Intel® Integrated Performance Primitives
Deliberate problems developing high-performance vision, signal, security, and storage applications.
Announcements
FPGA community forums and blogs on community.intel.com are migrating to the new Altera Community and are read-only. For urgent support needs during this transition, please visit the FPGA Design Resources page or contact an Altera Authorized Distributor.
6814 Discussions

Error - "Cannot find IPP_EC echo canceller".... What is this error?

suvidya2000
Beginner
1,471 Views
Hi i am new to Intel IPP. I tries running the sample program for Intel IPP Echo Canceller but i am getting an error saying "Cannot find IPP_EC echo canceller". All the environment variables are properly set. I am not able to run the sample program.Can anyone please help?
0 Kudos
13 Replies
ligoban
Beginner
1,471 Views

Hi.

What is the operational system?

What software did you use to compile IPP examples?

0 Kudos
suvidya2000
Beginner
1,471 Views
Hi Ligoban,
Thanks a lot for your reply. I am honoured. As of now, i am successful in getting some output from the Echo Canceller sample that is provided in Intel IPP speech coding sample. The input to that sample is supposed to be 3files as command-line arguments. i.e.
usc_ec[options]
where,
r-inPCMFile and s-inPCMFile files processed to perform echo path parameter estimation and estimated echo is subtracted from s-in sample and echo-free PCM data is written to s-outPCMFile file.

My problem here is i got the output but in a strange way. i.e. when i executed the program with proper parameters (where my r-inPCMFile and s-inPCMFile files were the outputs of a telephone conversation), i got s-outPCMFile with one of the voices missing in the conversation and that voice was replaced by a bad disturbance. Echo was never removed.

I would be glad if you help me on this. I use Fedora Core-III and gcc to compile IPP programs.


0 Kudos
suvidya2000
Beginner
1,471 Views
I need a solution on this as early as possible... Thanking you in anticipation..
0 Kudos
ligoban
Beginner
1,471 Views

Is your OS in the list of supported OS?

Another question - did you try to use different algorithms in your work?

[-a ]- specify an algorithm: fb (fullband), sb (subband) or sbf (subband-fast). Default: = sb.

Was the result the same withdifferent values of options?

Vladimir Dudnik, can you help in this question?

0 Kudos
suvidya2000
Beginner
1,471 Views
Yes i have tried all possible combinations but the intended echo is not getting subtracted from the input file. Can the following be the reason for it ? :: my r-inPCMFile and s-inPCMFile have same pitch (amplitude) of voice. The echo is supposed to be of a lesser pitch than the actual voice. But in my case the pitch of voices in the r-inPCMFile and the s-inPCMFile [THIS VOICE in s-inPCMFile IS TO BE SUBTRACTED FROM THE r-inPCMFile] are equal. Can this create any problem for the algorithm in detection of correct echo and/or removing it? (I think Acoustic Echo Cancellers differentiate 2 voices (sounds) on the basis of the amplitude -- NOT VERY SURE!!)
If not can you suggest some different files where i can test the echo removal?
Any mails regarding the same to me ( suvidya2000@gmail.com ) will be appreciated.

0 Kudos
Vyacheslav_Baranniko
New Contributor II
1,471 Views

Suvydia

Can you please attach your r-in and s-in files so I can play with them?

Make sence also to decrease R-IN level, thus to lessenthe echo replica inS-IN. Also, an echo level is desired to belower thana talker voice.That is, before r-in and s-in files recording, nice todecrease a speaker volume.

Slava

Cite: ITU G167 5.2.2.2 Signal levels:

The tests described in the sequel apply to level signals between [30 dBm0] and [10 dBm0] at Rin, and acoustic signals with levels 4.7 dBPa [ 10 dBPa] at Sin

Slava.

0 Kudos
suvidya2000
Beginner
1,471 Views
Could you please provide me your e-mail address so that i can send you those files...
0 Kudos
ligoban
Beginner
1,471 Views
Would you like to tell about the reason of the problem when it'll be solved?
0 Kudos
suvidya2000
Beginner
1,471 Views
Yes definitely. I am working in an IP based Voice Communication area where there is a need to eliminate an inherent echo from the transmitted and received voice data. If the sample gives some significant immunity from echo, then planning for implementation of Intel IPP in a regular course could be done. Hence the effort.
0 Kudos
ligoban
Beginner
1,471 Views

I used IPP vocoders in my work. We could share our IPPexperience if you don't mind.

My email: pseudor@yandex.ru

0 Kudos
suvidya2000
Beginner
1,471 Views
Hi ligoban,
I would be the happiest. I am sending you the files right away on your above mention e-mail address.. Which operating system do you work on?
0 Kudos
ligoban
Beginner
1,471 Views

I work under Windows XP Prof. and Linux ASP 11.

0 Kudos
ligoban
Beginner
1,471 Views
I wonder -what is the end of the story?
0 Kudos
Reply