[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: MLS code for bounty?
From: |
Butler |
Subject: |
Re: MLS code for bounty? |
Date: |
Fri, 07 Nov 2008 13:21:29 -0500 |
User-agent: |
Thunderbird 2.0.0.17 (Macintosh/20080914) |
Thanks for the reply Jordi. Details, then:
Again, I should mention up front that my math skills are pretty much
non-existent so if I seem to know anything about the below that is just
due to me trying hard to understand my problem and researching but does
not indicate any level of true understanding. I am weird in that while
my math skills don't exist, I have some understanding of programming.
I am a sound engineer attempting to create some flat panel sound
diffusers (based on ideas from a paper given at an Audio Engineering
Society convention in NYC some years ago-historically, diffusion panels
must have great depth to work, often causing an architectural dilemma
since who wants to give up several feet of space next to a wall in their
control room. The idea behind this new-ish design is essentially to
substitute absorption for depth) for use in my recording studio. This
involves creating a (acoustically) reflective mask with a pseudo-random
arrangement of holes to cover an absorptive panel. Grid cell dimensions
are immaterial to this process but in this initial case will be 1" squares.
My understanding is that I need to generate a binary m-sequence a/k/a
maximum length sequence:
http://en.wikipedia.org/wiki/Maximum_length_sequence
... to be mapped onto a specific grid (in this initial case 49x30) via
the Chinese Remainder Theorem.
Not having much of a budget for MatLab, it seems like Octave, as an open
source solution, would be ideal. What I would like is code to generate
an m-sequence and next a grid, based on the grid dimensions. So grid
dimensions are input first, then an m-sequence that can be mapped to
them generated, then that m-sequence mapped. It's my understanding that
the top/side dimensions of the grid must be coprime for this to work,
which (yes, this is the er, depth of my ignorance) I believe 49x30 would be.
There may also be some considerations vis a vis a convenient output
format, but first-thoughts? Does this sound possible? Difficult/involved
(possibly beyond my means in terms of $$)? I'm hoping maybe this could
result in some new function(s) for Octave as a bonus to the community if
not all functionality needed for this exists.
Best,
CDB
Jordi GutiƩrrez Hermoso wrote:
2008/11/7 Butler <address@hidden>:
So this is not a question re: help with octave, more a
question asking whether there is a place where I could offer a bounty
for working code? I could post details but don't want to start an
off-topic discussion...
This is probably ontopic in this mailing list, but you could also post
the details of your problem to the Octave maintainers list at
address@hidden
I'm eager to hear the details.
- Jordi G. H.
--
Chris D. Butler
address@hidden
718-417-0690
917-902-6736