Wavewizard v0.54b

[johnman]

Thx for your testing report. Its very detailed, which is very helpfull to narrow down those nasty bugs.

Since a and b seem to work im going straigth to c.



c) wav6 -> ac3 (Soft Encode)

- Channel Mapping: Disable (See Configure notes)
- Stream manipulation: Enable (Mono Streams)
- Output Format: Wave PCM (Don't work with wav_ex)
- Sample Type: Work at least with: 16bit int, 32bit int, 32bit float IEEE
- Configure notes:
1) Channel config: with L-0, R-1, C-2, LFE-3, LS-4, LR-5 is not necessary enable the ww Channel Mapping (only necesary when you make a re-ordered wav6 to be open directly in SoftEncode)
2) LFE Enable (necessary for 5.1). Then you insert, in SonnicAc3.ini, "UseLFE=On". SoftEncode make a 5.0 ac3, because a bug or a bad chosen name need "UseLFE=Off" to make a 5.1 ac3. I use:
Sonic Foundry Soft Encode Version 1.0 (Build 19)
Dolby Digital Encoder Version 6.2.2 December 2, 1997


1)The channelmappings supplied are from other users, they are not tested by me. I added them since they might be usefull to others.
2)If i understand you correctly, you are saying i have to set the UseLFE to off to enable it? Im checking it out myself to see whats going on.


d) wav_ex 6chan <-> 6 mono wav_ex

Correct split in 6 mono wav_ex preserving the ChannelMask.
Incorrect Merge of 6 mono wav_ex ignoring the ChannelMask.


Only the splitting preservers the mask, when merging or stitching the mask is set to 0. This is done intentionaly, because what should ww do if 2 files with the same mask were added. So to quickly solve this is just set the mask to 0.
I am thinking about an option to manualy set the mask, and give a hint based on the existing masks.



e) wav > 4GB

Opened when enable "Ignore invalid wav size". Length limited to 4GB. I think, in this case, the length must be calculated with the file size.

Whenever the size in the header is ignored, everything behind the start of the datachunk is treated as sampledata. So it should read until the end of the file.


I know the problem with the extrachunks at the end of file, but is better to have little clicks at the end, than lose MB of correct data.
When split in 6 mono wav only the first 4 GB are converted, the rest are ignored.
If not all data is converted, then something is wrong. ww doesnt cut away a couple of MB's just to be sure you dont get a pop. I just created a 6 channel file of 6 gb and splitted it into 6 monostreams.. Alls of them have the correct length :confused:
Are you sure you had enough diskspace?

[tebasuna51]

1)The channelmappings supplied are from other users, they are not tested by me. I added them since they might be usefull to others.
This ww channelmapping is supplied for me, before know the channelmapping inside the Configure of Soft Encode. Is only to explain when must be used or not.
2)If i understand you correctly, you are saying i have to set the UseLFE to off to enable it? Im checking it out myself to see whats going on.
Yes. I tested it. Maybe a bug of my version of Soft Encode, Version 1.0 (Build 19)
Whenever the size in the header is ignored, everything behind the start of the datachunk is treated as sampledata. So it should read until the end of the file.

If not all data is converted, then something is wrong. ww doesnt cut away a couple of MB's just to be sure you dont get a pop. I just created a 6 channel file of 6 gb and splitted it into 6 monostreams.. Alls of them have the correct length :confused:
Are you sure you had enough diskspace?
Yes, I work in a NTFS partition with more than 20 GB of free space.
The same file is separated by BeSplit with correct length (but with a error in the BlockAlign field, keeping the wav_ex header with a incorrect Channelmask corresponding to the wav_ex 6 chan.).
The file is a wav_ex generated by Faad with only 4.2 GB (130 min., 48 KHz) and the mono wav extracted have exactly 4/6 = 0.666 GB (124 m. 16 sec., 48 KHz).
The two erroneous fields in the wav_ex 6chan (RIFF_chunk_size and data_chunk_size) have values corresponding to the excess over 4 GB (aprox. 6 min., 48 KHz). Without extrachunks at the end of file.

The only check options in ww Preferences are:
- Ignore invalid wav size
- Save output
- Analyse files inmediately
- Stream manipulation -> Mono streams
- Output format -> Wave PCM

Thanks.

[johnman]

Im just guessing here, but it might be that the size is vallid. If the size is set to 4 GB ww will not use the full file. That is also why ww sets the size to 0. A way to test this is to unselect ignore invallid size, and then to drop the file. If it accepts it, then the size is vallid.

Im going to solve this "bug" by adding the options
- always ignore the size and
- automaticaly ignore the size when its larger then 2gb/4gb

If this is not the problem, it might be handy if you could post the first 100 bytes of the wav. Then ill can experiment with it myself.
it would look something like this:

52 49 46 46 44 CB 36 83 57 41 56 45 66 6D 74 20
10 00 00 00 01 00 06 00 44 AC 00 00 30 13 08 00
0C 00 10 00 64 61 74 61 20 CB 36 83 01 00 01 00
01 00 01 00 01 00 01 00 FF FF FF FF FF FF FF FF
FF FF FF FF 01 00 01 00 01 00 01 00 01 00 01 00
FF FF FF FF FF FF FF FF FF FF FF FF 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00


EDIT and i can confirm the bug in sonic surcode. I have to set the option to off to enable it :confused: I think i will add an option to cb to invert the flag.

EDIT 2 and i got the same verion of surcode as you

[johnman]

I've uploaded ww v0.53 to needfullthings.

No real changes, only a couple of bugfixes
- option to invert includeLFE flag in cb (default is to switch the flag)
- new options to ignore size in header
- and 3 other minor issues.
- the channelmask is still set to 0 when merging/stitching file. This will be changed when i add an config option for settig the channelmask on all wavs created.

please check the new version tebasuna51

[tebasuna51]

Sorry. I mistake my precedent comment. WW don't stop at 4 GB, stop at the invalid length.
I decode a 130m12s.aac with Faad v2.1b to 6faad_ex.wav, and with Foobar2000 v8.3 to 6Foobar.wav.
The two wav are open like Type: Unknown with Ignore invalid wav size Disabled.
With Ignore invalid wav size Enabled can split them in mono wav.
I report the data in WW (I add the first line, aac, to compare length), and the wav headers of two wav 6 channel.


Filename Size Length Channels SampleRate Type
130m12s.aac 112.028 KB 2:10:12.459 6 48000 aac
6faad_ex.wav 4.394.508 KB 2:04:16.540 6 48000 Wav Ext.
6faad_ex_ch1_R.wav 699.051 KB 2:04:16.540 1 48000 Wave PCM
6Foobar.wav 4.394.508 KB 05:55.918 6 48000 Wave PCM
6Foobar_ch1.wav 33.367 KB 05:55.918 1 48000 Wave PCM
-------------------------------------------------------------------------
Offset 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6faad_ex.wav
------------
0000 52 49 46 46 00 FF FF FF 57 41 56 45 66 6D 74 20 RIFF.ÿÿÿWAVEfmt
0016 28 00 00 00 FE FF 06 00 80 BB 00 00 00 CA 08 00 (...þÿ.._»...Ê..
0032 0C 00 10 00 16 00 10 00 3F 00 00 00 01 00 00 00 ........?.......
0048 00 00 10 00 80 00 00 AA 00 38 9B 71 64 61 74 61 ...._..ª.8>qdata
0064 00 FF FF FF B9 FF 00 00 00 00 00 00 00 00 00 00 .ÿÿÿ¹ÿ..........

FileSize: 4.499.976.260 = 0x10C383044
RIFF_chunk_size = FileSize - 8 = 0x10C38303C Invalid: 0xFFFFFF00
data_chunk_size = FileSize - 68 = 0x10C383000 Invalid: 0xFFFFFF00
-------------------------------------------------------------------------
Offset 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 6Foobar.wav
-----------
0000 52 49 46 46 24 30 38 0C 57 41 56 45 66 6D 74 20 RIFF$08.WAVEfmt
0016 10 00 00 00 01 00 06 00 80 BB 00 00 00 CA 08 00 ........_»...Ê..
0032 0C 00 10 00 64 61 74 61 00 30 38 0C B9 FF 00 00 ....data.08.¹ÿ..

FileSize: 4.499.976.236 = 0x10C38302C
RIFF_chunk_size = FileSize - 8 = 0x10C383024 Invalid: 0x0C383024
data_chunk_size = FileSize - 44 = 0x10C383000 Invalid: 0x0C383000


Maybe WW take like valid anything distinct of 0xFFFFFFFF or 0x00000000, but you can see there are two programs than fill data_chunk_size with different values.
I think if FileSize > RIFF_chunk_size + 8, the Length can be calculated with FileSize - 8 - Offset_data_chunk (ignoring possible extrachunks at the end of the file).

Thanks for your interest.

[johnman]

Thx again for your detailed report. I managed to pinpoint the exact problem, and im pretty sure it is fixed in v0.53. The problem was that there are multiple chunks, which each have their own size . The RIFF chunk is the filesize -8, and because the riffchunk was wrong ww did not accept the file. But the data chunk size was "correct". It is to small, but only a size that's to big, or a size of 0 would be rejected. Since ww sets the size to 0 if an sizeoverflow occures, it will reject the size, but other programs just set the size to an invallid number, so the size was not rejected.

Anyway, if you use the new version and select "ignore the size in the header" you're ok.

And to check if the problem is fixed, you only need to drop the wav into ww. If the length is displayed correct (2:10:12.459) , then it should run fine.

[johnman]

Although i doubt im going to get any help, im gonna ask it anyway :rolleyes:

I just finished a chunk analyser to show all the chunks in the wav, and i would like to test it a little, so again i need some testfiles. If anyone know a site like this (http://www.tsp.ece.mcgill.ca/MMSP/Documents/AudioFormats/WAVE/Samples.html) with some example files i would like to get a link.

And here is a little screenshot:

http://www.rarewares.org/wavewiz/chunks.png

And since im posting anyway, im thinking about letting ww also show all the data of other RIFF formats like avi. Both avi and wav's have the same basic RIFF structure, so it wouldnt be that much works (i think). But this doesnt really fits into ww so im only doing this if there isnt any other tool that can do this. So.... anyone wants to comment about this? Should ww be able to analyse avi's?

[tebasuna51]

I think is enough with the examples in this site. Of course can have many types with obsolete compression and different kinds of parameters and extrachunks, but for me is not necessary.

About the avi I think is a bit more complex. Do you know abcAVI Tag Editor? http://abcavi.tk/

[johnman]

I think is enough with the examples in this site. Of course can have many types with obsolete compression and different kinds of parameters and extrachunks, but for me is not necessary.

About the avi I think is a bit more complex. Do you know abcAVI Tag Editor? http://abcavi.tk/


I decided to support it, since im also thinking about letting ww demux whatever audio is present in the avi (although ww can only use the audio itself if its wav ofcourse)

[Zyphon]

I decided to support it, since im also thinking about letting ww demux whatever audio is present in the avi (although ww can only use the audio itself if its wav ofcourse)

This is good news johnman. I wish you luck implementing this and I hope the coding goes well. :)

[bobcat56458]

Hello johnman: Thanks for the great program, I’m using version 0.53. It works well converting a 5.1 Wave file into a DD5.1 AC3 file using the linked to ac3enc.dll, but when I try to do the same using Soft Encode version 1.0 (Build 19) I get the error message [CB error on: “name of 5.1 wave file.wav”. The configuration of conversion batcher requires 6 monostreams(s).] Now I’m able to load this same 5.1 Wave file directly into Soft Encode and it is able to convert it to a DD5.1 AC3 file. I could convert this 5.1 Wave to 6 mono files using BeSweet, but it would be much better to do all of the job in Wavewizard & its Batch Converter. Is what I’m doing possible using your program, and using Sonic Foundry Softencode as the Output type?

[johnman]

Hello johnman: Thanks for the great program, I’m using version 0.53. It works well converting a 5.1 Wave file into a DD5.1 AC3 file using the linked to ac3enc.dll, but when I try to do the same using Soft Encode version 1.0 (Build 19) I get the error message [CB error on: “name of 5.1 wave file.wav”. The configuration of conversion batcher requires 6 monostreams(s).] Now I’m able to load this same 5.1 Wave file directly into Soft Encode and it is able to convert it to a DD5.1 AC3 file. I could convert this 5.1 Wave to 6 mono files using BeSweet, but it would be much better to do all of the job in Wavewizard & its Batch Converter. Is what I’m doing possible using your program, and using Sonic Foundry Softencode as the Output type?

Im sorry the error message is not clear enough.

You have to go to options and check stream manipulation + monostreams. Then you get monostreams. If you want to use sonic, you must always convert the files to monostreams (unless it has only 1 channel already) and the same goes for surcode. This should fix it, but if it does not, post a msg again.

[daphy]

Hi Johnman,

I played a little with ww these days. The possibiltlies are exciting! :)
I also played with the interface to surcode. One tiny bug always accured see screenshot:

http://img291.imageshack.us/img291/2719/wwbug2wr.jpg (http://imageshack.us)

evertime I pressed convert I run into this error, after closing the ConversationBatcher manually and again pressing the convert button everything went fine :rolleyes:
As you mentioned in your last posting - I would suggest to add a routine which sets (if necessary for the encoder) the output automatically to mono files.
Next thing which makes me wonder: I did a DTS-WAV but the file ended up with a DTS extention - file itself was correct, only the the extension has to be renamed.
I guess that is because of the source file has the same name 123.WAV (PCM 6 channel) => 123.DTS (DTSWAV) should be a WAV means two times the same file name in one folder - thats not possible!
What do you thing about a ww prefix or something like: 123(DTSWAV-16-44.1).WAV?
Another question - I played around with some CUE sheet tools these days - is it possible to add some functions of them in ww, too? This is a general question, I guess we could collect some ideas what CUE function would be nice to add here on the board later :)

that´s it folks ;)

[johnman]

Hi Johnman,

I played a little with ww these days. The possibiltlies are exciting! :)
I also played with the interface to surcode. One tiny bug always accured see screenshot:


evertime I pressed convert I run into this error, after closing the ConversationBatcher manually and again pressing the convert button everything went fine :rolleyes:


You dont have to close it, pressing convert again also works. The problem is that ww starts cb and then waits 100 msec to see if it really started. If it didnt, it gives the error. I also noticed 100msec is to short. When windows has to read the file from the disk it often takes more then 100 msec. After starting it once, windows caches it (i think) and then 100msec does work. Ill fix this for the next release.



As you mentioned in your last posting - I would suggest to add a routine which sets (if necessary for the encoder) the output automatically to mono files.


I thought about this, but should ww also set the sample rate correct when its wrong, and the sampletype? I dont think it would be a good idea to do to much automaticaly, since the setting for cb might be wrong itself, and not the preferences.



Next thing which makes me wonder: I did a DTS-WAV but the file ended up with a DTS extention - file itself was correct, only the the extension has to be renamed.
I guess that is because of the source file has the same name 123.WAV (PCM 6 channel) => 123.DTS (DTSWAV) should be a WAV means two times the same file name in one folder - thats not possible!
What do you thing about a ww prefix or something like: 123(DTSWAV-16-44.1).WAV?

I never thought about this :). Cb just adds .dts since it is dts (or dts-wav).
ill change it for the next release. And its not a problem when the file already axcists. It just adds a (1) to the end of the file, just like ww. If everything is correect, ww and cb never overwrites any file.


Another question - I played around with some CUE sheet tools these days - is it possible to add some functions of them in ww, too? This is a general question, I guess we could collect some ideas what CUE function would be nice to add here on the board later :)
that´s it folks ;)

Well i havent ever used it myself, but from what i know about it, its pretty simple. Ill look into this.

Thx for all the suggestions and bug report's daphy. I'll fix them asap.

EDIT: I fixed this stuff, but since i dont regard the mentioned bugs as critical, i wont release a nre ww it immediately.

[tebasuna51]

This post is for Jonhman and all users of Command Line Operation of Sonic Foundry Soft Encode (used for WaveWizard).

This mode uses a Parameters.ini to set all options of the encoder (defined in an Appendix on the Soft Encode help).
All parameters with numerical value runs ok, but all parameters On/Off works on the contrary. For instance:

-UseLFE=Off -> insert the LFE channel in the ac3 (like I say in previous post).
-Deemphasis=Off -> use the filter and high frequencies are attenuated.
-DCfilter=On -> don't use the filter DC.
...
-TimeStamp=Off -> use the frame type SMPTE (16 bytes before the normal header). Played ok, but rejected from GraphEdit and DVD Authoring software. (default in WaveWizard).
-NonIntel=On -> make a file with Intel byte order rejected always (frame type Intel). If present TimeStamp=Off, frame type SMPTE override frame Intel.

The only parameter don't tested is BandWidthFilter (I don't know how), all others are verified.

I use v1.0 (build 19) of Sonic Foundry Soft Encode.

And i can confirm the bug in sonic surcode. I have to set the option to off to enable it :confused: I think i will add an option to cb to invert the flag.

I think the option to invert the flags must be general not only for UseLFE. (Soft Encode, not Surcode).

A final note for Jonhman: when I set -20 in Dialog Normalization (Configuration of Soft Encode) you put DlgNorm=19 in ini file and must be 20.

[johnman]

This post is for Jonhman and all users of Command Line Operation of Sonic Foundry Soft Encode (used for WaveWizard).

This mode uses a Parameters.ini to set all options of the encoder (defined in an Appendix on the Soft Encode help).
All parameters with numerical value runs ok, but all parameters On/Off works on the contrary. For instance:

-UseLFE=Off -> insert the LFE channel in the ac3 (like I say in previous post).
-Deemphasis=Off -> use the filter and high frequencies are attenuated.
-DCfilter=On -> don't use the filter DC.
...
-TimeStamp=Off -> use the frame type SMPTE (16 bytes before the normal header). Played ok, but rejected from GraphEdit and DVD Authoring software. (default in WaveWizard).
-NonIntel=On -> make a file with Intel byte order rejected always (frame type Intel). If present TimeStamp=Off, frame type SMPTE override frame Intel.

The only parameter don't tested is BandWidthFilter (I don't know how), all others are verified.

I use v1.0 (build 19) of Sonic Foundry Soft Encode.

I think the option to invert the flags must be general not only for UseLFE. (Soft Encode, not Surcode).

A final note for Jonhman: when I set -20 in Dialog Normalization (Configuration of Soft Encode) you put DlgNorm=19 in ini file and must be 20.

Thx for the bugreport. Ill release a new version later today, which will fix these problems.

EDIT: te Dialog Normalization is correct if i interpret the manual correctly :

DlgNorm 0 to 31 (-1 dB to -32 dB)

But taken into account all the ON's that should be OFF's and reverse, im not to sure about this either :(.

[daphy]

Hey guys,

what do you think is a useful function for CUE support?

some ideas
- for added (drag´n´dropped) single files a CUE sheet should be created (options -> gap between tracks, Albumname, Interpret, ... )
- for added (drag´n´dropped) playlists (m3u f.e.) a CUE sheet should be created including all infos from the playlist + the possible options mentioned above ...
(meanwhile I found some simular functions in the latest foobar2000 0.9b6+ version)

do you think these functions will be senseful?

or are there other/better ideas? :rolleyes:

[johnman]

Hey guys,

what do you think is a useful function for CUE support?

some ideas
- for added (drag´n´dropped) single files a CUE sheet should be created (options -> gap between tracks, Albumname, Interpret, ... )
- for added (drag´n´dropped) playlists (m3u f.e.) a CUE sheet should be created including all infos from the playlist + the possible options mentioned above ...

do you think these functions will be senseful?

or are there other/better ideas? ::


I sended you a pm about this but you havent answered it :(

I think its a good idea BTW :)

If someone else has any ideas about this, dont hesitate to share it with the rest of us

[daphy]

sorry my fault - I was a little stressed because of business :(

(but maybe collecting ideas on the board won´t be a bad idea, don´t you think? :) ) <- obsolete you´ve edited your posting :D

I played around with CUE Splitter (see needfulthings (http://www.needfulthings.webhop.org)), it offers a nice GUI for most of it´s functions - I don´t know wether that much function are needed in ww, but some won´t be bad.
As I mentioned at PM there are some further information possible to add inside MP3 (f.e. ALBW) or WAV (Audition) - as basis for a CUE file if possible? :rolleyes:

[johnman]

As I mentioned at PM there are some further information possible to add inside MP3 (f.e. ALBW) or WAV (Audition) - as basis for a CUE file if possible? :rolleyes:

Im still in doubt about the functionality of cue files, so you can imagine i dont know anything about the "album format". But from wht you described, its a handy feature, so i might add it someday.

Currently i want to implement cue functionality and vst hosting. Yesterday i spended a couple of hours trying to get a GUI from VI.dll, but the only thing i figured out is that the steinberg (= "inventor" of vst) documentation is really bad.

I "fixed" the bugs from tebasuna51. The problem is that the settings are not really reliable. For now i have added a setting to invert uselfe,ephasis,DC filter and timestamp and i disabled the intel byte order, so sonic will use the default byteorder (is this correct tebasuna51?).

But i still think the supports is a little shaky since i dont got much confidence sonic is using the .ini file correct. I have to look into this, and i possibly will
a) remove support for sonic
b) use something more reliable then the .ini file

For now i did what tebasuna51 suggested

I will upload ww v0.54b to needfull things in a minute

EDIT
i also tried to get bitidentical output from sonic in normal mode, and sonic through cb, and somehow i never get the same output. I tried to change different settings, but i never got identical files. This means i cant compare the settings from sonic+ cb with sonic in normal mode which again means, the settings are not 100% reliable.

[tebasuna51]

The Dialog Normalization is correct if i interpret the manual correctly :
DlgNorm 0 to 31 (-1 dB to -32 dB)

From a52 doc:
"Valid values are 1–31. The value of 0 is reserved. The values of 1 to 31 are interpreted as -1 dB to -31 dB with respect to digital 100 percent. If the reserved value of 0 is received, the decoder shall use –31 dB. The value of dialnorm shall affect the sound reproduction level."

There are other errors in Soft Encode help:
CodeMode 1 to 8 (1/0 to 3/2)
may be:
CodeMode 0 to 7 (Ch1 + Ch2, 1/0 to 3/2)
From a52 doc:
"If acmod is 0, then two completely independent program channels (dual mono) are encoded into the bit stream, and are referenced as Ch1, Ch2."

And:
MixLevel 0 to 32 must be 0 to 31 (5 bits field, max 31)

The settings not captured in your Configuration window, taken the default values (not the actual settings in Soft Encode):
BSMode=0 (Complete Main)
CopyRight=1 (set)
Original=1 (set)
RoomType=2 (Small Room, flat monitor)
MixLevel=105 (dB SPL)


But taken into account all the ON's that should be OFF's and reverse, im not to sure about this either.
I'm sure, after many, many test, but you are free to believe me.
May be others users can test this, and post any confirmation.

Another test using 0/1 instead Off/On don't work, the parameter are ignored and use the default.

[daphy]

uploaded :cool:
John, can plz check the programlist, I think it must be outdated?!? ;)

[johnman]

From a52 doc:
"Valid values are 1–31. The value of 0 is reserved. The values of 1 to 31 are interpreted as -1 dB to -31 dB with respect to digital 100 percent. If the reserved value of 0 is received, the decoder shall use –31 dB. The value of dialnorm shall affect the sound reproduction level."

There are other errors in Soft Encode help:
CodeMode 1 to 8 (1/0 to 3/2)
may be:
CodeMode 0 to 7 (Ch1 + Ch2, 1/0 to 3/2)
From a52 doc:
"If acmod is 0, then two completely independent program channels (dual mono) are encoded into the bit stream, and are referenced as Ch1, Ch2."

And:
MixLevel 0 to 32 must be 0 to 31 (5 bits field, max 31)

The settings not captured in your Configuration window, taken the default values (not the actual settings in Soft Encode):
BSMode=0 (Complete Main)
CopyRight=1 (set)
Original=1 (set)
RoomType=2 (Small Room, flat monitor)
MixLevel=105 (dB SPL)


I'm sure, after many, many test, but you are free to believe me.
May be others users can test this, and post any confirmation.

Another test using 0/1 instead Off/On don't work, the parameter are ignored and use the default.

I believe you, i dont really need confirmation. The problem is that i used the manual of sonic as a reference (i thought that someone who can make an encoder surely can make a proper manual), and as you mentioned, it has a couple off errors. I know i dont support all settings (like copyright) but i thought i supported the most important ones.
Anyway, im thinking about dont using that .ini file anymore and use something else. I cant seem to verify all the settings reliable , and seeing there are already enough issues with it, im hoping i can use an other method to transport the settings into sonic.



uploaded
John, can plz check the programlist, I think it must be outdated?!?


If you mean that thing we talked about earlier (version checking), it slipped my mind :(.

[tebasuna51]

Use the left column for SonicAc3.ini and right column for Sonic in normal mode. The files generated are bitidentical (for me).

[Parameters] [Soft Encode Settings]
DataRate=448 Data Rate = 448
SampleRate=0 Sample Rate = 48 KHz
CodeMode=7 Audio Coding Mode = 3/2 (L,C,R,l,r)
UseLFE=Off LFE enable = Checked
Bit stream mode = ...Complete Main
DlgNorm=20 Dialog Normalization = -20 dB
NonIntel=Off Save fr. in Intel byte order = Unchecked
CMix=0 Center Mix level = -3 dB
SurMix=0 Surround Mix level = -3 dB
DolbySurround=0 Dolby Surround mode = Not indicated
Copyright bit = Checked
Original bit stream = Checked
Info exists = Checked
Mix levels = 105 dB SPL
Room type = Small room, flat monitor
Deemphasis=On Digital deemphasis = Unchecked
DCFilter=Off DC high-pass Filter = Checked
BandWidthFilter=Off BandWidth low-pass Filter = Checked
LFEFilter=Off LFE low-pass Filter = Checked
Phase90=Off 90 Deegre Phase shift = Checked
Cut3dB=On 3 dB attenuation = Unchecked
CompressionType=0 Compression characteristic = None
RFCmpInUse=On RF overmodulation protection = Unchecked
TimeStamp=On Add Time Stamp = Unchecked

Sonic uses the defaults for not supplied parameters, then you need always:
Bit stream mode = ...Complete Main
Copyright bit = Checked
Original bit stream = Checked
Info exists = Checked
Mix levels = 105 dB SPL
Room type = Small room, flat monitor

[johnman]

Ahhh thanks tebasuna51, i see i made a mistake with a couple of defaults. I unchecked them while they should be checked. I have to say your feedback is very helpfull, and detailed. It really helps me to fix the bugs faster :). :thanks:


Im was/am working on a written tentamination (some really heavy stuff), and doing wavewizard devlopment together . Since that doesnt go to well together (it gave me a little headache :eek: ), im now doing only the tentamination. I've got 5 days for it, so until it is finished, i dont have to much time to spare.

When i get some free time again, im going to triple check all settings for surcode to make sure they are ALL correct.

Although everyone still can send bug reports, im afraid i cant fix them immediately.

EDIT
Has anyone checked if the settings in sonic in normal usage (without cb) are correctly applied in the resulting file?
If i would compare the output from sonic + cb with the output from sonic, and the output from sonic itself is incorrect, its still useless.

[tebasuna51]

Testing Wavewizard 0.54 for Soft Encode

[Normal] [Inverting]
UseLFE=On UseLFE=Off Ok
DlgNorm=19 DlgNorm=19 I set 20
RFCmpInUse=Off RFCmpInUse=Off ?
Deemphasis=Off Deemphasis=On Ok
DCFilter=On DCFilter=Off Ok
BandWidthFilter=On BandWidthFilter=On ?
LFEFilter=On LFEFilter=On ?
Phase90=On Phase90=On ?
Cut3dB=Off Cut3dB=Off ?
TimeStamp=Off TimeStamp=On Ok

You believe in me only 40%.
But, always, thanks for your job.

Edit: Before send this post I read your new post, well, you believe in me a little more. But I want the 100%.
You say:
Has anyone checked if the settings in sonic in normal usage (without cb) are correctly applied in the resulting file?
If i would compare the output from sonic + cb with the output from sonic, and the output from sonic itself is incorrect, its still useless.
And me:
"The only parameter don't tested is BandWidthFilter (I don't know how), all others are verified."
This include the effects in file output.
Let me more time for this last test (BandWidthFilter) and post later.

[johnman]

i believe you 100%. If you say its ok... it IS ok.

Im sorry, but i didnt read you post carefully enough. I missed theat part about you verified the other settings. Like i said, got much on my mind, and my tentamination is going kinda slow...

What i would like to know is
a) the bandwith filter
b) you said in an earlier post something about that intel byte order. I didnt fully understand it, and i thought its better to just disable it and use the default setting. But if you can clarify it, i would very much appreciate it :).
c) the other settings, are they correct now, or dont you know if they are correct> You got all those '?' next to them.

And thx for the effort in trying to get sonic work reliable

[tebasuna51]

a)Test for BandWidthFilter
Using a test wav 6 chan 48000 Samples/sec. (24 seconds tone test from 0 Hz to 24 KHz)

Encoded at 448 kbps (Nominal BandWidth = 20.3 KHz)
BandWidthFilter = On -> Progressive cut from 19.4 to 19.8 KHz
BandWidthFilter = Off -> Irregular cut from 20.3 to 20.4 KHz

Encoded at 320 kbps (Nominal BandWidth = 15.8 KHz)
BandWidthFilter = On -> Progressive cut from 15.0 to 15.5 KHz
BandWidthFilter = Off -> Irregular cut from 15.8 to 15.9 KHz

Then the switch work, another question is recommend it.
Before the test I think not see so clearly the differences.

b) Just like other switch you must invert On -> Off
This switch is ignored when you check TimeStamp. In Soft Encode Help:
"Time stamped frames are automatically not stored in Intel byte order."

Soft Encode uses 3 frame types:
-Non-Intel, First byte 0x0B -> run Ok in all soft. Default or with:
Add Time Stamp = Unchecked
Save frame in Intel byte order = Unchecked

-SMPTE, First byte 0x01, byte 17 0x0B (rest of frame like Non-Intel) -> Are played ok (ignoring the first 16 bytes), rejected from GraphEdit and DVD Authoring software. Activated with:
Add Time Stamp = Checked

-Intel, First byte 0x77 -> Only know the proper Soft Encode to open it. Activated with:
Save frame in Intel byte order = Checked
Add Time Stamp = Unchecked

I don't know the use of Time Stamp or Intel byte order.

c) Switches with ? aren't inverted by the switch in cb Configure, then don't work properly.
All switches must be inverted. And the Dialog Norm must be the same value than selected, now is value - 1.

EDIT: (ignoring the first 16 bit) must be (ignoring the first 16 bytes),

[johnman]

Thx tebasuna, this is some info i can use immediately in ww. The next release will have correct support for sonic softencode thanks to you :D.

[johnman]

Im trying to let ww also decode ac3 and dts so anyone can transcode them to any other format in only a single step. Since the code of tranzcode is freely available, im thinking about changing it a little so it can work together with ww, but since this is not the case with azid, im thinking about using the azid.exe from the azid1.9 package. The problem is that i havent used it very much, and i remember from a long time ago the azid.dll is sometimes better then the azid.exe (or was it the other way arround?). Does anybody know if there really is a difference between azid.dll vs azid.exe? Or should i maybe use an other ac3 decoder? Any comment/info is welcome :)

[jm duchenne]

Hi Johnman,

First, thank you for your work. It is really a must have tool for everyone who does multichannel sound.

I've read in the previous pages that you are working on a VST host feature ???

I make multichannel plugins (http://acousmodules.free.fr) and I wonder how it could be wonderful to use them inside Wavewizard !

Have you some beta version to test ?

[johnman]

Hi Johnman,

First, thank you for your work. It is really a must have tool for everyone who does multichannel sound.

I've read in the previous pages that you are working on a VST host feature ???

I make multichannel plugins (http://acousmodules.free.fr) and I wonder how it could be wonderful to use them inside Wavewizard !

Have you some beta version to test ?

I havent got a beta with vst hosting unfortunately, but it would be indeed a wonderfull idea to use your stuff it in ww. I looked at your site and its really interesting.
Have you some indepth knowledge about making vst's ? Im having a little problem with the implementation (of the host), and i could use some tips/pointers sometimes.

(@ all Although im also busy with my study (currently i have to make a sort of a simple graphic card driver) ww is still being developed, but a little slower then before.)

[jm duchenne]

Have you some indepth knowledge about making vst's ?
I'm sorry, but no. :(

I use Synthedit or other graphical environments, so I have not to deal with the basic codes...

But perhaps could you see with Herman Seib : http://www.hermannseib.com/english/vsthost.htm
or Tobias Fleischer (Toby Bear) : http://www.tobybear.de/developers.html
They have free hosts and I'm sure that they can share some resources or tips.

Perhaps could you send me an email (sonart@free.fr) when you will have made some progress with the VST host implementation ?
Thanks !

[johnman]

I'm sorry, but no. :(

I use Synthedit or other graphical environments, so I have not to deal with the basic codes...

But perhaps could you see with Herman Seib : http://www.hermannseib.com/english/vsthost.htm
or Tobias Fleischer (Toby Bear) : http://www.tobybear.de/developers.html
They have free hosts and I'm sure that they can share some resources or tips.

Perhaps could you send me an email (sonart@free.fr) when you will have made some progress with the VST host implementation ?
Thanks !

I already tried to get some info on vst on a number of ways, but so far wasnt very succesfull. I just have to find out everything myself :(.

And when i got a working beta host ill send you a mail. Im developing the host and ww seperate for now, so the first beta will probabely be only a hosting program. Later ill merge them together.

[johnman]

Im experimenting a little with recoring from the soundcard and i noticed that somehow all regular samplerates can be used to record from the soundcard. Even my onboard soundcard can record at 192khz :confused: . Does this mean my soundcard can really support this samplerate natively, or is windows resampling. How can i find out what the real hardware suported samplerates are? (BTW cooledit also shows my onboard soundcard is capable of recording @ 192khz)

And here is a little screenshot of ww recording in simplemode if anyone is interested:

screenie (http://www.rarewares.org/wavewiz/recording.png)

[daphy]

And here is a little screenshot of ww recording in simplemode if anyone is interested
Does the enhanced mode support multichannel recordings?
-> There meanwhile lots of soundcards on the market which support multichannel input, so let´s feed them :D :D :D

[johnman]

Does the enhanced mode support multichannel recordings?
-> There meanwhile lots of soundcards on the market which support multichannel input, so let´s feed them :D :D :D

It was already on my mind, but i think i need asio stuff for this. Normal windows recording only supports up to 2 channels (i think). I dont know exactly how hard/easy asio is to support, but if it is as poorly documented as the VST-interface (both are from steinberg if im not mistaken) it will take while :(.

My soundcard has an option for 6 channel recording (and i already captured 6 channel audio from powerdvd (digitally) using krystal audio not to long ago) and i intend to support this in ww :). Maybe i eventually create some sort of "virtual soundcard", so the output from ANY program can be recorded digitally using ww.

But how and when the support comes, im not sure. Im trying to gather some good info on recording, but this isnt easy. The more support i get, the faster the development will go :D .

[daphy]

you already know this (http://www.asio4all.com/) one?

[johnman]

Unfortunately i still can not release a new version, but to the ppl that are interested in ww: Im pretty busy with other stuff, but i still i work on ww from time to time. Currently I've finished another rewrite on ww. I've totally changed the internal structure of ww. Now i just create a "sample-pipeline" @ runtime by simply connecting together several functions. What this means is that its possible to do i.e.:

upsample -> dither -> downsample -> vst -> downsample -> normalize -> amplify->channelmap->vst-> etc (not really usefull, but still possible)

Previously the "sample-pipeline" was very static, and to support unlimited vst's was kind of difficult, so thats why the rewrite.

Currently ive not made a GUI to configure it, so configuring is currently hardcoded in. But if i would make a GUI, ww will get similair functionality as bidule :).

This internal stuff will only be visible in the next release by the number of vst's which can be linked together (virtual unlimited), but the change is really much bigger then that. Any function or vst can be seperately configure and used as often as needed.

Thats all for now :).....

[ursamtl]

john this sounds really promising. Take your time and I'm sure ww will be great once you get this all together.

[johnman]

Perhaps could you send me an email (sonart@free.fr) when you will have made some progress with the VST host implementation ?
Thanks !

I sended you an email, but i still havent got any responce. Could you please check your mailbox to see if they arent blocked?

[johnman]

Here is a screenshot of ww hosting a vst. NB notice the option to save as many configurations as you want for every vst. In the VI thread an idea was mentioned to add saving options to VI. When ww is used as a host, the settings can already be saved (and exchanged if needed).

screenshot (http://www.rarewares.org/wavewiz/vsthost.png)

[daphy]

Hi Johnman,

looks promising :D I´m looking forward to the next release!

[ursamtl]

Hi Johnman,

looks promising :D I´m looking forward to the next release!

Me too! John, tell me, would it be possible to add realtime file playback to wavewizard or some sort of preview function? It would be useful for making adjustments. For example, I have a 5.1 soundcard/speaker setup, so when I'm using V.I in Bidule, I'm able to immediately hear the results of any slider adjustments I make. Without monitoring capability, it's very much a "hit and miss" or trial and error process. Ideally, one could preview what the output file would sound like, and then switch to something equivalent to Bidule's "offline" mode to do the actualy processing as fast as the CPU allows.

Regardless of the monitoring issue, this latest screen grab looks very good!

Regards,
Steve.

[johnman]

Multichannel playback (and recording) will be supported :). This is indeed a must have for ww if vst's are used. But i wont promise its already available in the next release .

johny

[ursamtl]

Great news! No need for promises. You're already doing some great work!

[daphy]

Hiho Johnman,

I tried something simple with conversation batcher simply: want to convert an existant WAV to MP3 with out any changes on the source file. Hmm -> I have to let wavewizard (at least) copy the file to another place, is it possible to feed cb directly or did I missed something http://smilies.neo101.nl/smilies/2160.gif (http://www.maxthon.com/)

[johnman]

sry daphy, this cant be done...

when you feed wavs to ww and ww accepts it you know you will get a vallid mp3. Its a bit more reliable, but also more time consuming...

but i agree its not pretty, but what i actually want to do is to integrate cb in ww and just select output format in ww. Dont know how, but that would be a big improvent i think

[daphy]

-obsolete-

[johnman]

ty daphy :)