Clip+ Pitch and Playback Issues (Rockbox Solution?)

I am currently in the process of acquiring a new digital audio player. Being that I am on a tight budget the Clip+ has caught my attention. However, I have read that the original Clip has a low pitch/slow playback issue and that the Clip+ has either A) a slow playback/low pitch or B) a fast playback/high pitch issues. I am aware that the Clip+ can be Rockboxed. 

So here is my question: Does Rockboxing the Clip+ completely correct the playback issues? I am hoping that it not only fixes the playback speed to the proper speed but I also hope that it fixes the pitch to the proper pitch.

Your responses would be greatly appreciated. Thank you for taking the time to read my post.

At this point, the pitch variance even with the OF is generally undetectable to most people, apart from musicians and the ilk.  With Rockbox, even more so.

I am a musician and would probably notice. What I’d hope is that Rockboxing it completely solves the problem.

I get a kick that every couple of weeks or months someone, out of the blue; brings up this issue.  I dunno… Sounds like maybe a competitor that wants to bad-mouth the product.


I’m a musician too and have found the Clip+ (rockboxed) is close to pitch, doing nothing and can be adjusted (see http://download.rockbox.org/daily/manual/rockbox-sansaclipplus/rockbox-buildch4.html#x7-630004.3.3)).  If that isn’t good enough, spend a few more bucks and purchase a device guaranteed to be perfect pitch.  I have never read in any documentation supplied by Sandisk that their players are guarranteed to be perfect pitch. 

 I have  owned a clip+ for  a little over a week and that little player has amazing sound quality for the price. The pitch error with the original firmware is a mere 0.25% (how much is that in cents?), and the overwhelming majority of people don’t even notice that this thing has a pitch problem! I also “Rockboxed” it and found that the pitch error was reduced by a factor of 5 (to 0.05%). I will challenge anyone to detect any  audible pitch difference between this budget player  and a known audiophile grade expensive model like the cowon. I only know that there is still a pitch error because I can measure it with my  frequency counter. Otherwise I can’t hear any audible problem.So as far as I am concerned this pitch bug thing is very much a non issue. 

This may more properly belong in a Rockbox forum, but…:

Rockbox 3.7 for Clip+ appears to offer a pitch adjustment, but Rockbox manual says this tweak disappears when the unit is powered off…however the manual also says pitch setting resides in a bookmark, and it says somewhere else that a bookmark may be assigned to playlist, and it says somewhere else that a folder (aka directory) is one type of entity that may be assigned a bookmark, and finally it says somewhere that a bookmark may be made persistent (i.e. through power cycling).

So, it appears that pitch may be adjusted in a bookmark assigned to a root folder containing ALL music on the Clip+ internal memory and/or added micro SDHC card, and this pitch adjustment may be configured to persist through power off/on cycles unless user changes it, if I connect the dots correctly here.

Can anyone describe the exact details of how to use Rockbox to configure a “permanent” pitch adjustment for Clip+ that persists through power cycling, affecting ALL music files on the Clip+?  This would be THE solution to the Clip+ pitch bug.  Rockbox manual does not describe this adequately…

The 0.25% pitch error in Clip+ factory firmware/etc may be undetectable to some, but if there is a way to reduce the error to 0.05% for all music on the Clip+ or less then why not do that?  Sonic fidelity is my goal…“good enough” is NOT good enough if there is an economical way to get better.

@sandclip wrote:

This may more properly belong in a Rockbox forum, but…:

 

Rockbox 3.7 for Clip+ appears to offer a pitch adjustment, but Rockbox manual says this tweak disappears when the unit is powered off…however the manual also says pitch setting resides in a bookmark, and it says somewhere else that a bookmark may be assigned to playlist, and it says somewhere else that a folder (aka directory) is one type of entity that may be assigned a bookmark, and finally it says somewhere that a bookmark may be made persistent (i.e. through power cycling).

 

So, it appears that pitch may be adjusted in a bookmark assigned to a root folder containing ALL music on the Clip+ internal memory and/or added micro SDHC card, and this pitch adjustment may be configured to persist through power off/on cycles unless user changes it, if I connect the dots correctly here.

 

Can anyone describe the exact details of how to use Rockbox to configure a “permanent” pitch adjustment for Clip+ that persists through power cycling, affecting ALL music files on the Clip+?  This would be THE solution to the Clip+ pitch bug.  Rockbox manual does not describe this adequately…

 

The 0.25% pitch error in Clip+ factory firmware/etc may be undetectable to some, but if there is a way to reduce the error to 0.05% for all music on the Clip+ or less then why not do that?  Sonic fidelity is my goal…“good enough” is NOT good enough if there is an economical way to get better.

Re-read the above 2 posts.  Rockboxed, the pitch error drops to 0.05% or less.  There shouldn’t be any need to do anything further; meaning no getting into the pitch setting menu.

@fuze_owner_gb wrote:

 



Re-read the above 2 posts.  Rockboxed, the pitch error drops to 0.05% or less.  There shouldn’t be any need to do anything further; meaning no getting into the pitch setting menu.

 

Has that been confirmed (by measurement and/or comparison to factory Clip+ firmware)?  Can anyone post data/results?

@sandclip wrote:


@fuze_owner_gb wrote:

 



Re-read the above 2 posts.  Rockboxed, the pitch error drops to 0.05% or less.  There shouldn’t be any need to do anything further; meaning no getting into the pitch setting menu.

 


Has that been confirmed (by measurement and/or comparison to factory Clip+ firmware)?  Can anyone post data/results?

 

 

I’m sure it has (probably numerous times). Try the Rockbox site/forums.

@tapeworm wrote:


@sandclip wrote:


@fuze_owner_gb wrote:

 



Re-read the above 2 posts.  Rockboxed, the pitch error drops to 0.05% or less.  There shouldn’t be any need to do anything further; meaning no getting into the pitch setting menu.

 


Has that been confirmed (by measurement and/or comparison to factory Clip+ firmware)?  Can anyone post data/results?

 

 


I’m sure it has (probably numerous times). Try the Rockbox site/forums.

A search on the Rockbox forums for “Clip+ pitch” yields nothing on this.  Thanks.  Anybody else have anything to offer?

@sandclip wrote:

 


@tapeworm wrote:


@sandclip wrote:


@fuze_owner_gb wrote:

 



Re-read the above 2 posts.  Rockboxed, the pitch error drops to 0.05% or less.  There shouldn’t be any need to do anything further; meaning no getting into the pitch setting menu.

 


Has that been confirmed (by measurement and/or comparison to factory Clip+ firmware)?  Can anyone post data/results?

 

 


I’m sure it has (probably numerous times). Try the Rockbox site/forums.


A search on the Rockbox forums for “Clip+ pitch” yields nothing on this.  Thanks.  Anybody else have anything to offer?

 

 

This thread should be of some value: 

http://www.anythingbutipod.com/forum/showthread.php?t=48208

 

@fuze_owner_gb wrote:

<snip>

 

This thread should be of some value: 

http://www.anythingbutipod.com/forum/showthread.php?t=48208

 


 

Thanks, saw that one.  The stuff posted about frequency accuracies is over a year old, way out of date - there have been both SanDisk firmware and Rockbox updates since then.

The topic really has not been discussed in quite awhile, likely the reason you’re not seeing anything more recent.  I’m not aware that there have been any “advances” in tightening the frequency response further. 

@sandclip wrote:

 


fuze_owner-GB wrote:

<snip>

 

This thread should be of some value: 

http://www.anythingbutipod.com/forum/showthread.php?t=48208

 


 


Thanks, saw that one.  The stuff posted about frequency accuracies is over a year old, way out of date - there have been both SanDisk firmware and Rockbox updates since then.

 

If there’s nothing later than that, the same data would apply. Any firmware updates by either SanDisk or Rockbox are not going to change these readings detrimentally.

@tapeworm wrote:

 


@sandclip wrote:

 


@fuze_owner_gb wrote:

<snip>

 

This thread should be of some value: 

http://www.anythingbutipod.com/forum/showthread.php?t=48208

 


 


Thanks, saw that one.  The stuff posted about frequency accuracies is over a year old, way out of date - there have been both SanDisk firmware and Rockbox updates since then.

 


If there’s nothing later than that, the same data would apply. Any firmware updates by either SanDisk or Rockbox are not going to change these readings detrimentally.

 

Respectfully, I disagree:  Software regression testing is designed to uncover just this very sort of thing:  Reintroducing a previous fault or unwanted output, associated with a code revision.  Most of the time this would be an inadvertant (translation:  erroneous/unplanned) condition in the code, with an unwanted result, but not always.

I read somewhere in a forum discussing Rockbox that a previous revision had improved the AMS frequency error, but then a later revision - possibly the current release - had re-introduced it, according to one tester’s measurements.  The poster thought this had been done as part of attempting USB functionality that was subsequently removed pending further work.

When people make assumptions like “a later release is not going to cause ______ problem”, and don’t test to confirm, we end up with buggy software.

(“Bug” is a euphemism, BTW.  It should properly be call “fault”, or, if people pay money for it, “defect”.)

@sandclip wrote:

 


@tapeworm wrote:

 


@sandclip wrote:

 


@fuze_owner_gb wrote:

<snip>

 

This thread should be of some value: 

http://www.anythingbutipod.com/forum/showthread.php?t=48208

 


 


Thanks, saw that one.  The stuff posted about frequency accuracies is over a year old, way out of date - there have been both SanDisk firmware and Rockbox updates since then.

 


If there’s nothing later than that, the same data would apply. Any firmware updates by either SanDisk or Rockbox are not going to change these readings detrimentally.

 


 

Respectfully, I disagree:  Software regression testing is designed to uncover just this very sort of thing:  Reintroducing a previous fault or unwanted output, associated with a code revision.  Most of the time this would be an inadvertant (translation:  erroneous/unplanned) condition in the code, with an unwanted result, but not always.

 

I read somewhere in a forum discussing Rockbox that a previous revision had improved the AMS frequency error, but then a later revision - possibly the current release - had re-introduced it, according to one tester’s measurements.  The poster thought this had been done as part of attempting USB functionality that was subsequently removed pending further work.

 

When people make assumptions like “a later release is not going to cause ______ problem”, and don’t test to confirm, we end up with buggy software.

 

(“Bug” is a euphemism, BTW.  It should properly be call “fault”, or, if people pay money for it, “defect”.)

 

Well, I disagree with that.  It’s one thing if Sandisk made a claim that their players are perfect pitch, but they haven’t; so, the player is operating exactly as designed…even if it doesn’t meet your standards.


It is pretty much known that Sandisk isn’t going to spend vast amounts of time or resources on a “problem” that only a miniscule number of people can detect without test instruments.  That leads to the Rockbox team.  If you can’t find your answer here, I would suggest you go over there where there is a mechanism in place to contact the developers directly.

<Edit:  Client side “bugs” produced pile of wreckage.  Deleted by Sandclip.>

Can anyone post data regarding current Rockbox frequency error on the Clip+? (I know, I know - this is not a Rockbox forum - but the thread was started in this forum.)

  Anyone can do a simple test to roughly determine the pitch accuracy of a music player. You will need the nch tone generator or similar software to generate a steady sine wave tone. You will also need a frequency counter capable of at least 0.1hz  resolution. The Radio Shack 22-811 is ideal for this purpose: an inexpensive yet full featured multimeter that has a frequency counter useful for audio up to radio frequency (0.1hz-4MHZ).

  I used nch tone generator to generate a 200 hz sine wave as a wav file. I loaded this file on the sansa clip+ , and measured the frequency on playback coming off the pc sound card, the clip+ original firmware, and Rockbox. The output of the pc sound card was dead on 200.0 hz The Clip+ OF measured 200.5 hz approximately (some bobbling in the last digit).The Rockbox  player measured 200.1 hz, 1/5 th the error of the OF, which comes out as approximately 0.05%

  For more discussion on Rockbox  for the Clip+ and the pitch bug issue, see this link: http://www.hydrogenaudio.org/forums/lofiversion/index.php/t84643.html  Here is a quote from that link: (Rockbox)3.7 enables a special PLL clock providing an optimal ratio for 44.1kHz PCM playback on ASMv2 devices (Fuzev2, Clipv2, Clip+). AFAIK all those devices played 1.1% flat with rockbox, with 3.7 the pitch error is reduced to 0.04% (which is more precise than the original firmware).

@mroberts200 wrote:

  Anyone can do a simple test to roughly determine the pitch accuracy of a music player. You will need the nch tone generator or similar software to generate a steady sine wave tone. You will also need a frequency counter capable of at least 0.1hz  resolution. The Radio Shack 22-811 is ideal for this purpose: an inexpensive yet full featured multimeter that has a frequency counter useful for audio up to radio frequency (0.1hz-4MHZ).

  I used nch tone generator to generate a 200 hz sine wave as a wav file. I loaded this file on the sansa clip+ , and measured the frequency on playback coming off the pc sound card, the clip+ original firmware, and Rockbox. The output of the pc sound card was dead on 200.0 hz The Clip+ OF measured 200.5 hz approximately (some bobbling in the last digit).The Rockbox  player measured 200.1 hz, 1/5 th the error of the OF, which comes out as approximately 0.05%

  For more discussion on Rockbox  for the Clip+ and the pitch bug issue, see this link: http://www.hydrogenaudio.org/forums/lofiversion/index.php/t84643.html  Here is a quote from that link: (Rockbox)3.7 enables a special PLL clock providing an optimal ratio for 44.1kHz PCM playback on ASMv2 devices (Fuzev2, Clipv2, Clip+). AFAIK all those devices played 1.1% flat with rockbox, with 3.7 the pitch error is reduced to 0.04% (which is more precise than the original firmware).

Thanks mroberts200, that is really useful info.  It will take me a while to sift through that forum thread at hydrogenaudio.org.  I’d say that considering that these sorts of measurements are done on consumer-grade equipment, 0.05% error is within the range of measurement equipment error/uncertainty, so the Clip+ frequency accuracy can’t be determined any better than that (outside of an accredited lab).