[Dspforum] IIR lab question.

[Wenstrand, Doug S.]

ping-pong
 
though it definitely makes it a little more complicated, without the
ping-pong, the EDMA won't be much help, because even though you'll get
big blocks of data at a time, you won't have any time to deal with them.
 
For instance, lets say you get 1 second worth of data in an EDMA block.
You'd really like to take 1 full second to process that data, and write
the results somewhere else.  If you don't have ping-pong buffering, you
have to deal with that whole pile of data in only 1 sample time, since
the EDMA will start messing with your buffer then.
 
Similarly, on the output side, you can't overwrite the buffer of output
samples until they have all been used.
 
 

--------------------------------------- 
D. S. Wenstrand 
Johns Hopkins University 
Applied Physics Laboratory 
(240) 228-4282 

 


________________________________

	From: dspforum-bounces at echelonembedded.com
[mailto:dspforum-bounces at echelonembedded.com] On Behalf Of Pablo Pombo
	Sent: Tuesday, November 04, 2008 3:08 PM
	To: dspFORUM
	Subject: [Dspforum] IIR lab question.
	
	
	Do we need to implement ping-pong buffers in input and output or
single buffers are acceptable?
	
	
	
	
	
________________________________

	Date: Wed, 29 Oct 2008 10:45:19 -0400
	From: doug at echelonembedded.com
	To: dspforum at echelonembedded.com
	Subject: [Dspforum] Today's Class
	
	Just wanted to let everyone know our schedule for class today so
that they can plan accordingly.
	
	We have no assignments due this week, so no solutions will be
discussed, and we will have no quiz.  There will be a quiz next week
which should cover what you did for the STFT lab.
	Most of the class will be devoted to working on the EDMA lab,
which I understand is a little bit strange to debug. We will briefly
discuss the new assignment (IIR filter lab), but there will not be an
extended lecture on IIR filters.  
	
	Class Schedule:
	
	1. 10 minutes of examples of practical IIR filter performance
	2. Distribute/Discuss IIR Lab (if you don't come to class, don't
download until classtime, I'm modifying this)
	3. Off to lab to work on codec interface with EDMA
	
	I really want everyone to have a robust EDMA solution for the
codec interface, because we will be using it for most future labs.
Using EDMA and doing block processing (rather than sample by sample) is
clearly a preferable method that I want everyone to be quite comfortable
with.
	
	Summary :  If you are very comfortable with the EDMA lab, and
you are essentially finished, class tonight is probably something you
could miss.  I encourage anyone who is not comfortable to make sure that
they attend tonight so that we can devote some lab-time to helping you
out.
	
	We'll still meet in the classroom first.
	
	Doug
	
	
	
________________________________

	Invite your mail contacts to join your friends list with Windows
Live Spaces. It's easy! Try it!
<http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.
aspx&mkt=en-us>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://echelonembedded.com/pipermail/dspforum_echelonembedded.com/attachments/20081104/c175752d/attachment.html>