SanDisk Sansa c200v2, m200v4, clipv1, clipv2, clip+, and fuzev2

[funman]

Some bench results:

FS#11246 definitely helps a little bit with Clipv1.

It seems to help as well with Fuzev2 although the bench didn't complete.
I have recording problems on all models too: Fuzev2/Clipv2/Clip+.

I don't know if the shorter runtime comes from µSD, only testing from internal storage can tell

Btw when I run my tests I don't have headphones plugged, so perhaps I'm cheating?

[EDIT] : Adding Fuzev1 benches:

• test with FS#11246 didn't complete (crash)
• I got a pretty large difference of 2 hours between 2 benches (both running unpatched SVN) , not sure why
• I didn't bench the OF on this Fuzev1, but will do since it seems our runtime is almost the same than OF or better according to SansaRuntime

[funman]

right, my battery test has just finshed. 14h21m in the end.

sansa clip8gb v2
r25799
ogg q5, 11 track album, avg 143kbps
cpu boosted
volume -14

test repeated with back light on: 9h39m

any other tests you want doing, give us a shout. i'm not much use for anything else. :/

Can you give a try to FS#11246 with CPU boosted?

With the 2nd patch you will need to update the bootloader from SVN, come on IRC if you don't feel confident with doing it

[marc2003]

i'll give it a bash. i'm just reading through the instructions on the wiki for patching.

edit: sorted. i'm just waiting for the battery to charge and then i'll begin benching.

[bilditup1]

Ah, I was going to ask if having headphones in affects battery life as well. I had mine in the whole time.

[cris]

To all the developers...huge thanks, progress in the last few days has been great.  Tried several builds in the last few days and for playing music, looking at photos etc things are very stable (few crashes with graphical eq).

Funman - I did not respond to your comments responding to my comment about the sound with no EQ on Rockbox compared to stock OS - have been busy.  I did a back to back listening test with the earlier build and the sound is different (to my ears) than the stock OS.  I won't go as far as saying I am an audiophile but I think I have a good pair of ears...without wanting to use adjectives, the sound on the stock OS was a bit 'warmer' and perhaps a bit more forward.  This does not mean it is better, just different. Of course I could be imagining it.  I haven't compared later builds because I don't care to boot into the original OS since I have got used to the Rockbox sound and the features are SOOOOO much better than the original OS.

Cris

[saratoga]

I did a back to back listening test with the earlier build and the sound is different (to my ears) than the stock OS.  I won't go as far as saying I am an audiophile but I think I have a good pair of ears...without wanting to use adjectives, the sound on the stock OS was a bit 'warmer' and perhaps a bit more forward.  This does not mean it is better, just different. Of course I could be imagining it.  I haven't compared later builds because I don't care to boot into the original OS since I have got used to the Rockbox sound and the features are SOOOOO much better than the original OS.

I did RMAA tests the other day and the two are identical, so you are probably imagining it.

[Timar]

Volume on the Clip+ is still considerably lower than on OF, despite this commit: http://svn.rockbox.org/viewvc.cgi?view=rev&revision=25831 (doesn't seem to have had any effect for me)

[funman]

should be fixed for good with r25890

[marc2003]

With the 2nd patch you will need to update the bootloader from SVN, come on IRC if you don't feel confident with doing it

oh FFS. i can't even read. i totally missed the part in bold. i only applied the first patch.  :/

having said that, i still got 15h22m which is an extra hour. i'll try the 2nd patch later.

[Timar]

should be fixed for good with r25890

Sorry, but I can't hear any difference with r25894. Volume is still considerably lower than on the OF (without the EU volume cap and with all replaygain, EQ settings etc. turned off on both)

UPDATE: while unboosted playback on r258614 - r25869 was perfectly stable for me, I get frequent crashes again with r25894 (including "temporary bricks"!)... there must have been some changes made between -69 and -94  which result in the same situation as in r25799.

[funman]

With the 2nd patch you will need to update the bootloader from SVN, come on IRC if you don't feel confident with doing it

oh FFS. i can't even read. i totally missed the part in bold. i only applied the first patch.  :/

having said that, i still got 15h22m which is an extra hour. i'll try the 2nd patch later.

Thanks, the 2 patches have the same functionality so it didn't really matter. I just wanted to be sure it gave a bit more runtime, I got 15hours+ too on my Clipv2.

I have committed this patch, so we'll release new bootloaders for Fuzev2/Clip+/Clipv2 sometime soon.

should be fixed for good with r25890

Sorry, but I can't hear any difference with r25894. Volume is still considerably lower than on the OF (without the EU volume cap and with all replaygain, EQ settings etc. turned off on both)

Hm then I don't know, my Clip+ had high volume before r25890.

I tried modifying each of the 8 bits of AUDIOSET2 register, and only having the bit 6 set had an effect.

Since this issue appeared with r25821, then the problem must be in AUDIOSET2 register. Try modifying this register until you can tell what affects volume.

firmware/driver/audio/as3514.c => look for AUDIOSET2, the code executed on as3525v2 is the code inside "#ifdef HAVE_AS3543".
There is only 8 bits, so only 255 possibilities. And since the value currently written is 0, only 254 possibilities left.

UPDATE: while unboosted playback on r258614 - r25869 was perfectly stable for me, I get frequent crashes again with r25894 (including "temporary bricks"!)... there must have been some changes made between -69 and -94  which result in the same situation as in r25799.

I can't see anything between those 2 revisions which would affect crashes.

Remember these crashes just seem to happen randomly so they could be affected by the speed of wind, the color of my horse (he's white btw), or what you ate the day before ...

So far people reported that messing with DSP parameters (pitch, equ) would make crashes much more likely, this will help when I try to fix them but I don't know when since i'm not sure what's going on down there in this chip..


EDIT:
YES WE KNOW THE VOLUME IS LOW
DON'T REPORT IT AGAIN AND AGAIN, READ THE PREVIOUS REPLIES BEFORE POSTING PLEASE

[cris]

Funman

Thanks for the note on the sound tests.

25904 Build - Had a crash today (hopefully didn't damage my speakers :-( ).  I was trying to change the stereo width on the fly and it just crashed...lesson learnt.  To all of you connecting your Rockbox fuze to your hifi, if you care about your speakers, turn down the volume before trying to change settings whilst it is playing. I've reverted to the previous build for now.

Cris

Edit:  found this old thread so it does not appear that this is an unknown issue, some other users have had it -  http://www.anythingbutipod.com/forum/showthread.php?t=17166

It reports exactly the same problem I had.  I also just extract the new build over the old one without deleting the old one first which one of the users seems to think might have caused the problem...not sure.

[7o9]

funman, with version r25906-100509 i see a lot more crashes on my Clip+ than before (the older more stable one i used was r25831-100505).

These are two recent ones:

Code: [Select]

Undefined instruction at 3004996C

Code: [Select]

Data abort
at 30049DEC
FSR 0x8
(domain 0, fault 8)
address 0xE3530044
The first one is just after starting rockbox and entering the main menu. The second one is while playing mp3's.

My bootloader is r25906-100509 as well.

Edit: another one while playing the same mp3's:

Code: [Select]

Undefined instruction at 3002496C

[PockyMaster]

Played off a class 4 4gb generic microSD card with 14 mp3s encoded in VBR in repeat.

Sorry for the crude graph.

[Xanikseo]

There has been an improvement in sound quality in the recent builds on the Fuze v2! It sounds warmer, tighter and somehow more natural, in fact very much like the OF.

EDIT: Does r25890 have anything to do with it? If so, then yes, funman, there is a change in quality, but *definitely* for the better!