[AG-TECH] Echo Cancellation

[shudo at ni.aist.go.jp]

Markus and all,

> >  Just wondering what your analysis involved and what results you saw?  I have used both, but never really tried to compare the two.
>
> We had an acoustic engineer work with us on our first 3 nodes (and others),

> But basically it was things like larger buffers, faster CPU, more complex filtering capabilities, more flexible multi-pathing, complex digital I/O - and at about 30-50% less cost (he wasn't the supplier ;-) ). 

I have a comparing table of Polycom and ClearOne products.
The table has many rows and significant and easy-to-understand items
are as follows:

  tail time (maxmum latency the product handles)
  Polycom EF2210/2211  270 ms
  Polycom EF2241       270 ms
  Polycom EF2280       270 ms
  ClearOne XAP800      128 ms
  ClearOne XAP400      128 ms

  convergence time
  Polycom EF2210/2211  40 dB/sec
  Polycom EF2241       40 dB/sec
  Polycom EF2280       40 dB/sec
  ClearOne XAP800      30 dB/sec
  ClearOne XAP400      30 dB/sec

I got this table from a reseller of Polycom products, so it should be
favorable to Polycom.  But I am sure those are different somewhat.
We need a neutral comparisons.

As I have written once, the only reason many of us bought ClearOne
(Gentner) products is reference to ClearOne products in the recipe of
AG nodes provided by ANL people.
It's desirable that the recipes are enhanced but it's hard to
gain experience on using many kinds of echo cancellers because they
are pretty expesinve.

  Kazuyuki Shudo
  Grid Technology Research Center
  National Institute of Advanced Industrial Science and Technology (AIST)