Log from #neuros IRC channel

[00:00]

chreekat: yeah, i did notice some very specialized instructions when i last looked at the data sheet

[00:01]

chreekat: though that complexity might make your porting work more complicated, i guess

[00:05]

It'll make it more interesting, that's for sure.

[00:05]

nerochiaro: "Just to give an update, we are still planning on sending you an emulator, but more equipment is needed in addition to the emulator. Michael tested out one set, but needs to try another to make sure we send you equipment that will work for you."

[00:05]

from K, last october

[00:08]

GarBage: that's what i am saying, reminding that to them might be useful

[00:09]

<Koron>

yeah nerochiaro

[00:09]

<Koron>

i constantly get disconnected

[00:09]

<Seed>

nerochiaro: I should cross fingers for your connection to be more stable :)

[00:09]

Koron: you use which client ?

[00:09]

<Koron>

it's not my connection; other irc servers work absolutely fine

[00:09]

<Koron>

mIRC

[00:10]

<Koron>

i am using the exact same program instance to connect to two networks; this is the only one that gets disconnected all of the time

[00:10]

<Koron>

must be some kind of funky router issue

[00:10]

Koron: that was the same for me. the only client that works is irssi, for those with that disconnection problem

[00:10]

<Koron>

does it still get disconnected but it just realizes it sooner or does it not get disconnected as much?

[00:11]

<Seed>

mIRC can't tell when it's disconnected

[00:11]

<Koron>

it can on another network of mine

[00:11]

doesn't get disconnected at all. only times i get disconnected is when my ISP plays tricks on me

[00:11]

nerochiaro: i'd rather wait for this outsourcing hires to reach some good point, the equipment is not cheap (~ USD 3000) and it may come handy to them

[00:11]

<Koron>

perhaps the server sets how often the client is supposed to do a ping to check the connection?

[00:11]

<Seed>

the best you can do is start a timer that repeats every 10 minutes with something like "/lusers" in it

[00:12]

<Seed>

Koron: it does.. it pings the client in an attempt to get a response

[00:12]

GarBage: nice point. that's a worthy cause and they need it probably more than you

[00:13]

Koron: the only suggestion i can give you is: try irssi. that's the only thing that worked for me.

[00:13]

<Koron>

well, i may try the timer thing

[00:14]

<Koron>

i'd rather not have to use another client as I am pretty accustomed and content with mirc

[00:14]

<Seed>

Koron: the timer thing works for me at work

[00:14]

Koron: of course (*hushed* damn, my irssi evangelism failed once again)

[00:14]

;)

[00:17]

<Koron>

well, I am going to try the /lusers things...5 minute repeat

[00:17]

<Koron>

we'll see how that goes

[00:18]

<Seed>

try the command with the shortest output

[00:18]

<Seed>

don't abuse peernode

[00:18]

<Koron>

lusers has about 5 lines of text

[00:18]

<Seed>

do /whoid ChanServ :)

[00:18]

<Seed>

it's 4 lines

[00:18]

<Seed>

Whois

[00:19]

<Koron>

good idea

[00:19]

<Seed>

or.. /dns Koron

[00:20]

<Koron>

brb

[00:20]

<Yono>

alright, back

[00:20]

<Yono>

hmm still no joe

[00:20]

Seed: www.audiocoding.com

[00:20]

Seed: could you tell me what's wrong with it?

[00:21]

<Seed>

wrong?

[00:21]

<Seed>

I just don't like the variety of AAC encoders

[00:21]

<Seed>

they aim at semi-deaf semi-retarted Apple losers

[00:22]

<Yono>

GarBage: how difficult is it to add a lossless codec to the neuros fw, generally?

[00:22]

Seed: and then tune to some of those low-bw 32-64 kbit streams in here www.tuner2.com

[00:23]

Yono: it shouldn't be difficult, given it is coded in plain C, lossles or no, given you have the skills and the equipment to do that

[00:23]

<Yono>

ok

[00:26]

Seed: so, the "wrongness" lays in the variety of MPEG4/AAC encoders (it is still a evolving, open format prone to change), not the involving patents or anything?

[00:28]

<Seed>

hmm.. yes.. those are good points

[00:29]

<Seed>

Shy has many more.. I must sleep and I can't type

[00:31]

i'll go to bed. see you people.

[00:32]

nite all, i'm leaving as well

[00:34]

Heh, midafternoon here...

[00:35]

<Yono>

yeah, evening here

[00:35]

<Yono>

CiTay, are you going to stick around much longer?

[00:36]

<CiTay>

maybe 2 hours

[00:36]

<Yono>

ok

[00:36]

<ShyK>

anyone said aac? which implementation? :) AAC LC, Main, LTP, SSR, LD, HE AAC, PS AAC? or maybe MPEG-4 container formats m4a, m4p, m4b, mp4...

[00:36]

<Yono>

lol, too late shy

[00:36]

<ShyK>

:) yes. good night here too.

[00:39]

<CiTay>

Yono, can you send me a message when he's here? that will produce a sound

[00:39]

<Yono>

ok, I'll do that

[00:39]

<CiTay>

thx

[00:41]

Good night to everybody who has left/is leaving

[00:47]

<Yono>

CiTay: he's online

[00:48]

<Yono>

he'll probably come on here in a minute or so

[00:49]

hello

[00:49]

<Yono>

CiTay has the news

[00:50]

<CiTay>

hi

[00:50]

hello

[00:50]

<CiTay>

i spoke to Josh Coalson of FLAC

[00:50]

great

[00:50]

<CiTay>

his answer is as follows

[00:51]

<CiTay>

yes, I would like to see FLAC support on the neuros, but I will not be able to do it alone. I have made some effort integerizing the all the code, including the encoder, which will be in the upcoming release.

[00:51]

<CiTay>

but the big hurdle after that is the fact that the neuros dev environment has no 64bit types so I would have to go through all of libFLAC to work around it, which is significant work.

[00:51]

<CiTay>

beyond that is the task of integrating libFLAC into the player s/w which I am probably not the best person for.

[00:51]

<CiTay>

end quote

[00:51]

oh, I see

[00:52]

the dev environment he's talking about is ccs?

[00:52]

this is not an inherent limitation of the processor?

[00:52]

<CiTay>

not sure about that, i can ask him though

[00:53]

<CiTay>

i also spoke to some people concerning MPC

[00:53]

do we have a candidate for integrating it into the firmware, or would we have to do that internally?

[00:53]

<CiTay>

well, sounds like Josh isn't to keen on doing it himself

[00:54]

<CiTay>

i'm not sure how many other capable people there are to do the FLAC job, there might be some

[00:55]

if we have to do it internally, then I would prefer to do it once we've established our new firmware division

[00:55]

Concerning 64 bit, I bet it's both. I assume no C type maps to a 64bit datum in the TI development kit, and that's because doing anything with 64 bits would probably give a huge performance hit

[00:56]

<CiTay>

about MPC, i spoke to the former main dev Frank Klemm to ask about feasibility, telling that you use a 16 bit fixed-point DSP. he expressed concerns about that, saying a majority of the intermediate results needs 24 to 32 bits precision

[00:56]

CiTay: That, specifically, wouldn't be a problem.

[00:56]

<CiTay>

he also estimated it would take 10 workdays of programming and 3 days of testing to implement it

[00:57]

<CiTay>

it won't, chreekat? how come?

[00:57]

It has two accumulators that can do 32 bits

[00:57]

<ShyK>

really? :o

[00:57]

<ShyK>

that's great

[00:57]

If I've understood the specs correctly. :)

[00:57]

<ShyK>

(not sleeping yet)

[00:57]

<ShyK>

chreekat: if this is the case, it's fantastic

[00:58]

<Yono>

(20:53:32) unknown_lamer: GarBage: can the neuros dsp use 32-bit ints at all, or is it all 16-bit?

[00:58]

<Yono>

(20:53:55) GarBage: unknown_lamer: 32 bits are fine

[00:58]

<Yono>

I'd say ur correct chreekat

[00:58]

<ShyK>

wow. then i see things are much less complicated than we first thought.

[00:59]

Oops, I'm supposed to be takig a break from the computer, back in a bit

[01:00]

* chreekat tests the capabilities of the Neuros for recording drum practices

[01:01]

<Yono>

so actually MPC would be much easier than FLAC

[01:01]

<CiTay>

we're looking for someone who could implement MPC. i already know someone who would likely do it, but only if we can't find a more capable coder than himself, he says. he has a bit of a low confidence about this job :)

[01:01]

oh my!

[01:01]

<ShyK>

we've posted about the need for a firmware programmer to give people who might be capable a chance

[01:02]

<ShyK>

in some days we'll see

[01:02]

<CiTay>

he did a lot of programming before, don't get me wrong. audio-related of course

[01:03]

well, please keep me posted.

[01:03]

<CiTay>

when i told him about it, i didn't know that the 16 bit limitation isn't such a limitation after all, so now things look brighter

[01:04]

<CiTay>

i will

[01:04]

<ShyK>

where can we get all the info about the hardware?

[01:04]

<ShyK>

to make sure 32bit is fully supported, etc

[01:05]

does anyone have the datasheets?

[01:05]

<Yono>

yeah

[01:05]

<Yono>

K gave me the original hw page for open.neurosaudio.com that never made it

[01:05]

<Yono>

let me see...

[01:06]

<Yono>

neuros.php-help.com

[01:06]

<CiTay>

ironically, the former lead developer Frank Klemm works at Carl Zeiss Microscopy... programming DSPs. :)

[01:07]

<ShyK>

Yono: thanks

[01:07]

keep me posted, its very exciting.

[01:07]

<ShyK>

sure

[01:08]

things like this could have a significant impact.

[01:08]

<CiTay>

indeed

[01:08]

<ShyK>

a VERY significant one

[01:08]

<Yono>

Joe: is there any point in the day that your phone is on that you aren't busy?

[01:09]

It will speak very powerfully to TI and the industry in general if we got FLAC and MPC implemented by open source developers

[01:09]

<Yono>

most people here are from europe and are usually asleep by the time you get on a computer

[01:09]

<ShyK>

JoeBorn: exactly what i think. it'll benefit everyone.

[01:09]

as a result of open source, Ogg Vorbis was implemented on TI long before Sigmatel.

[01:10]

as a result of open source developers, I should say.

[01:10]

yes, I can talk during the day.

[01:10]

<Yono>

ok

[01:10]

tomorrow is pretty bad though, until at least late afternoon.

[01:11]

<Yono>

alright, I'd feel bad messaging you on ur phone and u being in an important meeting or something

[01:11]

that's ok, you can message. it's better than a call :)

[01:12]

<Yono>

lol, alright

[01:13]

some of the proprietary codec providers want to license Ogg from us.

[01:13]

it speaks pretty powerfully to the potential impact.

[01:13]

<ShyK>

JoeBorn: CiTay probably mentioned it, but i'll say that Musepack is exactly the kind of format fit for portable players. fastest modern lossy format, very high quality (as has been constantly proven), and completely free, and, the request for hardware support is more huge than most people think.

[01:13]

<Yono>

well, you and iriver are the only ones that really seem to support alternative codecs

[01:14]

<Yono>

sorry, rio, not iriver

[01:14]

well, the Karma has a few.

[01:14]

iRiver too.

[01:14]

<Yono>

yes, but Karma was the first "popular" player to support flac

[01:15]

but I think one difference is that we embrace the movement. Not only to we implement open source codecs, but we open source the code that does it as well.

[01:16]

if we can demonstrate the power of that, it will really demonstrate something.

[01:16]

<Yono>

and after TI is done evaluating, that will be even moreso

[01:16]

as of now, TI probably views the open source stuff as more of a nuisance, but all this could give them an advantage

[01:17]

that they can benefit from.

[01:18]

well, I should run, but please keep me posted and I'll be available late afternoon tomorrow.

[01:18]

<ShyK>

that's what i like about projects like yours, that can really prove just how big an impact supporting widely used open source formats can be.

[01:18]

<ShyK>

and finally someone is doing it

[01:18]

<ShyK>

JoeBorn: see you

[01:18]

<CiTay>

sure, JoeBorn. see you then

[01:18]

<Yono>

later joe

[01:18]

thanks, talk to you soon.

[01:26]

<CiTay>

off to bed..

[03:15]

Heh

[03:15]

<Yono>

hello

[03:16]

<noiz>

hey

[03:16]

Since the DSP implement entire algorithms with a single instruction, it would be cool if a compiler could catch if the same algorithm is implemented in C

[03:17]

I mean, I'm assuming not. But it would be awesome

[03:20]

<eozl>

isn't conventional wisdom "c is a professional language, do exactly what the professional programming me wants"?

[03:35]

<_Silmaril>

CUT THAT MEAT....CUT THAT MEAT

[03:37]

<noiz>

...

[03:37]

<_Silmaril>

hi there.

[03:37]

<noiz>

hey

[03:38]

<noiz>

so how is that meat coming?

[03:38]

<_Silmaril>

good, real good

[03:38]

<_Silmaril>

any idea when the next sorune is set to release?

[03:38]

<noiz>

in the next couple of days

[03:39]

<noiz>

you can sign up for the releases mailing list if you want

[03:40]

<_Silmaril>

word

[03:40]

<_Silmaril>

thanks man

[03:40]

<noiz>

its on the main page if youre wondering where it is

[03:41]

<noiz>

so far there are 2 people that have signed up for it

[03:42]

<_Silmaril>

I'm a big fan of it

[03:42]

<_Silmaril>

the rebuilds are clutch on a device that corrupts as much as my baby :)

[03:42]

<noiz>

:)

[03:47]

<noiz>

hey

[03:47]

hey

[03:48]

<noiz>

i think i got another person that will sign up for the releases mailing list

[03:48]

<noiz>

up to 3 people now :)

[05:08]

* SmrtJustin needs to sign up for the releases mailing list.

[05:09]

<noiz>

theres a link on the main sorune.com page

[05:34]

night all

[05:34]

<noiz>

later

[05:34]

noiz: I've been meaning to ask, what host did you guys switch to?

[05:36]

<noiz>

totalchoice hosting

[05:36]

<noiz>

very good

[05:36]

<noiz>

and cheap :)

[05:36]

cool

[05:37]

I suggested to sorune sourceforge because of the free hosting, but thats good you guys found someone

[05:38]

well, night

[07:58]

<pjz>

howdy, anyone awake?

[07:59]

<pjz>

I was considering a Neuros and wondering if I should bother with the Neuros II or is the 3 about to come out?

[09:43]

pjz: Nothing official has been said about the 3, I think

[11:36]

The N3 wont be out utnil late summer at best

3 or mini ?

n3

mini is coming soon

well, the summer is not too bad

[11:42]

i hope they take all the time to make things right this time

[11:44]

I hope I can get one easily in Aus :o)

[11:45]

pwarren: there's not much difference in getting one to australia or to europe i guess

[11:46]

outside the US, is hard

[11:46]

not really, except the Euro is worth more than AUDs

[11:47]

We make as easy as possible

[11:47]

AmericanTechpush: i know, i got my N2 from you and you did a great job with it

[11:48]

well, I'll either go with you guys or Cool4u2view

[11:48]

A big part was getting the Credit card for International

[11:48]

I have not seen Jeff in quite some time

[11:48]

neither. ah well.

[11:48]

really smart guy

[11:48]

Off at college most likly

[11:49]

heh, I go back in a few weeks.

[11:49]

AmericanTechpush: i think that one fundamental thing for international orders is to guarantee some in-house testing before sending. faulty units are major pain in the arse for us on the other side of the pond

[11:50]

by doing so we have to open the product

[11:50]

Some might think that they are getting used

[11:50]

which DI says is tested before leaving them!

[11:51]

AmericanTechpush: well, would you do that in case a customer explicitly asks for that testing, though ?

[11:51]

If they asked for it sure I wouild

[11:51]

AmericanTechpush: that's good to know

[11:52]

be it already takes a long time to get the orders out. IT will add to a already long time

[11:52]

well, it's kind of an added insurance, so i guess it might be worth the wait

[11:53]

it's surely better than to have to send back a faulty product across the atlantic

[11:53]

depends on the person

[11:53]

of course. i'm speaking for myself here

[11:54]

When did you order?

[11:55]

this autumn

[11:55]

Garbage got his from us as well

[11:56]

i was convinced the got his from jeff.

nope

From us

hi DeepB

<DeepB>

hi nero

[20:26]

i am testing audioscrobbler. looks like a nice idea

[20:26]

<DeepB>

seen last.fm as well?

[20:26]

i'm not really into internet radio

[20:26]

man, USB disk mode kills my battery

[20:27]

but i might give it a shot later on

[20:27]

JoshMalone: i rarely use the USB disk without the power cord plugged

[20:27]

Q-W-Y: wow! same 1KHz distortion here :(

[20:27]

<Q-W-Y>

eh? Really??

[20:27]

looks like a DSP problem or maybe voltage problem :(

[20:28]

<Q-W-Y>

what about 15kHz?

[20:28]

<DeepB>

brb

[20:30]

Q-W-Y: dunno - but 1KHz @ 0.3db peak is BAD

[20:31]

<Q-W-Y>

:/

[20:31]

i'm beginning to doubt that my ears are really bad, then

[20:31]

and it's nothing to do with output volume

[20:31]

i really can't hear the distorsion when i listen to the test sine signals

[20:31]

nerochiaro: below about 2-2.5 db there's no distortion

[20:32]

JoshMalone: how do you raise the gain ?

[20:32]

when you create the file, or useing mp3gain on a mp3 file

[20:33]

(mp3gain == linux app)

[20:33]

will replaygain do it, too ?

[20:33]

maybe - prolly

[20:33]

is it the "emphasis" field in the mp3 files, or am i completely off-base ?

[20:34]

<Seed>

hi guys

[20:34]

ooo! it's the EQ!

[20:34]

hi seed

[20:35]

seed: somehow your arrival inspired in me the need for a cup of tea. don't ask me why. i'll go prepare one.

[20:36]

I turned off the EQ and it cleaned up

[20:36]

<Q-W-Y>

my EQ is off

[20:37]

yeah - so my unit doesn't have the distortion your does :-/

[20:37]

<Seed>

nerochiaro: I just finished my cup

[20:38]

Seed: must be some kind of telepathy then

[20:38]

<Seed>

could be

[20:38]

<Seed>

maybe we even listen to the same band

[20:39]

possible, but harder. what do you have playing ?

[20:39]

<Seed>

ABBA

[20:39]

mmh, no. Tiamat

[20:39]

<Seed>

I'd never listen to Tiamat

[20:39]

<Seed>

that's it, we're through

[20:40]

same for me with ABBA

[20:46]

<Q-W-Y>

i am clueless

[20:46]

Q-W-Y: are you ?

[20:47]

<Q-W-Y>

good thing is, that i now now that i hear to 20kHz

[20:47]

Q-W-Y: maybe you have defective brain?

[20:47]

heheh

[20:47]

<Q-W-Y>

yes, probably i have defective brain. .. :D

[20:48]

sorry to hear than, man

[20:48]

I'll hafta check 15KHz on my unit

[20:48]

<Q-W-Y>

problem is, that i live in CZ

[20:48]

:( ^ 3

[20:48]

<Q-W-Y>

i've created new test files

[20:48]

<Q-W-Y>

full range

[20:48]

<Q-W-Y>

from 20 hz to 20 kHz

[20:56]

<Q-W-Y>

whow VERY interesting

[20:57]

<Q-W-Y>

hmmm it was OK on lower gain

[20:57]

yeah - maybe it's the 3V supply where 3.3 is more nominal

[20:57]

<Q-W-Y>

my battery has 4 volts

[20:57]

(we had a almost identical problem with ethernet on one of our boards - biasing magnetics with 3.0 instead of 3.3)

[20:58]

Q-W-Y: no - the schematic shows 3.0V powering the DSP in a spot where 3.3 is nominal

[20:58]

<Q-W-Y>

ah

[20:58]

<Q-W-Y>

sorry

[20:58]

bias problem in the nternal headphone amp, I'll bet

[21:00]

they need to get the HPVDD up to 3.3 I'll bet

[21:00]

once you guys nail down the problem for good, it would be interesting to hear what NA's engineers think about it

[21:00]

question is - can it be changed easily... :-/

[21:00]

<Q-W-Y>

hmm so maybe i could connect it to 3.3 V

[21:01]

weird - the schematics don't seem to show the power supply :(

[21:02]

wait - that's proll y in the backpack

[21:03]

<Q-W-Y>

HMM very interesting

[21:03]

<Q-W-Y>

(wave in soundforge

[21:03]

<Q-W-Y>

i will upload it somewhere

[21:17]

<Q-W-Y>

there is a screenshot

[21:18]

<Q-W-Y>

ceskyinterier.cz

[21:19]

Q-W-Y: what firmware?

[21:19]

<Q-W-Y>

garbage *13

[21:19]

huh?

[21:19]

<Q-W-Y>

i've tried 2.28 too

[21:20]

sounds like you've got major EQ bugs....

[21:20]

similar to the EQ corruption seen in 2.23 official firmware back when

[21:21]

<Q-W-Y>

i dont use eq...

[21:22]

but, it may be doing things behind your back

[21:22]

N 2?

[21:22]

<Q-W-Y>

yes

[21:23]

I can't help but notice the similarites btwn your behaviour and the 2.23 EQ bugs

[21:36]

<Yono>

hey all

[21:36]

hi

[21:37]

<Q-W-Y>

yono look at this

[21:37]

<Q-W-Y>

ceskyinterier.cz

[21:38]

<Q-W-Y>

it is full range freq from 20 Hz to 20kHz on my neuros

[21:38]

<Yono>

interesting...

[21:39]

<Yono>

hmm, I'm not very firmiliar with sound forge

[21:40]

<Yono>

but from what I see neuros is better with certain freqs

[21:40]

<Yono>

is this considered a good range for a portable player?

[21:42]

<Q-W-Y>

sure

[21:42]

<Q-W-Y>

it is bug of my player

[21:42]

<Q-W-Y>

others dont have this problem

[21:42]

<Yono>

oooo

[21:42]

<Yono>

now I understand

[21:43]

<Yono>

was this through the same line-in mic as the source?

[21:45]

<Q-W-Y>

yes

[21:45]

<Q-W-Y>

i will upload the samples

[21:46]

<Q-W-Y>

20 minutes remains :p

[21:46]

<Yono>

this might make an interesting wiki topic

[21:47]

<Yono>

what do you think nerochiaro?

[21:47]

<Yono>

if it is enough of a widespread problem

[21:48]

<Q-W-Y>

I've witten to support, but the are ignorat

[21:49]

<Yono>

yeah

[21:49]

<Q-W-Y>

They are talking about EQ... pfff my EQ is turned off

[21:49]

<Yono>

I doubt it has to do with eq

[21:49]

<Yono>

think its the jack?

[21:50]

<Yono>

no, that wouldn't be it

[21:51]

hmm odd how bashpodder isnt noticing all the files that the site says there is...

[21:52]

<Yono>

maybe a DSP problem