JackOfAll wrote: 
> KERNEL-WAND-3.0.35-CM34+
> > 
Code:
--------------------
  >   > 
  > [  252.698435] spdif_set_sample_rate wanted spdif clock rate = 64915200     
   
  > [  252.704802] spdif_set_sample_rate got spdif clock rate    = 64962406     
   
  > [  252.711161] set sample rate to 44100                                  
  > 
  > [  397.592835] spdif_set_sample_rate wanted spdif clock rate = 129830400    
   
  > [  397.599288] spdif_set_sample_rate got spdif clock rate    = 113684210 
  > [  397.605734] set sample rate to 88200
  > 
  > [  517.848749] spdif_set_sample_rate wanted spdif clock rate = 259660800 
  > [  517.855204] spdif_set_sample_rate got spdif clock rate    = 227368421
  > [  517.861648] set sample rate to 176400
  > 
  > [  340.299932] spdif_set_sample_rate wanted spdif clock rate = 227328000
  > [  340.306386] spdif_set_sample_rate got spdif clock rate    = 227368421
  > [  340.312831] set sample rate to 96000
  > 
  > [  576.823473] spdif_set_sample_rate wanted spdif clock rate = 454656000
  > [  576.829925] spdif_set_sample_rate got spdif clock rate    = 454736842
  > [  576.836369] set sample rate to 192000
  > 
--------------------
> > 

Looking at the errors here - the 88.2 and 176.4 cases are off by 12%?? 
I think we need to try some other values for the divider in the spdif
block in these cases.  I agree the patch is what I would have done
though...  I think we need to sweep through the values for the divider
and find cases where the actual clock is closest to the desired clock
for these two values.  [my theory is the 23 and 37 were picked on this
basis as I can't derive them via any other way!]  

Perhaps John can help us measure the actual achived frequencies.


------------------------------------------------------------------------
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=98544

_______________________________________________
unix mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/unix

Reply via email to