I wish for...Audio Cues

Okay so I'm driving into work this morning with the top down on my car, the Empeg blasting, and grew tired of
the playlist I I was listening to. With the top down on the convertible I can't read the Empeg screen at all. I can't
pull off the freeway everytime I want to squint at the Empeg and run down through the menus to change playlists.
So I was stabbing at the Empeg blindly, trying to find a new playlist that fit my mood. Bah!!

Initially I was wishing for a way to just jump to a different playlist, I could find one that I wanted that way
but then a better interface occured to me:

The menus could all include audio equivalents. So as I step though the menus the text is said over the
speakers. I imagine that each menu could have a tiny mp3 associated with it. I'm thinking an English woman's
lightly accented voice:

Volume...Fader...Next Visualization...Playlists......Self-Destruct Sequence Initiated....

Of course that let's you step through the standard menus but how to step through what may turn out to be
hundreds of playlists? Have Emplode generate a tiny mp3 attached to each playlist when it is defined. By
default it's a computer voice saying the playlist title or optionally the user overides with a relevant music sample.
Then you could step though your Playlists:

Classic Rock...Alternative...Electronic...Dance...

One open-source text to speech project is available here.

yours, Dean

Voice prompts, including a self-recorded audio tag for your own presets (and all the "factory" voice prompts being in a replaceable/updateable playlist, so you can rerecord the prompts yourself) are on the to-do list, you'll be glad to know :)

Hugo

> Voice prompts, including a self-recorded audio tag for your own presets (and all the "factory"
> voice prompts being in a replaceable/updateable playlist, so you can rerecord the prompts
> yourself) are on the to-do list, you'll be glad to know :)

Very sweet!! I can't wait. =)

Off subject but I'm sitting here happily shoving muisc at my Empeg through the new Linux interface.
I'm happy since the few win98 (and win2000) platforms I had access to dried up a few weeks ago and
I still have a LOT of space to fill up.

yours, Dean

Festival is a bit big, instead I downloaded rsynth and cross compiled it.
Unfortunately, I can't get it to work on the empeg, though it works on my redhat
box. When I strace it it all appears to be functioning; it does some ioctls on
/dev/dsp and then writes data to it, but no sound. I also compiled aumix
and checked that the volume hadn't been turned down, but no luck. Anyone
got any idea? Does /dev/dsp have to be written in 4608 byte chunks as well as
/dev/audio? I modified rsynth to do that but it didn't make any difference.

Some other useful links to links here

Paul

You have to unmute the audio - Mike has a program to do the correct ioctl, I'll prod him tomorrow :)

Hugo

You have to unmute the audio - Mike has a program to do the correct ioctl, I'll prod him tomorrow :)

Well, he hasn't prodded me yet but here's some untested code that might prove useful.



--
Mike Crowe
I may not be speaking on behalf of empeg above :-)

> here's some untested code that might prove useful.

Thanks, but no luck. It said it was already unmuted and rsynth's "say" didn't
say anything.

Does the following indicate anything obvious to you?:

./strace -fetrace=stat,open,write,ioctl ./say hello
open("/etc/ld.so.preload", O_RDONLY) = -1 ENOENT (No such file or directory)
open("/etc/ld.so.cache", O_RDONLY) = 3
open("/lib/libm.so.6", O_RDONLY) = 3
open("/lib/libc.so.6", O_RDONLY) = 3
open("/dev/dsp", O_WRONLY|O_NONBLOCK) = 3
ioctl(3, SNDCTL_DSP_SPEED, 0x201924c) = 0
ioctl(1, TCGETS, {0x1002 /* B??? */ opost isig icanon echo ...}) = 0
write(1, "Actual sound rate: 8000\n", 24Actual sound rate: 8000
) = 24
write(3, "\200\200\200\201\201\200\200\200\200\201\200\200\200\201"..., 4608) = 4608
ioctl(3, SNDCTL_DSP_SYNC, 0xbffffdd4) = 0

A couple of things spring to mind:

a) We don't support 8000Hz, only 44.1khz. You can expand the sound data yourself though - and we do intend to offer other rates at some point.
b) The data looks vaguely 8 bit and probably mono. We offer 16 bit stereo only. Again, you could post-process the output buffer yourself.

hugo

That's great Hugo! So that means that i can make my own sounds for each function, and playlist, or use default ones? That's neet because you wont have to look down at the playlists anymore wile driving.


-CHiP

Can rsynth and the player co-exist on the sound hardware? Assuming I
build rsynth (and sox to get the sample rate&size right), can I run it
while the player is going? Or can only one program hit the sound devices
at a time?

(I'd try it out myself, but I left my freakin' serial cable at work)

brian

> Can rsynth and the player co-exist on the sound hardware? Assuming I
> build rsynth (and sox to get the sample rate&size right), can I run it
> while the player is going? Or can only one program hit the sound devices
> at a time?

Right, I've tried it. It seems only one device at a time. I think I'll put
esd on it. Any chance empeg can put esd (enlightenment sound daemon) support
into the player? :-)

Here's what I did to get rsynth to work:

Cross compiled rsynth and sox on linux, transferred to empeg.
./say hello there paul -l hello # Puts output in file hello
./sox -s -w -r 8000 -t raw hello -r 44100 -c 2 baz.au # convert to empeg fmt
dd if=baz.au of=/dev/dsp bs=4608 # send to dsp in 4608 byte chunks

Paul

you guys at empeg are animals. now i only i wish i could afford one. i'd be listening to it right now if i could have. o well. maybe next time i'm contacted. any idea when mid 1000's defers will have the chance to reorder?


oh, i wish i wish i had an empeg...