Post by thread:[EE] Audio from serial port.

> I promised somebody I'd post here when I was finished with my "audio > over RS-232" project: > > http://vivara.net.nyud.net:8090/blog/?p=24 > > Looks like hackaday.com has picked it up. But they got my name wrong. :( > > Regards, > Mark > markrages@gmail Pretty clever! I especially like the user comment: "I think, for your next trick, you should hook up a modem and send the audio over a phone line." Harold -- FCC Rules Updated Daily at http://www.hallikainen.com

On Tue, 25 Oct 2005, Mark Rages wrote: > I promised somebody I'd post here when I was finished with my "audio > over RS-232" project: > > http://vivara.net.nyud.net:8090/blog/?p=24 > > Looks like hackaday.com has picked it up. But they got my name wrong. :( This is not fair. They should wait until *i* get around to write *mine* up. I made such a thing months ago and I had proper log compression too, as well as oversampling and resampling to kill the whine (I was playing back 8kHz au at 115kBauds), and shuffling equivalent-weight patterns to kill the whine some more, AND noise mix-in when kerning to avoid pops. I used lookup-table PCM not ADPCM, at an equivalent resolution of 4 bits (3 bits per character, 2 characters per symbol out). I still have it. His photos look impressive but ime you can't hear a thing with a 4 ohm speaker driven with 20mA. Even headphones sound very quiet. The best I could muster was a telephone piezo speaker. Peter

On 10/25/05, Peter <spam_OUTplpTakeThisOuTactcom.co.il> wrote: {Quote hidden}> > > On Tue, 25 Oct 2005, Mark Rages wrote: > > I promised somebody I'd post here when I was finished with my "audio > > over RS-232" project: > > > > http://vivara.net.nyud.net:8090/blog/?p=24 > > > > Looks like hackaday.com has picked it up. But they got my name wrong. :( > > This is not fair. They should wait until *i* get around to write *mine* > up. I made such a thing months ago and I had proper log compression too, > as well as oversampling and resampling to kill the whine (I was playing > back 8kHz au at 115kBauds), and shuffling equivalent-weight patterns to > kill the whine some more, AND noise mix-in when kerning to avoid pops. I > used lookup-table PCM not ADPCM, at an equivalent resolution of 4 bits > (3 bits per character, 2 characters per symbol out). I still have it. > His photos look impressive but ime you can't hear a thing with a 4 ohm > speaker driven with 20mA. Even headphones sound very quiet. The best I > could muster was a telephone piezo speaker. > Comments: 1. I am doing resampling via libsamplerate. 2. Whine is killed by moving start and stop bits inside the feedback loop. 3. Sigma-Delta != ADPCM. Sigma-Delta is not a differential coding. It is more comparable to PWM, with the noise shaped up into higher frequencies when possible. 4. Audio is quiet but audible into 4 ohm speaker. Painfully loud into headphones. I suppose this depends on the electrical specs of the serial port. Still, I would expect my laptop to have less output than a desktop with real +/- 12V supplies. 5. Post your code and I'll link it from my page. Regards, Mark markrages@gmail -- You think that it is a secret, but it never has been one. - fortune cookie

The link no longer works: "Error establishing a database connection" :-( ----- Original Message ----- From: "Mark Rages" <.....markragesKILLspam@spam@gmail.com> To: "Microcontroller discussion list - Public." <piclistKILLspammit.edu> Sent: Tuesday, October 25, 2005 12:42 PM Subject: Re: [EE] Audio from serial port. {Quote hidden}> On 10/25/05, Peter <.....plpKILLspam.....actcom.co.il> wrote: >> >> >> On Tue, 25 Oct 2005, Mark Rages wrote: >> > I promised somebody I'd post here when I was finished with my "audio >> > over RS-232" project: >> > >> > http://vivara.net.nyud.net:8090/blog/?p=24 >> > >> > Looks like hackaday.com has picked it up. But they got my name wrong. >> > :( >> >> This is not fair. They should wait until *i* get around to write *mine* >> up. I made such a thing months ago and I had proper log compression too, >> as well as oversampling and resampling to kill the whine (I was playing >> back 8kHz au at 115kBauds), and shuffling equivalent-weight patterns to >> kill the whine some more, AND noise mix-in when kerning to avoid pops. I >> used lookup-table PCM not ADPCM, at an equivalent resolution of 4 bits >> (3 bits per character, 2 characters per symbol out). I still have it. >> His photos look impressive but ime you can't hear a thing with a 4 ohm >> speaker driven with 20mA. Even headphones sound very quiet. The best I >> could muster was a telephone piezo speaker. >> > > Comments: > 1. I am doing resampling via libsamplerate. > 2. Whine is killed by moving start and stop bits inside the feedback loop. > 3. Sigma-Delta != ADPCM. Sigma-Delta is not a differential coding. > It is more comparable to PWM, with the noise shaped up into higher > frequencies when possible. > 4. Audio is quiet but audible into 4 ohm speaker. Painfully loud into > headphones. I suppose this depends on the electrical specs of the > serial port. Still, I would expect my laptop to have less output than > a desktop with real +/- 12V supplies. > 5. Post your code and I'll link it from my page. > > Regards, > Mark > markrages@gmail > > -- > You think that it is a secret, but it never has been one. > - fortune cookie > > --