Vg 4 Dialogic Sound Format Quality And...

[ktruk]

Whats going on with Vg4Dialogic sound files?

 

I was under the impression that the "Dialogic" version would play the files at better quality (even if only slightly better) than the TAPI Wave driver based system.

 

First, I ran the Credit-Card script without converting the wavs down to 8KHz, and I got a slow drawling mans tone. This is the result of a 11Khz file being played by VG at 8Khz.

 

Then I ran CoolEdit and converted the WAV to 8Khz/8Bit/Mono/PCM. Now they play okay, but the quality seems worse than the TAPI 11khz/8Bit/Mono/PCM format. I used the "Edit/Convert Sample Type" function and get a mushy voice copy of the original file.

 

So...

1. what is the best way to get the best sound played by Vg4Dialogic? Start with a high-sample rate (like 22KHz/44KHz) first? Or record the original files at the target spec (ie: at 8KHz)?

 

2. Is there a better way to do the conversion?

 

3. Does setting the resolution improve things (or spoil thisngs), ie: setting 16bit vs 8 bit.

 

4. When you state (elsewhere on the forum postings) that a slight improvement is available, just how good should it be? no better/no worse (virtually the same) or barely noticeable or obvious improvement?

 

5. If there is an improvement, why is a 8Khz file better than an 11KHz file?

 

6. Also, I've noticed that I get all my script WAVs copied over to VOX files, at half the size! These copies are created automatically whenever VG plays them. So, is this VOX conversion built-in? It does not seem to need both, because the script designer can't play VOX and if I delete the WAV, then VG complains that the WAV is missing, so VG is NOT even using the VOX??? Some of my scripts use a lot of WAVs and I can really do without this duplication!

 

Please help explain the above points, this info is needed desparately. Thx - Tim.

[SupportTeam]

Then I ran CoolEdit and converted the WAV to 8Khz/8Bit/Mono/PCM. Now they play okay, but the quality seems worse than the TAPI 11khz/8Bit/Mono/PCM format. I used the "Edit/Convert Sample Type" function and get a mushy voice copy of the original file.

Do not convert 11kHz to 8kHz - you will get a poor 8kHz file.

As per instructions in "VG for Dialogic" - when installing VG chose the "TAPI" options so that VG installs 8kHz system sound files.

1. what is the best way to get the best sound played by Vg4Dialogic? Start with a high-sample rate (like 22KHz/44KHz) first? Or record the original files at the target spec (ie: at 8KHz)?

For best quality go to a studio - they can record far better then anything you will achieve with a PC and microphone. Ask studio to supply in 8kHz PCM (or 8kHz VOX) and you will not need to do any conversion yourself.

 

2. Is there a better way to do the conversion?

From 11kHz conversion is tricky - and as I understand the approach would even depend on what type of voice is recorded. The only advice is to experiment with different Audio software settings and filters. This is all best left for the studio or audio experts to do.

 

3. Does setting the resolution improve things (or spoil things), ie: setting 16bit vs 8 bit.

Makes no difference from VGs/Dialogic's point of view. If 16 bit is used then VG/Dialogic would just not look at the 8 extra bits.

 

4. When you state (elsewhere on the forum postings) that a slight improvement is available, just how good should it be? no better/no worse (virtually the same) or barely noticeable or obvious improvement?

I'd say barely noticable.

 

5. If there is an improvement, why is a 8Khz file better than an 11KHz file?

With 11KHz files Dialogic drivers do extra on-the-fly conversion of the 11kHz files. The on-the fly conversion introduces some noise into the played sound.

 

6. Also, I've noticed that I get all my script WAVs copied over to VOX files, at half the size

VG converts WAV to VOX. VOX is what the Dialogic plays (without any further conversions). You can specify VOX sound files as the files to be played in VG for Dialogic, then .WAV files are not even needed.

 

When recording Dialogic supplies a VOX file. VG for Dialogic then converts it to 8kHz WAV. There are no distortions in VG's WAV<->VOX conversion - you can use good sound editor software to confirm that.

[ktruk]

Okay, couple of points...

 

As per instructions in "VG for Dialogic" - when installing VG chose the "TAPI" options so that VG installs 8kHz system sound files.

I am not using VG voice prompts in my apps, so this is irrelevent. Besides, I thought TAPI was 11Khz.

With 11KHz files Dialogic drivers do extra on-the-fly conversion of the 11kHz files. The on-the fly conversion introduces some noise into the played sound.

So why the conversion? Why can't we just prepare "native" format files?

VG converts WAV to VOX. VOX is what the Dialogic plays (without any further conversions). You can specify VOX sound files as the files to be played in VG for Dialogic, then .WAV files are not even needed.

Okay, so if I specify WAV in the script, VG generates a VOX if not already available and plays that instead of the WAV. If I delete the WAV file, but keep the script as WAV then VG does not play the VOX but complains no WAV. Change the script to VOX in the script and then VG will play the VOX, but the script editor won't. So... Can it be made to play the VOX somehow? (assuming I have a VOX player on my dev PC).

When recording Dialogic supplies a VOX file. VG for Dialogic then converts it to 8kHz WAV. There are no distortions in VG's WAV<->VOX conversion - you can use good sound editor software to confirm that

Okay, VOX<>WAV do look (and sound equivalent) but I intend to record 100s of files (may be 1000s) a day, and really don't want this file duplication process. Can it be avoided?

 

I assume the problem lies with your new OCX or DLL files for Vg4Dialogic, that is behind this frustrating change to sound files? I had hoped (and have been led to believe) that there would be an improvement to sound quality with Vg4Dialogic as you were no longer frustrated with the limits of the TAPI environment.

 

I just wish that we had a high-quality format that sounded as good as other systems do. I do not want to be fobbed off with over-simplified statements like "telephones can only do 3khz and 8 bit volume" - If this statement were true, we would never have got beyond 2400 baud modems several years ago. I just want better quality, professional prompts that make the best use if the bandwidth available. And I don't want this tiresome file duplication business. Is there anything I can do to avoid this? This potentially is a disaster waiting to happen for me.

[SupportTeam]

So why the conversion? Why can't we just prepare "native" format files?

In VG for Dialogic you can - use VOX 8kHz files.

Can it be made to play the VOX somehow? (assuming I have a VOX player on my dev PC).

Change VG.INI's RecEditor entry from SNDREC32.EXE to the sound editor of your choice.

Okay, VOX<>WAV do look (and sound equivalent) but I intend to record 100s of files (may be 1000s) a day, and really don't want this file duplication process. Can it be avoided?

Just recoed in .VOX - but conversion in VG is done transparently, so you may as well record in .WAV...

 

.VOX files are what Dialogic card's sound encoder/decoder actually uses to play sound files - this is what all other Dialogic based systems have to feed to the Dialogic card in the end. Note however that the 8kHz .WAV files will get converted with no quality loss at all, so you may as well use them...

[ktruk]

Okay...

 

Change VG.INI's RecEditor entry from SNDREC32.EXE to the sound editor of your choice.

 

Tried CoolEdit, but didn't work - do you know of anything that does? (Audacity?) - If I find one first, I will let you know...

 

It would be helpful if you could outline the command-line params for the SNDREC32 app if you have them handy? Is it something like "..\SNDREC32 /play /close filename.wav"?

 

Just recoed in .VOX - but conversion in VG is done transparently, so you may as well record in .WAV...

 

Is that "record" or "recode"? Anyway, how do I do this? do I just change the extention in recording module to ".VOX"? - As I said I DO NOT want duplicated wavs and voxs, but one or the other. So by recording .VOXs I won't get WAVs ? - I will try this right now...

 

However, if I change the ext will this upset the Voice-Mail system - I want recordings to be accessible via customised VM scripts and the Mail-box "player". Any thoughts or issues arising from this?

[SupportTeam]

This will bring up Audacity:

 

[PlayRecordConfig]

RecEditor="C:\Program Files\Audacity\audacity.exe"

 

Is that "record" or "recode"?

record.

So by recording .VOXs I won't get WAVs ?

Correct.

do I just change the extention in recording module to ".VOX"

Yes.

However, if I change the ext will this upset the Voice-Mail system - I want recordings to be accessible via customised VM scripts and the Mail-box "player". Any thoughts or issues arising from this?

That's why you should just stick with using the recommended format (8kHz WAV) instead of VOX.