From alsa-devel-owner@alsa.jcu.cz  Thu Jan 21 16:59:10 1999
Received: from nic.funet.fi (nic.funet.fi [128.214.248.6])
	by marvin.jcu.cz (8.9.1a/8.9.1) with ESMTP id QAA05531
	for <alsa-devel@alsa.jcu.cz>; Thu, 21 Jan 1999 16:58:41 +0100
Received: from localhost (user: 'kouhia', uid#241) by nic.funet.fi id <6927-2279>; Thu, 21 Jan 1999 17:58:11 +0200
From: Juhana Sadeharju <kouhia@nic.funet.fi>
To: alsa-devel@alsa.jcu.cz
In-reply-to: <01BE44CC.41B12C60@MUSE> (richard@muse.demon.co.uk)
Subject: Re: hi-res pcm sound support/IEEE floats
Message-Id: <19990121155814Z6927-2279+2926@nic.funet.fi>
Date: 	Thu, 21 Jan 1999 17:58:11 +0200
Reply-To: alsa-devel@alsa.jcu.cz
Sender: alsa-devel-owner@alsa.jcu.cz
Precedence: list

>From:	"Richard W.E. Furse" <richard@muse.demon.co.uk>
>
>Hmmm. I'm going to ramble a bit and then argue to suggest that we should do 
>8bit audio, 16bit audio and then go straight to IEEE 32bit floats.

>ARGUMENT 1 (signal/noise, limits of human hearing, logarithmic/linear 
>audio)
>
>A) Much as I hate to throw a spanner in the works, I think the current 
>discussion rather pointless and flawed. There is NO POINT in more than 
>24bit linear audio on playback.

Where that discussion was done? Can you quote an article which suggest
to use more than 24-bit in playback or recording?


>playing a square wave at 1bit is just inaudible, i.e. "0dB" as the 
>environmental sound people record it. Then play another square wave at 
>24bits over the same amp. This will give you about 140dB. This is into the
[ ... ] 
>By this same scale, 16bit audio gives a lot of dynamic range. There's an 

The distortion to audio at 1-bit level is major and nobody will use it at
high volume levels. 24-bit is enough, in my opinion, but with 16-bit you
cannot reach the volume levels high enough if you want less distortion to
quiet places.

You're putting it wrong way: first make a loud sound at level 140 dB (rock
concert level), then make a quiet sound and adjust its distortion by
increasing the bits.

Lets face the fact that CD makers are using noiseshaper ditherers with
16-bit audio. It gives 19-bit dynamics and sounds much better than 16-bit
alone. Is 19-bit enough to give enough dynamics? If yes, then why not
move to use 20-bit converters instead of using 16-bit with lotsa additional
noise (caused by ditherer)?


>taking you from perceived silence to hearing damage.

This kind of arguments make you sound less professional.
4-bit mantissa and 8-bit exponent is also "taking you from perceived
silence to hearing damage" but that has nothing to with quality.


>using words that small, IEEE floats have seemed to be the standard. I 
>gather there has been a little more argument about double-precision floats, 
>but not much.

Some dsp stuff require you to use more than 24-bit precision, i.e., float
is not enough. Some operations are not possible without enough precision,
i.e., routine will fail badly with float.


>16bit audio should remain the staple diet of Linux because most audio is 
>playback audio and 16bits on a good amp gives 0dB to 96dB as silence to 
>neighbour-disturbing loud. 96dB/16bits is good for playback.

Are you working for Cakewalk or are you trying to freeze Linux professional
audio projects?

>but playback making full use of 24bits means you just damaged your 
>audience's hearing. Do you want that on your conscience?

Nonsense. The extra bits can be used at the low volume end, not at the
high volume end only as you seem to suggest!

ALSA drivers should support whatever audiocards supports. IMHO, conversions
from audiocard formats to float or such should be done outside the drivers.
If card gives out X-format, then a programmers should have a possibility
to get that format to his program directly.

There are software such as PSL which can make the conversions, channel
splits etc. if wanted.

Yours,

Juhana

