Topic: Core Sound PDAudio-CF

[stunt_show]

Hey there,

On the Core Sound website (www.core-sound.com) it says the PDAudio-CF is available now.  That's not true is it?

I think they mean it is available to be ordered, actual release date unknown at this stage.

Anyhow, will this new recorder be the next greatest thing in the digital-era?  What are people's thoughts on it, and the future upgrades it may allow?

Regards,

stunt_show

[creekfreak]

I will believe it when I see a working model.

[John Kelly]

Well apparently someone here already has a demo, but I doubt I'll ever use one.

[Craig T]

I run the following hardware/software configuration:
HP iPAQ 3835
HP dual pc card expansion pack
Toshiba 5GB pcmcia hard drive (FAT32, 32kb clusters)
PDAudio-CF (& CF-to-pcmcia adaptor)
PocketPC 2002

I can offer the following updates on recording software:
Pocco's Wichita beta v1.02 records at 16/44.1 without any noticeable flaws.  I haven't done serious testing since I just installed it a few hours ago.  I ran a wmod-ua5 into it for a few recordings - seems to work ok and is very easy to use.

Gidluck's Live2496 latest release supports up to 24/48 - I just got it today and don't have the codes to enable recording yet.

More info to come over the next few days.

[goose]

Craig, would you post a pic of that set up when you have a chance?  I'm not familiar with all those gadgets...

[Brian Skalinder]

I run the following hardware/software configuration:
HP iPAQ 3835
HP dual pc card expansion pack
Toshiba 5GB pcmcia hard drive (FAT32, 32kb clusters)
PDAudio-CF (& CF-to-pcmcia adaptor)
PocketPC 2002

I can offer the following updates on recording software:
Pocco's Wichita beta v1.02 records at 16/44.1 without any noticeable flaws.  I haven't done serious testing since I just installed it a few hours ago.  I ran a wmod-ua5 into it for a few recordings - seems to work ok and is very easy to use.

Gidluck's Live2496 latest release supports up to 24/48 - I just got it today and don't have the codes to enable recording yet.

More info to come over the next few days.

Craig T - great info.  Keep it coming?  

[Simp-Dawg]

yeah man...you're gonna be our guinnea pig/info bitch for a while it seems!  good work though, keep us well informed on how this goes for you...thanks!

[cpclark]

i would love to see how this comes out, much cheaper than buying the new SD HD recorder thats gonna be 2000

[Craig T]

I've been playing around with Pocco's Wichita recording long segments of ambient noise with dpa 4061>wmod-ua5>pdaudio-cf>ipaq.  No hiccups after 30min, 45min, 60min...  The dreaded 2GB limit is in effect, so we'll see how long I can go at 16/44.1.  I did notice the record meters slow down after about 30min to the point where you can't really use them for anything other than knowing you have a signal (no peak level, just graphic bars with no metering, although the bar does turn red at the tip where 0 would be).  The elapsed time also started going by 2-3 seconds at a time instead of the normal 1 second.  I guess the processor &/or video RAM is getting taxed pretty heavily.  Upon playback, there is nothing wrong, just visual on the screen when recording.  You MUST turn off the screen when recording so you don't bump the stop or pause buttons (the screen will register bumps from your hands etc, not just the stylus).  You can also set the buttons on the iPaq to function as record/play/stop/pause/etc.  I don't recommend this as they are not protected and could be accidentally bumped.

You must format your memory for large clusters - I reformatted my 5GB Toshiba pcmcia hard drive to FAT32, 32KB clusters.  At 512 byte clusters, I got split-second gaps every 8 seconds.  Changing to 32KB clusters seems to have solved that problem.  (I used a "storage tools" utility to do this - free trial download, nothing in PocketPC 2002 would reformat the drive)
http://www.softwinter.com/storagetools.html

PHM registry editor also was helpful when checking to make sure the PDAudio-CF drivers were properly loaded.
http://www.phm.lu/Products/PocketPC/RegEdit/

[dmonterisi]

nice report, craig, +t...re: the 2 gig limit...can you set it so that at say, 1.9 gigs it just creates a new file without any missed samples?

-damon

[Craig T]

nice report, craig, +t...re: the 2 gig limit...can you set it so that at say, 1.9 gigs it just creates a new file without any missed samples?

-damon

I don't think so.  I'll ask tech support if this is possible & if it isn't with this version, maybe they can add this to a future release.

[dmonterisi]

nice report, craig, +t...re: the 2 gig limit...can you set it so that at say, 1.9 gigs it just creates a new file without any missed samples?

-damon

I don't think so.  I'll ask tech support if this is possible & if it isn't with this version, maybe they can add this to a future release.

that would be really helpful because the second set of a lot of bands such as panic will go over 2 gigs at 24/96.  wavelab has that feature so that you just set whatever size you want it to start a new file and it does.  no samples lost, you can then split the files how you want and add the section that became the new file to the end of the previous section without missing anything.

[Brian Skalinder]

nice report, craig, +t...re: the 2 gig limit...can you set it so that at say, 1.9 gigs it just creates a new file without any missed samples?

-damon

I don't think so.  I'll ask tech support if this is possible & if it isn't with this version, maybe they can add this to a future release.

Another option to solve this problem:  ability to record L and R channels to independent files.

[MattD]

It's still not a solution in 24/96 land, though. I frequently go over 2 GB/file even when recording L and R to separate tracks. It's only 2 hours at 24/96.

Proprietary format or seamlessly creating a new track is what I'd recommend to the developer.

-Matt

[Brian Skalinder]

I frequently go over 2 GB/file even when recording L and R to separate tracks. It's only 2 hours at 24/96.

I hadn't done the math, Matt, but you're absolutely right - good point!  In that case, I agree with you completely - proprietary format or seamlessly creating a new track is the only way to go.

[Craig T]

Pocco is trying to implement automatic seamless file breaks in a future release of Wichita.  They also said a few more weeks until they expect to support 24bit recording.

Just got Live2496 up and running.  First impressions: just as easy to use as Wichita (GUI looks like a tape recorder with large "buttons"), better metering (has 0 to -80 markings), so far I don't see any way to create "shortcut buttons" with the iPaq buttons and 4-way pad.

[Lee]

I will believe it when I see a working model.

you said it

[zhianosatch]

This is SO not a simple process... I hope it works and gets streamlined.

[Craig T]

More Live2496 info:

Current release of Live2496 does not have working meters, and you have to change recording file destination in a text file before starting the program.  Future releases will add more GUI menus to eliminate the need to edit the text file.

Gidluck mentioned he would like to implement an auto-splitting feature, but does not think its possible due to the PDA hardware limitations (something about creating a "logjam" in the PDA bus).  Live2496 will write wav files >2GB, but some sort of recovery application would have to be used so they can be read into editing or playback software.  I'm not sure how this works, but I'll give it a try...

[Craig T]

First 2 recordings at 16/44.1 were successful.  Got DMB 9/5 and Gongzilla 9/7.  I used Pocco's Wichita for both recordings since I had a few problems with my Live2496 tests.  I just received a new version of Live2496, so hopefully the problems are fixed.  The level meters on the current version of Wichita are almost useless - as the recording gets longer the refresh rate slows way down to about 5-7 seconds.  Same with the elapsed time counter - at an hour it was counting by 5 second intervals.  Doesn't affect the recording, just the display while recording.

[Craig T]

First 2 recordings at 16/44.1 were successful.  Got DMB 9/5 and Gongzilla 9/7

Well, not quite.  I found a 'skip' about 20 minutes into the DMB show - sounds like the recording stopped for a fraction of a second and then started up again.

[JasonSobel]

a few questions for you, because I'm interesting in this for 24/96...

have you used Len's PDAudio Recorder (Linux) software.  The site says that this comes with the PDAudio-CF card along with all the drivers necessary.  The site also states:

"Core Sound's PDAudio Recorder (Linux) provides detailed signal level metering, and elapsed time and storage media capacity monitoring. It preserves the audio data file in the case of lost power and it records audio files larger than 2 GB."

How does it take care of the 2gig problem?  Does it seamlessly start a new track?  if not, how does Len deal with it?  The "preserves the audio data file" part seems like a great idea too.  In case the battery dies, at least the whole file is not lost.  Do the other software packages do this as well.

and you mention a skip in one of your recordings about 20 minutes in.  Was that the only one that you've noticed so far?  Did you also run DAT?  Have you done any bit-accuracy tests on the files?

I'm sure some of those questions would be good to ask Len as well, but I'd like to hear your opinion on things as well.

Thanks,
Jason

[Craig T]

Update 9/19

Gidluck's Live2496 seems solid at 16/44.1 & 48, level meters have also been added for 16bit.  I'm getting dropped samples in 24/48 resulting in what appears to be white noise.  I've been informed I may not be able to do 24bit recording with the iPaq 3835 by Gidluck, but this conflicts with what I was orginally told by Len at Core Sound when I was shopping for iPaqs.  I'm going to try some performance tweaks to solve the dropped sample problem.

Pocco's Wichita has been working well in 16/44.1 with the PDAudio-CF although I haven't received any software updates since my last report.  I still haven't been able to recreate the dropped sample that I got during one recording.  I tried recording a lecture from the built-in microphone of the iPaq and got a poor quality recording, sounding kind of like dropped samples, with regularly spaced dropouts in the sound (I had the auto-gain feature turned off and mic gain set to +0).

[zhianosatch]

 

[cpclark]

i hope the bugs get worked out, im liking this more and more everyday