEGAS 4. After all, Sonic Foundry knows a bit about audio.

Thanks,

Gary

[dumb13]

Originally posted by Matthew
I wouldn't trust TMPGEnc's internal routines...won't toolame do this? You can link it in TMPGEnc's options too.

Thanks for the hint, have never took a look into that part of TMPGEnc - certainly toolame and lame would do a much better job in layer 2 and 3 encoding.

May I have one question re this: I read in the TMPGEnc Help, that I am supposed to configure the external tool in order to avoid problems. Unfortunately, I have no idea of how to do this, although have been using lame and toolame, but in differnet frontends.

Thanks for any direction on this
Regards
D

[bb]

dumb13,

it's easy, provided that you know where it's hidden

- In TMPGEnc, select Option > Environmental setting
- Select the "Audio engine" tab
- Click "External program" under "MPEG-1 Audio Layer II encoder"
- Browse to your tooLAME.exe
- Click "External program" under "Sampling frequency convertor"
- Browse to your ssrc.exe
- Click "OK" to confirm

bb

[dumb13]

Thank you, BB.

- In TMPGEnc, select Option > Environmental setting- Select the "Audio engine" tab
- Click "External program" under "MPEG-1 Audio Layer II encoder"
- Browse to your tooLAME.exe

I already have done this, however

- In TMPGEnc, select Option > Environmental setting
- Click "External program" under "Sampling frequency convertor"
- Browse to your ssrc.exe

I have not yet set this - now it is also done.

Now all set, and working fine - do I assume right, that all other audio encoding parameter settings are passed from the TMPGEnc settings/template to Toolame automatically?

Thanks and Regards
d

[bb]

Yes, at least I assume it does
It will create a temporal wav file first, then execute the external programs to create the mp2. Takes longer, but quality is excellent.

bb

[tempetai]

Hi All, need some advise. I don't know whether I should post this under avisynth usage or CCE. But, I guest, it should be under CCE.

I tried VBR with CCE and CCE created three files (MPA, MPV and VAF). The problem that I got was the MPA have no sound. My source is DV PAL with 48kHz. I used avisynth to resample the audio to 44.1K as my final output will be xCVD. My avisynth script is as below:

Avisource("myfile.avi")
ConvertToYUY2(interlaced=true)
SeparateFields()
SomeFilterHere()
Interleave()
LanczosResize(352,288)
Weave()
ResampleAudio(44100)

First I thought I have to merge back the audio and video. So, I used TMPGENC simple multiplex function. But, the result still the same, no sound. Any idea what I have done wrong.

[bb]

tempetai,

if I were you I would not use CCE to create the MP2 audio file (MPA), because it is known to produce inferior quality. Instead save the audio as a WAV file from within your video editor, then convert it to MP2 using BeSweet (or tooLame, if you know how to deal with command lines).

Just for the record: the VAF file is necessary only if you do more than one video encoding pass in CCE. For single pass, you should disable the creation of VAF files. For xCVD you'll probably want to do more than one pass, though.

bb

[tempetai]

Thanks bb,

Since I don't really sure what parameter to pass to toolame command line, I have used tmpgenc to call the toolame to resample the audio for me. It works great. Yes, I used VBR-2 pass with CCE and the video also looks really good and fast if I were to compare with the speed that tmpgenc can offer. Then I used simple multiplex on tmpgenc to combine both mpa and mpv into mpg. They are in-sync.

[bb]

You can use the BeSweetGUI to encode with tooLAME. By the way: tooLAME is an MP2 encoder; if you need to resample your audio, i.e. change the sampling rate e.g. from 44.1 kHz to 48 kHz, then you can use SSRC. BeSweet (together with the GUI, if you like) allows to chain SSRC and tooLAME, so that it's only one step. So there's no need to use TMPGEnc if you want to encode with CCE anyway.

bb

[tempetai]

bb, Thanks for the info. It is a justification for me to try these tools.

By the way, the video that I encoded with CCE look good on PC but it is like a scratched CD on my DVD player. I've tried tmpgenc and mainconcept as well and both gave me a very good video on DVD player. Is there anything that I have to set on the video setting of CCE if I were to encode for CVD?

[bb]

Which bitrate did you use? The settings for CVD should be similar to those for SVCD, except for the resolution. Do you burn to CD-R media or DVD-R?

bb

[tempetai]

bb,

I'm using the same setup as I create SVCD except the horizontal resolution. By the way, I have succesfully create my CVD last night with CCE. I just need to untick all options on Video Setting tab.

Thanks.