Topic: iriver H1xx Tapers FAQ (beta)

[poppag76]

I'm using 08.05.06.  I hope there isn't a glitch with it.

[George]

I'm using 08.05.06.  I hope there isn't a glitch with it.

It's very possible.  Just try an older build and see what happens.

[petur]

I'm not aware of any glitches at the moment, playback should be quite stable.

So you were playing a big WAV file and hit the joystick.
- playback near the end of the file?
- pressed the joystick or hit it left/right (prev/next)?

It's always good to keep a 'last known stable' version somewhere, but I would encourage to keep up to date from time to time as bugs get fixed. But always test before going to a show

[BayTaynt3d]

Why not use the patched build of the guys working on the Recording Enhancement Pack (REP)? Go to Rockbox forums, find the iRiver section, then find the thread that discusses the Recording Enhancement Pack. In the very first post, or in the sig of one of the main developers who posts all over that thread, you will find a link to a build that includes the REP. It's got some cool extra features including a historical levels chart, a balance gage, and AGC functionality. Now, while the AGC functionality will probably be turned off for most of us, there is one useful setting called "Safety" that will only turn down the gain by .5 db if clipping occurs and then it leaves the gain alone. That's kind of a nice feature for stealthing. I mean, you hope you set you levels right to begin with, but it sure is nice to know that the unit will turn down the gain for you if you are clipping. This build is of course probably not as stable, but I think it's proven itself enough to be worthy. Anyone else using a build with the REP patched in?

[whatboutbob]

Petur: Big wav files freeze up on me semi-regularly, but I'm using 060719 (just after freeze lifted I think).  I pretty much just have to try to rewind and she freezes up. I thought there were known issues with big wav file playback so I didn't bother reporting.

[poppag76]

I'm not aware of any glitches at the moment, playback should be quite stable.

So you were playing a big WAV file and hit the joystick.
- playback near the end of the file?
- pressed the joystick or hit it left/right (prev/next)?

It's always good to keep a 'last known stable' version somewhere, but I would encourage to keep up to date from time to time as bugs get fixed. But always test before going to a show

Well, it wasn't really big.  It was about 39 minutes.  About 300 megs.  I was trying to turn the volume down and accidently pushed down on the joystick and it just froze up.

Also, I was recording an hour 1/2 long show from my JB3 and the levels disappeared and acted as if I was stopping then started back up again.  The time jumped forward as if it recorded the missing section(which it did do).  Anyone else experience the phenomenon?

[whatboutbob]

that's normal during buffer-writes.

[shaggy]

that's normal during buffer-writes.

I found myself like 3 times out of 10 looking at the remote with the peak display blank when I needed to check levels (either incredibly bad luck or it is something that needs to be fixed).  I did about 14 hours worth of taping the other weekend at Fuji Rock and it was mildly frustrating.

Oh, no remote noise associated with the remote with SPDIF in.  I was just seeing the Lemosax and modSBM noise floor (certainly not the mk4v's noise floor).

[whatboutbob]

There is no known patch or work-around.

This sounds far more aggressive than i intend it, but please be careful with your "tone" shaggy.  I'm confident you didn't mean to sound demanding...but remember that rockbox is an open-source project.  All developers are volunteers.  Remember that iriver firmware has no recording levels whatsoever, and without the rb guys the iriver is pretty much useless as a taping tool.  If you were to post that something "needs to be fixed" over at rockbox.org then you would be welcomed to learn the code and pitch in.

Off the top of my head, at 16/44.1 the buffer write takes about 14 seconds approximately every 2:50 mins...so yeah...3 out of 10 is pretty unlucky. I've found it marginally inconvenient at times...but not enough to annoy me.  Would i prefer it worked all the time?  Absolutely.  Has it been looked at extensively by the rockbox developers?  Yes. With any results?  No.  I wouldn't be shocked if it could be "fixed", perhaps at the expense of battery life, but I'm no developer so I'm likely talking out of my proverbial.

Petur: anything to add?  Prolly should take this discussion over to rb anyways...

[shaggy]

WAB, Yeah, I been spanked a few times already on the RB forums and at the mistic river forums.  I guess we are a little brash over here, tho I consider myself pretty tame and polite compared to others.  I am really grateful for what they've done (I ditched the JB3 for two H120s).  Things will continue to improve, it is open sourced as you mentioned and some people's hobbies seem to lean towards tinkering at the code level to perfect things.  I edited the comment so the developers reading this will not get pissed off at us.

I had issues reading the levels due to me not having the two H120s I was working with set on the same scale.  So I suppose I was looking more frequently than I usually do at the meters.

Which brings me to ask, what do people set their scales and the min and max if they are using the log scale (in lin mode, it defaults to 0-100%).  I find the linear scale a  bit more open between -12 and 0, much like the way the M1 meters are set up.

[petur]

Petur: anything to add?  Prolly should take this discussion over to rb anyways...

The levels are sampled by the same task that does the sample conversion (to prepare for writing to file) and the writing to disk itself if the buffer gets full. So when it starts writing to disk, there are no new peak samples...

So this is a design feature at the moment, nothing much to do about it at the moment (meaning: changing it is not a simple fix).

I must admit that the missing peak info is a little inconvenient but that's life for now

[cgrooves]

Which brings me to ask, what do people set their scales and the min and max if they are using the log scale (in lin mode, it defaults to 0-100%).  I find the linear scale a  bit more open between -12 and 0, much like the way the M1 meters are set up.

Shaggy - I set mine up at -12 & 0.  I've read that there was a levels enhancement put in around 07.04.06 that was beneficial if you use that setting as well.

General comment directed at nobody -  I have found that the levels are VERY responsive, and much better than many recorders out there, IMO.  The fact that you lose levels while writing the buffer is no biggie to me.  Once I set them, I try to not be so paranoid about it and checking them all the time.  I will look occassionally to see if the light has come on due to clipping, but that's it.  The soundman should have everything dialed in by the second or third song anyway.  People that are overly paranoid about clipping can always run the levels conservatively and just normalize in post.  My .02

[BayTaynt3d]

So is anyone else running with the REP installed? If so, what build/patch is most stable with REP?

[cgrooves]

So is anyone else running with the REP installed? If so, what build/patch is most stable with REP?

I have not, and have not planned on running the REP personally.  No help here.

[Humbug]

I've been running an H120 for a few weeks now and found the FAQ invaluable.

I particularly like the ease of use, the recording buffer, the size, the way you can flip the screen 180°, and that it plays FLAC files.

The question I have concerns gain. On a JB3, I read that gain is digital, and was (more or less) useless.

Is this the case on the H120? Should I set it at zero and add it in in post, or try and set correct gain in the field?

Does it increase the noise floor?

Apologies if this question has been answered somewhere else..