Topic: Archos Gmini 120

[jerryfreak]

alas, the longer test (23 minutes) with the v3 indicated a series of glitches around 12:00, a combination of random noise and dropped samples. I'm doing an ad2k test overnight...

[rustoleum]

I was able to update the firmware successfully last night. It looks like I somehow managed to save it as a different name when I did this a few days ago.    I now can select WAV from the recording option.

I still haven't successfully fed the thing a spdif signal, but I only tried using my d100 which is on it's way out and could be the problem (although I think I should have seen _something_ anyway).  I am now using the included cable (don't know how I missed that before).  I'll get the apogee out but it may not be for a couple days as I've got a full week ahead of me.  Pretty cool that the AD2K seems to be more reliable than the V3... gives me hope that the MME will behave differently than the V3 and possibly give a flawless recording.

I think either way I'll keep this thing around as it is pretty nifty to have a tiny 20 g harddrive that I can offload files to while taping on my lappy at a festival or other situation where the internal drive isn't cutting it.

Rusty

[JasonSobel]

Hey guys, thanks for testing this thing wtih me...  Riley, interesting thing about the 19 seconds...

Well, what now?  Is it time to talk to Archos and battle with tech support to see if we can make them understand what's going on here, or should we do a little more testing first?

YES, definitely.  I've been calling them and emailing them, with no real results.  it takes a few days for them to answer each email, and I don't think the same person is reading and responding to my emails.  and the phone conversations I've had with them don't seem to help either.  They'd like to write me off as an isolated case.  I think it would be great if everyone who is testing this thing calls and emails them and bring to their attention the deficiencies that currently exist, and that could probably be fixed with firmware updates...

I think the first priority is to get it to reliably record at 16/44.1 and 16/48.  once it does that, maybe we can get them to record 24/48 and 24/96...

if anyone has any more test results to post, keep them coming.  and also, please post if you have any positive (or negative feedback from archos)

I ran another quick test last night.  I generated a 5 second 440Hz wav, put it on repeat, and played it analog out of my computer, into the V3, then digi into the Gmini.  I recorded this for just over 10 minutes.  I wasn't testing for bit-accuracy, but simply the glitches.  I had the same results, although I didn't take the time to note exactly where the glitches occur.  sometimes the glitches were just one sample, but I've seen errors up to 6 samples in a row.  usually happens in both channels, but, again, I've seen it in just one channel as well.  

I may take it out tonight, because, hey, what do I have to lose?  if anything, it will be a good test for the battery.  

and +t for everyone else who's also testing this
 > Jason

[timP]

I really think that the pops and such are from the HD spinning...if it spins fast enough, couldn't it work up a charge or something that messes up the recording...

People with IRIVER mini HD recorders are experienceing the same pops and clicks, timed when the HD spins....
(edit: the iriver has a light that comes on only when the HD is spinning, and you can time it out to where the pops are....)

[seethreepo]

I may take it out tonight, because, hey, what do I have to lose?  if anything, it will be a good test for the battery.  

and +t for everyone else who's also testing this
 > Jason


+t's indeed,  if its not  too much to ask can you run 1 set via analog line in ,  (I'm in a huge minority  here)  but I dont use a a\d or pre just mics and deck and if  it can approach the quality of the dats a\d I'm sold .  

if the audible clicks occur via line in I'm out but ...

tia!
CB

[chase]

+t's to all the testers!

i am very curious to see how this turns out.  this dropped sample business is a bit disconcerning although creative overcame this through firmware updates.  hopefully archos can do the same, but you have to remember how small and complex these things are, and that electromagnetic fields are generated by basically anything that can carry a charge in there.  i think this is one of the major problems in making audio gear in such a small package, but i am hopefull that this will get worked out so maybe i can get rid of my d8.

[greenone]

Jason, spread, et al - Chase's post leads me tp ask this: are they *dropped* samples, or are they glitchy samples? I.e. are the files identical in length over time and match up at arbitrary points throughout the recording?

This may just be a red herring, of course - when Jamie Lutch claimed that the JB3 mis-recorded samples it turned out to be the CO2 (at least based on Brian's use of the Hosa ODL-276...I may have that number wrong). Still, if the Gmini is dropping samples then I'd say that's a more serious problem than glitchy ones...

[jerryfreak]

from the results I've seen with others testing, it is *misplacing* samples, not dropping samples.

the nomad didnt drop samples either, it also misplaced them (and still does with many spdif sources)

misplaced samples are much harder to detect in a live recording situation, becasue it could be mistaken for a clap, a drum hit, crowd noise, etc. whereas a dropped sample is really evident as a 'skip' (part of the data is missing)

my ad2k test last night didnt work out, I guess the device died before saving it. there was 4gb less space on my drive this morning, but no file. I did find a 4gb temp file with a recovery program I had, but it claimed it was corrupt.

[JasonSobel]

my tests indicate that they are just misplaced samples, not dropped samples.   with my bit-accuracy tests, when I trimmed the two .wav files at the exact same starting and ending point, resulted in two files identical in length.  just that the archos recording had glitches in it.

if its not  too much to ask can you run 1 set via analog line in

right now, I don't have any XLR to mini cables...  I've been meaning to make a set of right-angle XLRs to 1/8" to go analog out of the V3, but I haven't gotten around to it yet.  but I guess I could analog out of the D8, with my mini to mini cable...  as always, I'll post all results here.

[timP]

I'm going to record silence and see if it shows any glitches.....
fingers crossed.............

[timP]

I did 2 tests vis line in, with my mics in a slinet room with pillows and a blanket on them for silence, each 5+ minutes and all got when I put them into wavelabs all I got was flat line. So, this should mean good news for analog recordings.............

[Brian Skalinder]

I did 2 tests vis line in, with my mics in a slinet room with pillows and a blanket on them for silence, each 5+ minutes and all got when I put them into wavelabs all I got was flat line. So, this should mean good news for analog recordings.............

I'd suggest a bit more testing before declaring all is well for analog-in recordings.  And why the silent room / pillows / blanket?  Doesn't seem like a very realistic field test to me.  Just curious...

[timP]

I was trying to see if it would make skips and pops while in silence...
def don't want to jump the gun with it at all, just have made some analog recordings that I listened to on decent headphones and can hear no skips or pops...
what would be a better way for me to create a clean test? just trying out ideas..

[Brian Skalinder]

I was trying to see if it would make skips and pops while in silence...
def don't want to jump the gun with it at all, just have made some analog recordings that I listened to on decent headphones and can hear no skips or pops...
what would be a better way for me to create a clean test? just trying out ideas..

Well, since it's analog-in, there's no real way to compare the recording against a control.  So I think just more time testing is a good idea, and in more realistic conditions - a loud sound source, using different levels of gain on the Gmini, etc.

[timP]

right on. I was thinking that if it was making skips via SPDIF in when the HD saves every 30 sec or so then it might make the same skips via line in, and they would show up as pops or skips