Rockbox

Tasklist

FS#11304 - higher quality audio hw settings for AS3543 (Clipv2, Clip+ and Fuzev2)

Attached to Project: Rockbox
Opened by Marcoen Hirschberg (marcoen) - Friday, 21 May 2010, 14:28 GMT
Last edited by MichaelGiacomelli (saratoga) - Thursday, 28 March 2013, 22:04 GMT
Task Type Patches
Category Drivers
Status Closed
Assigned To No-one
Operating System Sansa AMSv2
Severity Low
Priority Normal
Reported Version Release 3.4
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

This patch changes some of the settings of the AS3543 (Clipv2, Clip+ and Fuzev2)

AUDIOSET2_HPH_QUALITY_LOW_POWER -> AUDIOSET2_HPH_QUALITY_HIGH

this is also set in the OF

AUDIOSET3_HP_BIAS_150

increases the bias current by 50%, should be equivalent to AUDIOSET3_IBR_HPH which is set for the AS3525. Supposedly useful to drive headphones with an impedance of 32 Ohm and more.

Would be great if these settings could be changed during audio playback to listen for any actual changes in quality :)
This task depends upon

Closed by  MichaelGiacomelli (saratoga)
Thursday, 28 March 2013, 22:04 GMT
Reason for closing:  Out of Date
Additional comments about closing:  As far as we can tell, enabling this has almost no effect on actual performance driving low impedance headphones (measured THD improved slightly but the difference is so small I'm not sure its real) but it does seem to reduce battery life by a fair amount (45 minute reduction was reported in one test). If someone does a new test showing a difference or that battery life is no longer negatively affected, we can reopen this but for now this patch seems like it should not be used.
Comment by MichaelGiacomelli (saratoga) - Friday, 21 May 2010, 15:34 GMT
This did not improve playback on the clipv2 for me. The tone I hear doesn't respond to volume changes, so changing the HPH settings is unlikely to have any effect on it. Increasing the bias current will probably reduce things like THD at high volume though.
Comment by Rafaël Carré (funman) - Friday, 21 May 2010, 15:37 GMT
Same here it doesn't change.

The clipv2 sound only happens on pcm playback, and it sounds like the sound i hear when using the buttons (playing or not), especially noticeable if you leave a button pressed.

I don't think I hear anything in the OF : it's more difficult to pay attention because the volume can't go very low.
Comment by Stephan Grossklass (keyb_gr) - Wednesday, 11 May 2011, 15:53 GMT
To clarify, this only has to do with nonlinear distortion levels in mixer and headphone amplifier. The AS3543 datasheet states the increased bias current should be helpful for loads of *less* than 32 ohm in particular, and I doubt they were going by audiophile standards when they wrote that. The distortion spectrum with a low-impedance load certainly suggests an underbiased output stage.

True, it would be useful to have an option to turn this on and off dynamically. "On" would be "audiophile mode" and "off" would be "power saving mode".
Comment by MichaelGiacomelli (saratoga) - Wednesday, 11 May 2011, 16:33 GMT
We tested it and it had no effect on distortion. What distortion are you offering to and dis this patch have a measurable effect?
Comment by sideral (sideral) - Monday, 04 July 2011, 16:25 GMT
The discussion over here
http://www.anythingbutipod.com/forum/showthread.php?t=53702&page=67
seems to indicate that this patch has a small, but measurable effect on audio quality, at the expense of a bit of battery life.

I'd be fine enabling this unconditionally. I believe this also would be an easier sell than an option for "audiophile" playback: I think the Rockbox project as a whole values playback quality slightly higher than extended battery life, whereas there's always push-back against adding yet another option.

Any other thoughts?
Comment by MichaelGiacomelli (saratoga) - Monday, 18 July 2011, 17:44 GMT
http://rmaa.elektrokrishna.com/Temp/Clip%2B%20Rockbox%20Default%20vs%20FS12111%2B11304.htm (thanks dfkt!)

Small improvement in THD leading to a grand total of 0.1 dB improvement in (already extremely low) noise floor. Unless there's literally no change in battery life I don't really see a point in having this. Such small changes in SNR are barely measurable, and certainly not audible. The AGC change in  FS#12111  is probably worth having though.
Comment by Bertrik Sikken (bertrik) - Tuesday, 20 September 2011, 18:41 GMT
A very slight improvement with this patch, see also http://forums.rockbox.org/index.php/topic,28848.0.html
Comment by Martin Sägmüller (dfkt) - Tuesday, 25 October 2011, 11:27 GMT
It seems both as3514-audio-hq.diff and pvdd1_avdd17.patch are broken against current SVN.
Comment by Martin Sägmüller (dfkt) - Tuesday, 25 October 2011, 11:58 GMT
Both patches resynced to 30835. Not sure the #endif in system-as3525.c is in the correct place. Please somebody check. :)
Comment by Bertrik Sikken (bertrik) - Tuesday, 25 October 2011, 12:10 GMT
The first part of the patch is unnecessary, because the pvdd1_avdd17.patch has already been applied to SVN.
The second part of the patch looks correct.
Comment by Martin Sägmüller (dfkt) - Tuesday, 25 October 2011, 12:22 GMT
Thank you for checking, Bertrik. Here's just the first patch resynced.

Loading...