Rockbox: Clip v2

[marc2003]

i've created a USB enabled build for the clip v2 here (no other patches).

http://cid-649d3bfeaf541fbb.office.l...clipv2-usb.zip

i haven't tested much but it appears to be working ok. i've transferred a few files and it shows up ok in windows.

you should see a USB logo like this when plugged in....



i'm using the latest official release bootloader which hasn't been updated since last july.

assuming you use windows, i'd restart after copying over the new build. mine went a bit funny not recognising the device at first but seems fine now.

[CrowdAds]

[marc2003]

i've noticed a few usb related things on the rockbox site going through SVN so i've updated my clip v2 with usb enabled build.....

http://cid-649d3bfeaf541fbb.office.l...clipv2-usb.zip

(not that anybody reads this thread... )

[shaggyZ]

I just read it, thanks my clip version 2 was the only sansa I had that didn't have rockbox with usb. Now I am complete.

[marc2003]

use the same link above if you want it updated with SVN at the time of this post..

[shaggyZ]

awesome, just what i was looking for.

[keyb_gr]

When I tried a 10 kHz test tone in both 44.1 kHz and 48 kHz samples rates on my ClipV2 (as MP3), the 48 kHz one showed an audible het which does not belong there. Can anyone reproduce this (3.7.1 here)?

Also, any news on these annoying ticking noises? They're just loud enough to be audible on a HD590, and for anything more sensitive an attenuator is necessary. Refreshing the database or running the FFT plugin generates even more noise.

My Clip also trips over its own feet once in a while, locking up upon a keypress, which I guess is due to the funny multiplexing of input and output. I'm not entirely sure whether the RAM is 100% OK, would there be any way to test that?

[saratoga]

Originally Posted by keyb_gr

When I tried a 10 kHz test tone in both 44.1 kHz and 48 kHz samples rates on my ClipV2 (as MP3), the 48 kHz one showed an audible het which does not belong there. Can anyone reproduce this (3.7.1 here)?

Not sure I understand your question, but if you're running at 48kHz you'll be resampled via linear interpolation to 44.1k.

[keyb_gr]

Originally Posted by saratoga

Not sure I understand your question, but if you're running at 48kHz you'll be resampled via linear interpolation to 44.1k.

Linear interpolation? Oh crap. That explains it then...

But didn't the Clip support 48 kHz natively at some point, back in the day when the pitch bug was still there?

Whatever, not like I'd need it, but it's good to know in any case.

[saratoga]

Originally Posted by keyb_gr

Linear interpolation? Oh crap. That explains it then...

But didn't the Clip support 48 kHz natively at some point, back in the day when the pitch bug was still there?

The OF does, but not rockbox, since our DSP engine assumes 44.1khz.

[keyb_gr]

Ah, OK - probably was in the OF then. That might be worth changing in the future, depending on whatever the device in question supports (the Clip with its PLL setup is quite flexible here). I imagine devices with a color screen would get more 48 kHz material, and linear interpolation really isn't good.

Speaking of device dependency... how is the default selection of fonts chosen? I'm asking because those supplied with the ClipV2 build at least aren't the very most useful. You get either utterly tiny (8) or fairly large (12+) sizes. I'd suggest throwing out the huge, useless 16-GNU-Unifont and including 10 and 11 Nimbus instead.
To my eyes, 11 Nimbus makes a good compromise between readability and amount of information.

Actually all I wanted to say here is thanks to the folks who got the database working again, whatever the problem was. r29642 works fine.

[eboyer93]

You can easily install the font package.

[keyb_gr]

Possible improvement? Ever since I upgraded from 3.7.1 to a recent build, the ticking background noises seem to have dropped in level. I have been consistently unable to detect them on my HD590 (approx. 107 dB SPL / 1 Vrms), even when listening very hard in a quiet environment. Previously I could pick them up when concentrating on them. (They're still readily audible on K26P and SE420.)

Oh, and yesterday I traced problems with some files exhibiting repeatable short skips to them being corrupted (nominally the same but not bit-identical). It turned out that little more than a dozen were affected, out of about 2000. While I can't exclude a bad crash with FS corruption in 3.7.1 or a brief try of 3.8 as a potential cause, the primary suspect is bit rot in the flash memory. This is MLC flash after all. (My 8 gigger is 2 years old now and has been notoriously full for a long time. I think I already hit a flaky area with the Rockbox binary in the past, too.)

So I'd say it can't hurt to check your files once in a while. The classic Total Commander's file synch worked fine here.

[orby]

i notice no improvement using the latest build r29686.
am using very sensitive phones i must admit.

i posted here as well

[funman]

Can someone with a Clipv2 test this patched firmware (unzip it first) ?

It contains a USB-enabled bootloader, version 4.0.

To test the USB mode: power off, then press center while plugging the Clipv2 on USB.

If someone can confirm that it works we can make new bootloaders release for all the Sansa AMSv2 (Clipv2, Clip+, Clip Zip and Fuzev2)

Thanks

[keyb_gr]

I'll give it a shot.

EDIT: Looks like it works fine. Drive was installed and I can access stuff.

EDIT^2: This really made me appreciate RB's boot times again. r29666 on this ClipV2 barely needs 2 seconds to boot. (My Clip+ with µSD in needs 3 secs or so.) It's about 5..6 for the OF. My old iAudio G3 even needs more than 8 seconds, which feels like an eternity - and that wasn't slow at the time.

[skip252]

Gave this a shot and

1. Connect when powered off- Connects in Rockbox USB.
2. Connect and press |<<= - Connects in sandisk USB.
3. Connect when powered off and holding center button - Connects and shows message "Bootloader USB mode"

Data transfers in all 3 modes. I've tried connecting each way at least 10 times and it's connected each time.

That's on Vista and Win7. I'm not near my Linux system to try it there.