From: Bluechip <csbluechip_at_gmail.com>
Date: Mon, 03 Apr 2006 09:35:42 +0100

Of course, as far as LAME is concerned, adding a "-q 0" will increase
the quality.

But in fairness, I doubt that is what you seek.

Any radio ham or CB freak will tell you that you can improve
PERCEIVED quality (that is psychoacoustic trickery) by adding reverb
or echo to the sample. (Before compression, ?obviously?)

BC

At 02:41 03/04/2006, you wrote:
>ScottLearned wrote:
>>Hello,
>>I just want to say thanks to all the Dev's out there who worked so
>>hard making Rockbox accessible for Blind/Visually impaired users. I
>>am really thrilled to have speech on such a small han-held player!
>>I'm not sure how much you guys realize how this profoundly affects
>>us users, but thank you again. Well anyways, sorry for the rambling.
>>My question is, can I improve the quality of my voice clips? I
>>changed the settings in Lame and checked out the scripts, but can't
>>for the life of me find where to create a higher quality .talk file.
>>Thanks for any help and keep up the excellent work!
>>Scccott
>
>The encoding options used by the voiceBox scripts are set in the
>variable LameOpts in the voiceBox.wsf file, like so:
>
> Const LameOpts = "-m m -b 24 -S"
>
>I'm not really sure what would provide the best *quality* voice
>clip. At one time we had some constraints on the clips that limited
>our choices, IIRC. I'm not sure if that is true any longer. I'd have
>to dig around in the lame docs and experiment a bit first to see
>what might work better. If anyone out there has used some different
>settings for Lame (or the SAPI5 voices) and gotten better results,
>please post them here. I'd be happy to make them the new default
>options in the voiceBox package.
Received on 2006-04-03