Re: [hatari-devel] Use correct white noise

[Nicolas Pomarède <npomarede@xxxxxxxxxxxx>]

Le 14/07/2012 22:01, David Savinkoff a écrit :
> Hi,
>
> This patch uses the correct LFSR for the YM2149. The previous must
> have had a typo. currentNoise ^= YM2149_RndCompute() is corrected.
> BTW the noise doesn't sound different.
>
> The code is also optimized in Galois form.
>
> see: (also note 17 stage table at end)
> http://www.newwaveinstruments.com/resources/articles/m_sequence_linear_feedback_shift_register_lfsr.htm
>
>
> David

Hello

sorry for the late reply, been quite busy lately. I will modifiy Hatari 
with your patch as soon as I can (I have some changes to finish/push before)

Regarding the fact that it doesn't really change the perceived result, I 
also noticed that some times ago while trying different random methods 
(the one you sent is similar to the one used in MAME for example). Even 
using the rand() function gives similar result.

In the end as this is white noise, the only thing that really matters is 
the period in reg6 and the frequency at which we compute a new value ; 
as long as the pseudo-random function is close enough to randomness, 
result is not really discernible.

Nicolas