[Fluxus] Noisy output from (gh)

[Glauber Alex Dias Prado]

Chris Targett <xin at xlevus.net> writes:

> Humm,
>
> Even after
> * Blitzing all of Jack's preferences
try increasing not only jack's buffer size(period) but also nperiod (3
or higher)
> * Using a buffer size in Fluxus of 2048
Buffer size of fluxus has to be the same as jack.
> * Using the 'dummy' JACK driver to bypass any dodgy audio devices
dont use this, try alsa and make sure it is configured for your card.

You can bypass jack with reading from a wav file directly as dave mentioned, with:
(process "somemusic.wav")

>
> I'm still getting the same result, everything is jumping around like crazy.
> It's as if Fluxus was getting a fed bunch of zeros and then a big
> surge of non-zeros to make up for it.
>
> Is there anything else I can try?
>
> Ta
> Chris Targett
>
>
>
> On Tue, Dec 22, 2009 at 2:04 AM, Dave Griffiths <dave at pawfal.org> wrote:
>> On Mon, 2009-12-21 at 12:52 +0000, evan.raskob [lists] wrote:
>>>
>>> On Dec 21, 2009, at 12:57 AM, Glauber Alex Dias Prado wrote:
>>> >>>
>>> >>> I thought it was the music I was listening to and used Jaaa to
>>> >>> generate a pure sinewave but I still get the same result.
>>> >>>
>>> >>> Is this likely to be an issue with Fluxus, or
>>> >>> something/bad-configuration in Jack?
>>> > isnt the gh code not dependendant on jack, only on the audio
>>> > sources?(audio file)
>>>
>>> This is true, but noisy audio in == noisy result, which can be due to
>>> jack adding glitches and noise (this certainly happens to me a lot).
>>
>> If you read from a wav file jack isn't involved.
>>
>> It's much more likely to be a problem with the fft implementation in
>> fluxus than anything to do with jack (which is specifically designed to
>> prevent clicks and glitches, and works pretty well in my experience).
>>
>> cheers,
>>
>> dave
>>
>>