Re: [linux-audio-dev] LADSPA 64bit FP support ?

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: Re: [linux-audio-dev] LADSPA 64bit FP support ?
From: James McCartney (asynth_AT_io.com)
Date: Wed Mar 22 2000 - 09:46:27 EET


on 3/22/00 12:36 AM, Paul Barton-Davis at pbd_AT_Op.Net wrote:

>> (Doubling and halving have no effect on the mantissa at all...?)

> ffoo = (float) PEAK_VAL;
> ffoo *= 2.0;
> ffoo /= 2.0;
>
> if (ffoo != (float) PEAK_VAL) {
> fprintf (stderr, "Oh well, thats just your FPU.\n");
> }

Does this fail on the PC ? Doesn't on the Mac.
Multiplies by powers of two should not change the mantissa at all.

>> I still think that the only reason we have a 24 bit integer is to overcome
>> the limitations in SNR of integer signals at low amplitude.
>
> Interesting perspective. I've never seen this voiced in any of the
> trade rags - they all seem to want the larger dynamic range that 24
> bits offers, because of the difficulty of keeping a widely varying
> signal from clipping at the same time as keeping it "hot".

You have a wider dynamic range than you can hear with 16 bits, only the SNR
sucks. Thus more bits are added.

> Who can hear sample accurate response to MIDI controller data ?

It is not for midi controllers that there is sample accuracy, it is for
starting voices. If you have some synth patch that you are using for grains
or simply offsetting the timing then you can hear the flanging between
voices. If you only can start at buffer boundaries then you lose the ability
to cause gradated spectral effect by timing of voices. You CAN hear this
effect very easily.

>
> Just as SC puts a premium on real-time, sample accurate performance, I
> think that software that does HDR should be able to do what dedicated
> digital recorders do. If you were about to buy or download a software
> HDR system, and then heard that unlike an ADAT or DAT recorder, it

On a straight through FP drops no bits just like integer format.
FP will lose bits if you add multiple full scale signals just like integer
format.
FP represents quieter signals better that integer format.

--- james mccartney james_AT_audiosynth.com <http://www.audiosynth.com>
SuperCollider - a real time synthesis programming language for the PowerMac.
<ftp://www.audiosynth.com/pub/updates/SC2.2.7.sea.hqx>


New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Wed Mar 22 2000 - 10:15:02 EET