On 09/02/2015 01:41 PM, Marcus D. Leech wrote:
> On 09/02/2015 01:35 PM, Marcus Müller wrote:
>>> I'd say we rather want better
>>> randomness than faster noise sources
>> If that's the case, I'd recommend that Stefan uses the
>> normal_distribution variate from boost (Stefan, see my example), rather
>> than doing his own "normalization" of the RV, and we use that.
>>
>> Regarding Boost's mt19937 and the ways Stefan and I make normal
>> distributed RV out of the uint32_t that this emits:
>> From getting "raw" uniform uint32_t instead of normal floats through
>> boost's normal_distribution(rng) variate is that, to little surprise,
>> 75% of time is spent looking up/interpolating/calculating based on the
>> uniform integers to normally distributed floats. Blind guess is that
>> this would be something that someone who really cares about speed might
>> implement in VOLK with sufficient accuracy, making WGN generation
>> X3x0-capable.
>>
>> Cheers,
>> Marcus
>>
> A call to SHA-256 can return 256 bytes at a time.... Map those
> bytes into {-1.0,1.0}. That may still not be fast enough, dunno. But
> the distribution
> should be good.
>
>
Gah, that should be 256 *bits*, or 32 *bytes* at a time. Blame my
cats. They keep me awake at night. Yeah, that's it.
_______________________________________________
Discuss-gnuradio mailing list
Discuss-gnuradio@gnu.org
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
No comments:
Post a Comment