Log from #neuros IRC channel

brenden: as far as i'm aware, they have found one or two people who are not extremely faithful in their own background. so you might be the most qualified person to date that is intent on working on it

[07:46]

i plan on using similar hardware as the neuros, with ogg encoding

[07:46]

brenden: it's always hearsay though of course

[07:46]

well, i'm just a student

[07:46]

i don't have a formal background yet

[07:46]

brenden: let me get you the addresses for the list and joe

[07:46]

alright

[07:46]

oh, joe is joe born, president of neuros audio btw

[07:46]

yes, i gathered

[07:46]

i'm actually part of a team

[07:47]

he's pretty closely involved with this stuff as you probably also gathered

[07:47]

me and at least 2 others are working on our project, but we'd love to help out the neuros project too

[07:47]

well if you could get the whole team involved in it that would completely kick ass

[07:47]

:)

[07:48]

brenden: getting a GCC target done would pave the way for more developers to work on the firmware itself

[07:49]

it would definitely make my life easier

[07:49]

joes's email is jborn@neurosaudio.com ...

[07:49]

do you work for neuro or something?

[07:49]

it might be best for you to join the neuros-firmware and gcc target project lists and then mail them

[07:50]

brenden: no, i'm just a customer that's very interested in how things go :)

[07:50]

i see

[07:50]

brenden: the better all of this works out, the better my neuros will be :)

[07:50]

brenden: they're talking about putting in MPC and FLAC support, for instance :)

[07:51]

lists.sourceforge.net

[07:51]

that's the subscription form for the neuros-firmware list

[07:51]

lists.sourceforge.net

[07:52]

there's the other

[07:52]

one question i have: is there a specific license that the firmware and schematics are under?

[07:52]

if i were you i would subscribe to both, and address your email to both lists and joe

[07:52]

i didn't see anything

[07:52]

gpl would be nice

[07:53]

brenden: i don't know about the schematics. neuros has been very receptive to requests about licensing. They basically want to make it as easy as possible for people to contribute

[07:54]

i want to know how they feel about us using their firmware

[07:54]

we will still respect the open source-ness

[07:54]

but we would like to use it for an entirely different purpose

[07:54]

brenden: they ended up licensing the firmware under the MPL. there was discussion on the list about it and many of us wanted that far more than the GPL. This is because NA was planning to give out the source under the GPL to the community, but then maintain their own fork independantly. They didn't want to infect their fork with GPL'd code because they wanted the opportunity to license the stuff to other parties

[07:55]

hmm

[07:55]

brenden: by using MPL they can re-license for other parties with non-free licenses. but the MPL tree will always be quite free. IIRC, it has a GPL re-licensing option too so that one could create a GPL fork if one desired

[07:55]

that doesn't really work in my favour

[07:56]

brenden: well, I don't know how much you can tell me about your purposes, but I would imagine the crux would be whether you are a potential customer, potential competition, or just somebody who could benefit irrespective of NA

[07:56]

well

[07:56]

all i can say is that we're not making a product which will be sold

[07:56]

brenden: for instance, if you are working for Apple and you want to use the neuros FW in the ipod, I'm sure they would rather license it to you for a lot of money ;)

[07:57]

ahh

[07:57]

well

[07:57]

what we are making will remain in our hands

[07:57]

hehe it will be hard for you to deal with NA without telling them what you're doing :)

[07:57]

let's just say you are making another portable player

[07:58]

i can tell them anything if they will sign an NDA

[07:58]

you would probably have two options

[07:58]

i'm in the oil business

[07:58]

not the consumer electronics business

[07:58]

use their firmware source for free, but you have to obey the terms of the MPL -- giving the source back to Neuros and the community -- or the GPL, giving your source changes back to the community

[07:58]

(in my spare time though, i'm a full time student)

[07:59]

or pay them to license the product under a non-free license and keep the source to yourself

[07:59]

well, i'm just giving an example

[07:59]

the only source changes we would make would be to add ogg encoding for audio recording, and remove all the other features

[07:59]

all we really need to do is record audio

[07:59]

basically, the GPL and MPL don't mind if you want to make money off the source, the trick that you have to give your developments back for the "greater good"

[07:59]

why would you have to keep that to yourselves, then?

[07:59]

ehhh

[08:00]

it's complicated

[08:00]

the device is only a small part of the project

[08:00]

well you wouldn't have to open up the rest of your components :)

[08:00]

afaik

[08:00]

well, we plan on doing so anyways

[08:00]

i'm not quite following

[08:01]

the reasons i have to keep "hush hush" are really quite complicated

[08:01]

you mean you plan on having all of the source for your whole project be open source?

[08:01]

yes

[08:01]

we don't care about OUR ip

[08:01]

then you would not be in violation of their license

[08:01]

we care about infringing on someone else's ip

[08:01]

www.sparteksystems.com

[08:02]

there is some stuff on that site

[08:02]

you can keep any other code private so long as you don't put it into the neuros firmware tree

[08:02]

which will tell you all i can tell you

[08:02]

nah, we don't care about any code we add

[08:02]

so i don't see the problem :)

[08:02]

well, great

[08:03]

you plan on taking the neuros firmware source, making some modifications, and making the result available to them licensed under the MPL, right?

[08:03]

pretty much

[08:03]

that's quite acceptable

[08:03]

the only changes we need to make are to a. remove almost all the features and b. add ogg encoding from audio recording

[08:03]

so long as you do that, they won't bother you about how you're actually using it

[08:04]

so

[08:04]

do you have any clue what temp the device is good to?

[08:04]

we need 150C

[08:04]

the ogg encoding would be cool and i'm sure they'd be happy to receive the change and integrate it into the product

[08:04]

it's usually hit or miss with this kind of stuff though

[08:05]

removing almost all features of course would be fairly useless ;)

[08:05]

hrm well

[08:05]

yes, well, we don't need them

[08:05]

would you be using the whole product or components from it?

[08:05]

what's that temp in F?

[08:05]

we need the following: memory, usb mass storage, audio encoding, pressure recording, and recording some info from accelerometers

[08:06]

uhm

[08:06]

yeah, i just mean useless to them :)

i don't know F

hrmm

100C is boiling

0 is freezing

so..it's hot

about 200F is boiling

[08:06]

32F is freezing

[08:06]

so like hrm maybe 250-300F

[08:06]

make it ~275

[08:06]

yeah

[08:06]

i've already ordered a device

[08:07]

and we're going to put it in the oven first thing

[08:07]

but it would be nice to know ahead

[08:07]

so will you try to stick the whole thing in the oven or take it apart some

[08:07]

well

[08:07]

we'll take all the extra junk off

[08:07]

so just like the board and stuff

[08:07]

how about the LCD

[08:07]

and just basically put the oven in the circuit and run some wires to a scope to see what happens

[08:07]

we don't need the lcd yet

[08:07]

ok lemme find you the link to the basic hardware

[08:07]

we like the idea of being able to use it in the future though

[08:08]

well actually the link i have is for the next generation which will be used in the neuros 3

[08:08]

it will be about the same, and it will use the same firmware

[08:08]

the specific chip isn't that important

[08:08]

we may even use a dsPIC instead

[08:08]

because they are known to work at temp

[08:09]

www.spectrumdigital.com

[08:10]

ohhhh

[08:10]

nice

[08:10]

<inkedmn_>

the neuros databases are mdb files, yes?

[08:10]

damn

[08:10]

inkedmn_: no clue

[08:10]

i wish i had talked to you 30 mins ago

[08:10]

brenden: this information is good news? :)

[08:10]

<inkedmn_>

ok

[08:10]

i would have just bought that

[08:10]

hehe

[08:10]

brenden: i don't know if that would work with the neuros firmware right out of the box

[08:10]

err

[08:10]

actually

[08:11]

that doesn't look right

[08:11]

brenden: and the current device is designed differently.. it has another board for the hard drive backpack

[08:11]

* inkedmn_ dl's positron source

[08:11]

right now we're not interested in the hard drive stuff

[08:11]

brenden: that link is an "evaluation module" for the next generation of the board

[08:11]

so thats the neuros 3?

[08:11]

brenden: the neuros 3 that uses it will be out supposedly around q3 2005

[08:12]

well, we'll start with what we have i guess

[08:13]

doing ogg encoding on a fixed point device will be tricky

[08:14]

www.spectrumdigital.com

[08:14]

i think that is the board in the n1 and n2

DA250

i could be wrong

no

it's 150

DA150?

yes

you mean the n1/n2 are DA150, or that link is DA150?

n1/n2 are da150

ah

hmm

looks like we will be writing the first fixed-point ogg encoder

[08:17]

well, kickass ;)

[08:17]

okay

[08:17]

does it look like you'll be writing the first c54x gcc target as well? :)

[08:18]

mmmmmmm

[08:18]

perhaps

[08:18]

so, how about if you stick it in the oven and it melts?

[08:18]

it really depends on how much time we have

[08:18]

btw, why ogg?

[08:18]

why not mp3, wav, or flac?

[08:18]

if it melts, we will buy more stuff and stick it in the oven to see if it melts

[08:18]

no royalties

[08:19]

we don't need lossless compression

[08:19]

you mean move on and try something else besides neuros?

[08:19]

well

[08:19]

the only thing we're unsure of with the neuros is the ti chip

[08:19]

everything else is known to work

[08:19]

flac doesn't have royalties and somebody else may be writing encoding for that very soon

[08:19]

wav doesn't have royalties and it works now but of course that is much larger

you could call TI

memory is limited

hey

we cannot afford to waste it on flac

[08:20]

try speex

[08:20]

and we don't care that much about quality

[08:20]

if you wrote speex encoding and decoding for the neuros, that would be completely great

[08:20]

much better than ogg encoding :)

[08:20]

actually, it's Ogg Speex

[08:20]

as opposed to Ogg Vorbis

[08:21]

speex isn't really appropriate

[08:21]

ah, too bad :)

[08:21]

we're not recording speech

[08:21]

how high a quality do you need?

[08:21]

speex is good for really low quality and speech, which it sounds like it won't be

[08:21]

ogg is good for like 32-128kbit

[08:21]

you might try MPC at higher bitrates

[08:21]

64 bit would probably be good for us

[08:21]

up to about 300kbit

[08:22]

ah, then ogg really is a good bet

[08:22]

kbit, that is

[08:22]

another issue is cpu overhead

[08:22]

ogg is kinda bad on that

[08:22]

hmm

[08:22]

you might like mpc better for that even though you may not get as good quality with it at the lower bitrates

[08:22]

it is designed for higher.. but the difference at lower bitrates might be nominal

[08:23]

for decoding at least i know that they are just about the cheapest among all the codecs

[08:23]

decoding isn't an issue really

[08:23]

i don't know about encoding but at least they are not likely to be as bad as ogg :)

[08:24]

since they are trying to get MPC support done too, i'm sure they would be happy if you worked on that as well :)

[08:25]

we only have so much time

[08:25]

and as much as i'd like to contribute to the neuros community, our main focus is our own project

[08:27]

certainly :) but perhaps your interests can intersect :)

[08:28]

doing either vorbis encoding or mpc encoding would make people pretty happy; i would say mpc encoding would be more useful for the neuros, and it might also be better for you since it's free and may also be lower overhead

[08:29]

ogg is free

[08:29]

it might take longer, though, if you are familiar with vorbis enough to write an encoder easily but not as familiar with mpc

[08:29]

yeah, mpc and ogg are both just as free

[08:29]

we're concerned more about memory usage then quality

[08:30]

we may need to record as much as 5 hours of audio, and be limited to about 1gb of memory

[08:31]

well with ogg, depending on what you need, you might be able to go down to as little at 16kbit