Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > General > Audio encoding

Closed Thread
 
Thread Tools Search this Thread Display Modes
Old 2nd September 2003, 00:18   #41  |  Link
TerraForce1
Registered User
 
Join Date: Apr 2003
Posts: 82
Food is nice

Hi,

First of all I would like to thank everybody pointing out on this.

At first glance I wanted to test if this method worked out for me. I have the following problem and it keeps doing it over and over.

The pitch of the created 6 channel wav is higher compared to the original stereo WAV file, so I end up with a 14 seconds less than the original. The songs ends like it has to end but the pitch is higher compared to its original. The created multichannel wav is like 2:45 as for the original is 2:59. I dont know if anybody has experienced this problem.

I know I just could alter the pitch, but I find it better that the program works flawless.

Could somebody point me out what I am doing wrong or is it bidule itself?
TerraForce1 is offline  
Old 2nd September 2003, 02:20   #42  |  Link
TerraForce1
Registered User
 
Join Date: Apr 2003
Posts: 82
Well, I love the way your mind works

Hi, well as the post subjects says it all:

I love the way your mind works. Meaning you are right.

Thx.
TerraForce1 is offline  
Old 2nd September 2003, 07:37   #43  |  Link
daphy
Miles Freak
 
daphy's Avatar
 
Join Date: Apr 2003
Location: Europe
Posts: 652
@ all

If you have a new profil that works fine for you please contact me via PM. Than send me your settings (***.bidule -file) via email, so I can give others the opportunity to download it and make further development possible

CYA Daphy
daphy is offline  
Old 2nd September 2003, 22:35   #44  |  Link
TerraForce1
Registered User
 
Join Date: Apr 2003
Posts: 82
Awesome results

Hi,

I want to report about the awesome results I got. I own an Eagles DVD. You can choose between DTS or PCM. I have expirmented on the song California Hotel (Acoustic version). Because I own the DVD, I downloaded a mp3 of this song and tried the guide. Unbelievable, you can hear the same effects as they come on the original DTS track as with the new encoded DTS sound. I don't know how, but it looks like the same. But however there is a but, and that is. In the original DVD I hear the BassTablets from Front Left, as with the encoding they come from Rear Right. As from the original DVD I get the hihats from Rear Right as from the Encoded they come from Front Left. Somehow it looks like the sound is swapped between the Speakers. The center and right front speaker also look to be just the way around. I had not much time for more analyzing. But the sound of the original DVD gives with the original speaker setting a better harmonic between the sounds.
My first thought is the setting of the Emigrator where you can set the conrolled opposites. But I may be wrong.

For your information.

TerraForce1

Last edited by TerraForce1; 2nd September 2003 at 22:37.
TerraForce1 is offline  
Old 2nd September 2003, 23:36   #45  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
@ TerraForce1: Thanks for the nice feedback. Three comments:

1) The channel-mapping with the method in the guide is correct (we checked this again and again before publishing). Despite the 'good' results with this method, it is very likely that results are different from those optained with special 5.1-mix (btw: I know California Hotel from the DVD very well too ).

2) The Ambisonic method is more about creating a 'sound-image', not about isolating effects & instruments to discrete channels. Given this, having effects coming out of the same speaker as with the DVD-5.1-mix (with emphasis on mix) should not be expected.

Another way to put this: B-Pan creates an Ambisonic 1st-order B-file (=WXYZ). The fact that we decided for the guide to outline decoding to 5.0 is just because most of us have a 5.1-system. However, decoding could be easily made for 4.0, 8.0, 12.0, etc. etc. speaker rigs

3) You can experiment with the `controlled opposite' (vs. 'strict soundfiled'). Richard Furse finds that the `controlled opposite' equations produce a larger listening area at the expense of some directional information (Ambisonic Decoding Equations . Keep us posted on your preference.

Regards,

Andreas
kempfand is offline  
Old 3rd September 2003, 19:36   #46  |  Link
specise_8472
Guest
 
Posts: n/a
Quote:
Originally posted by Eye of Horus
BTW I just did some tests with the Nuendo, but I don't get it to output anything..... Do you have an idea ?

kind regards,

EoH

It was just an idea - but am working on it at the moment

Also, is there any good reason why in Bidule you cannot just define 6 single channel outputs and output each channel as a single wave file. Then at the end you have your 6 waves all ready to go
 
Old 3rd September 2003, 22:10   #47  |  Link
bitsnbytes
Registered User
 
Join Date: Aug 2003
Posts: 57
Quote:
Originally posted by specise_8472
It was just an idea - but am working on it at the moment

Also, is there any good reason why in Bidule you cannot just define 6 single channel outputs and output each channel as a single wave file. Then at the end you have your 6 waves all ready to go

You can but it is time consuming.
I tried 6 recorders but some generated invalid files.

So I set the six recorders unlinkig all but the first... then unlink the first and link the second...etc
then in my encoder it warned me that there were files of different data lenghths... but song encoded correctly and sounded right.

ps I only did it this way to go from 32bit/48000
to 24 bit/48000 directly

with this manner you dont get normalizing

so a 5 min song gonna take you 30 min.....
results are not worth the extra time I dont think.... but maybe later when they fix it ;-)
__________________
[System] AVP-Rotel RSP-1066; DVD Audio/Video-Rotel RDV-1060; AMP-Rotel RMB-1075; Speakers JBL, Rock Solid And Polk Audio
bitsnbytes is offline  
Old 3rd September 2003, 22:33   #48  |  Link
Eye of Horus
Banned
 
Eye of Horus's Avatar
 
Join Date: Jun 2002
Location: The Netherlands
Posts: 326
Quote:
Originally posted by specise_8472
It was just an idea - but am working on it at the moment

Also, is there any good reason why in Bidule you cannot just define 6 single channel outputs and output each channel as a single wave file. Then at the end you have your 6 waves all ready to go
But..... I want to make AC3's from the 6-in-1 wav.....
I can load this file directly into Soft Encode !!

&@Kempfand
These are the pros and cons of hitchhiking :-)

EoH
Eye of Horus is offline  
Old 3rd September 2003, 22:47   #49  |  Link
bitsnbytes
Registered User
 
Join Date: Aug 2003
Posts: 57
New Test Subject

I got The Edgar Winter Group - "They Only Come Out AT Night" CD from guy at work... I have to say zzz Frakenstein is A masterful piece and made sweet love to my ears.... originally encoded for quadraphonics... I now see why
__________________
[System] AVP-Rotel RSP-1066; DVD Audio/Video-Rotel RDV-1060; AMP-Rotel RMB-1075; Speakers JBL, Rock Solid And Polk Audio

Last edited by bitsnbytes; 3rd September 2003 at 22:53.
bitsnbytes is offline  
Old 3rd September 2003, 23:01   #50  |  Link
specise_8472
Guest
 
Posts: n/a
Interesting find.
Load the resultant 6 channel wave into Adobe Audition and it automatically splits the file into 6 waves. Then add them to multitrack view.
The only problem then is that the only multichannel output formats supported at the moment are Windows Media 9 Multichannel. Will also export 6 wavs and multichannel wave.
Am still playing around with various options and programs.

BTW have you tried the TAPEIT vst plugin (get from vst central) as the recorder for individual wavs.
 
Old 3rd September 2003, 23:08   #51  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
TapeIt

Last edited by kempfand; 3rd September 2003 at 23:17.
kempfand is offline  
Old 3rd September 2003, 23:30   #52  |  Link
bitsnbytes
Registered User
 
Join Date: Aug 2003
Posts: 57
tapeit worked worse than the on board recorder

rendering 1kb files when configured 6 times
__________________
[System] AVP-Rotel RSP-1066; DVD Audio/Video-Rotel RDV-1060; AMP-Rotel RMB-1075; Speakers JBL, Rock Solid And Polk Audio
bitsnbytes is offline  
Old 4th September 2003, 02:52   #53  |  Link
bitsnbytes
Registered User
 
Join Date: Aug 2003
Posts: 57
I pulled the .wav out of a divx to convert but cant seem to get it done.

its microsft acm waveform mpeg layer 3?

cant get it into bidule
__________________
[System] AVP-Rotel RSP-1066; DVD Audio/Video-Rotel RDV-1060; AMP-Rotel RMB-1075; Speakers JBL, Rock Solid And Polk Audio
bitsnbytes is offline  
Old 4th September 2003, 06:45   #54  |  Link
daphy
Miles Freak
 
daphy's Avatar
 
Join Date: Apr 2003
Location: Europe
Posts: 652
Quote:
its microsoft acm waveform mpeg layer 3?
I always thought bidule works only with non compressed files?
Use for example cooledit´s save as option to save into a non compressed WAV (PCM 16bit 48/44.1khz)

CYA Daphy
daphy is offline  
Old 4th September 2003, 07:23   #55  |  Link
TerraForce1
Registered User
 
Join Date: Apr 2003
Posts: 82
a long the 24 bit 32 bit story

Hi,

Along the 24 bit 32 bit story I have the problem that the recorded amplitude with Bidule is very low. I don't know if it is possible to get the recorded audio at a higher amplitude. I know that Besweet is capable of doing this (btw I don't know the acceptable settings guess -13 db will do), but is there somehow a way in Bidule to get it correct from the beginning ? As for my second question, while maintaining a higher amplitude of the surround speakers I want my LFE output to be of a lower amplitude than the current settings. It's really too loud. I have to lower the input output ratio on my subwoofer otherwhise I get a distorted effect. It's like going to a houseparty with gabbers or to put it in other words real hardcore bassdrums. Well and that ain't what you are expecting when you listen to slow rock.

Watching the waveforms with Cooledit gives a clear image that the amplitude of the LFE is really much bigger than the other waveforms. So for the moment I have to manually cut 6 db every time.

As my conclusion is: from how I want to have it, the amplitude of the LFE had to be the amplitude of my surround speakers. And the amplitude of my surround speakers had to be the amplitude of my subwoofer.

Is there somehow an amplitude plugin for Bidule. If yes how do we integrate this in the current model.

@ daphy
Also compressed wave file are accepted. I saved my wave file in an ADPCM format and it played the file. Mp3's however dont.

TerraForce1

Last edited by TerraForce1; 4th September 2003 at 07:31.
TerraForce1 is offline  
Old 4th September 2003, 17:37   #56  |  Link
davidv@plogue
Registered User
 
Join Date: Aug 2003
Location: Montreal
Posts: 29
Hello

1) There is allready a Gain module in Bidule,
in the toolbar (or right mouse button menu) under mixing/gain...
its a simple 2 in 2 out gain module

2) Sadly while our multichannel wav files are following the wav standard to the letter, since that format is uncommon, not many apps read them properly (ex MS Media player). I knew about Cool Edit and Audacity, but i didnt know about Adobe Audition, which is good to hear.

3) Bidule is a realtime application, and while we are very happy to learn its being used in this fashion (im waiting to set up my own 5.1+ setup to try this very soon, when my home theatre is done), It wasnt
made specifically for this task.

That said, this thread has waken up the need for a Offline (render) mode which would allow to process much faster than real time
(say, if in realtime bidule reports 10% cpu use, the offline mode would able - in theory process that 10 times faster) But you wouldnt be able to monitor it at the same time.

4)I have to look at why the settings are not saved properly, it seems its a bug that seem to affect soem VSTs and not others.

Cheers
davidv@plogue is offline  
Old 4th September 2003, 17:45   #57  |  Link
davidv@plogue
Registered User
 
Join Date: Aug 2003
Location: Montreal
Posts: 29
Sorry other things just crossed my mind

5)we dont support mp3 natively, but a vst plugin does that: mp3play
(the site however down to protest against software patents) which we also are against.

6) on Bit depth. Bidule uses 32bit IEEE floating points data internally
and thats the signal that's passed though the graph. So even if a plugin uses 64 bits internally, it wouldnt make much of a difference outside of it. Secondly, wav files stored at 24 bits (linear) contain as much bit headroom as a Float 32 bit IEEE wav file.
(techy talk
A 32 bit Float used for audio uses the same precision as a 24 (linear) integer, since the remaining 8 bits are used for exponent.
davidv@plogue is offline  
Old 4th September 2003, 17:47   #58  |  Link
davidv@plogue
Registered User
 
Join Date: Aug 2003
Location: Montreal
Posts: 29
7) for a list of natively supported audio files in bidule,
have a look at this list:
http://www.zipworld.com.au/~erikd/libsndfile/#Features

(bidule uses this library in order to read/write audio files)
davidv@plogue is offline  
Old 4th September 2003, 19:11   #59  |  Link
Eye of Horus
Banned
 
Eye of Horus's Avatar
 
Join Date: Jun 2002
Location: The Netherlands
Posts: 326
Quote:
Originally posted by davidv@plogue
Hello

1) There is allready a Gain module in Bidule,
in the toolbar (or right mouse button menu) under mixing/gain...
its a simple 2 in 2 out gain module


Hi David,

But.... how much gain do you have to give ?
Is there a way like OTA's -g max ??

Quote:

2) Sadly while our multichannel wav files are following the wav standard to the letter, since that format is uncommon, not many apps read them properly (ex MS Media player). I knew about Cool Edit and Audacity, but i didnt know about Adobe Audition, which is good to hear.


Adobe Audition = Coold Edit Pro 2.1 (They bought the program.)

Quote:







3) Bidule is a realtime application, and while we are very happy to learn its being used in this fashion (im waiting to set up my own 5.1+ setup to try this very soon, when my home theatre is done), It wasnt
made specifically for this task.

But it does a good job !!!!

Quote:

That said, this thread has waken up the need for a Offline (render) mode which would allow to process much faster than real time
(say, if in realtime bidule reports 10% cpu use, the offline mode would able - in theory process that 10 times faster) But you wouldnt be able to monitor it at the same time.


The value on the bottom of the screen gives sometimes strange results.... i don't know what really happens then, or if it's a bug somewhere in the "memory and cpu management".
load a module and use it...... cpu : 2.6 %
reload the module, do some modifications (other file names) : cpu 80 % and once even 105.67 % > hard reset to be able to work again....


Quote:

4)I have to look at why the settings are not saved properly, it seems its a bug that seem to affect soem VSTs and not others.

Cheers
And why it changes one setting without that setting being touched !


kind regards,

EoH
Eye of Horus is offline  
Old 4th September 2003, 20:13   #60  |  Link
davidv@plogue
Registered User
 
Join Date: Aug 2003
Location: Montreal
Posts: 29
>But.... how much gain do you have to give ?

you can do -oo to +10 dB (nominal - unchanged - gain by default)

>Is there a way like OTA's -g max ??

You mean normalize?
Im afraid thats impossible to normalise a signal in
a live (streaming situation) unless someone invents crystal-ball VST,
as an audio signal normalizing algorithm must

1)first scan the audio signal entirely
2)remember the highest (peak) value it encountered
3)find a gain value x so that high val = max
4)apply gain x to whole file

none of which can be done in a live streaming app, like this.
(but are easily done in offline sound editors)

Engineers use Dynamics Compressors in realtime
(with which they can use to prevent sudden unexpected peaks),
but these, well change the source dynamics.. so you might not want that here.

The best way would be with trial and error, run a couple of source
files and find a gain setting that doesnt clip the signal.
(this plugin might help: http://www.pspaudioware.com/plugins/vmeter.html)

Our audio file player extracts the file at nominal volume, and
audio files have a max "normalized" value they can support
(well the 32 bit float can go beyond that and represent signal
out of the normalised [-1;1] range, but i dont want to get too techy again), So im sure there is a propper gain setting, but the overall gain setting might be influenced by the settings of any of the other plugins in the chain.

>Adobe Audition = Coold Edit Pro 2.1 (They bought the program.)

Yeah, youre right, my plogue collegue just told me.

>But it does a good job !!!!

Hey thanx!

>The value on the bottom of the screen gives sometimes strange results....
>i don't know what really happens then, or if it's a bug somewhere in the "memory and cpu management".
>load a module and use it...... cpu : 2.6 %
>reload the module, do some modifications (other file names)
>: cpu 80 % and once even 105.67 % > hard reset to be able to work again....

it can be many things, pentium 4 denormalisation bug in (any) VST,
I/O disk thread code having problems (bidule bug).
Or something else weve never encountered before.
If you can find specific steps to reproduce these errors, we would glady
try to fix them.

Last edited by davidv@plogue; 4th September 2003 at 21:29.
davidv@plogue is offline  
Closed Thread

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 22:36.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.