[hpsdr] Audio distortions during zoom/pan panadapter

[Roger Rehr W3SZ]

I had the same thought as George, namely that the issue is one of the 
granularity of the windows default timer.

I had a similar issue years ago [2007] when playing around with the 
Delphi IDE, and wrote a bit about it then.
http://www.nitehawk.com/w3sz/windowstimingissues.htm

I believe that you can prove the diagnosis as follows:

There is a program named skedgran01.exe that runs on Windows versions up 
to and including Windows 10 that shows you the timer granularity for 
your system under default and high resolution timing.  This is the first 
link on my webpage.

You can run skedgran01.exe [testing the "DefaultTiming"] both when you 
are running DPC Checker and when you are not, and see if in fact the 
timer granularity is affected, as George suggested, by the DPC Checker.  
It is my assumption that it will be affected.

I just checked to make sure all of this works on Windows 10, and it does 
EXCEPT that DPC Checker won't run on Windows 10 here, even with 
compatibility modes, and so I need to leave the last check to you as 
outlined above.

The images below will give you an idea of what to look for [substitute 
"With DPC Checker" for "With YahooFinance"]:

See the skedgran01.exe result here for Default timing
With: http://www.nitehawk.com/w3sz/DefaultTimingWithYahooFinance.png
and
Without: http://www.nitehawk.com/w3sz/DefaultTimingBaseline.png
the Yahoo Finance page opened.

Compare this with high resolution timing result
With: http://www.nitehawk.com/w3sz/HighResolutionTimingWithYahooFinance.png
and
Without: http://www.nitehawk.com/w3sz/HighResolutionTimingBaseline.png
the Yahoo Finance page opened.

Also note that running Default Timing with the Yahoo Finance page opened 
is NOT equivalent to running with High Resolution Timing.

Please post your results back to the list!

Oh, the Code Project c++ program source given in the Code Project 
article linked to from my web page does compile on Win10 using Visual 
Studio 2013, and it runs perfectly.  But I think using the 
skedgran01.exe program as outlined above is a superior approach to 
figure out what is going on here.

Hope the above helps, and

73,

Roger Rehr
W3SZ

On 12/9/2015 9:02 AM, ghbyrkit at chartermi.net wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
>
>
>
> I could be wrong, but I suspect that possibly running the DPC checker 
> (and various other multimedia software) COULD switch the computer 
> timing resolution from about 16 ms to 1ms or so, or some other 'high 
> resolution' value that allows multimedia stuff to behave better 
> (faster, more responsive, less subject to latency issues)
>
>
> I'm not sure how to check this, but it could explain the results.
>
>
> 73,
>
> George K9TRV
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20151209/3878856f/attachment-0002.htm>