Closed GoogleCodeExporter closed 9 years ago
Please attach first and second audio files
Original comment by oifa.yul...@gmail.com
on 22 Oct 2012 at 6:40
I have attached the two audio files.
Original comment by quintana...@gmail.com
on 22 Oct 2012 at 7:39
Attachments:
Please take a look at :
https://ccrma.stanford.edu/courses/422/projects/WaveFormat/
wav file format has 44 bytes header and then data chunk
this means that data in file should have fileLength-44 bytes
in your case its 17726-44=17682
and 65870-44=65826
for pcm codec frame length is 320 bytes
for a/u law its 160 bytes
therefore for both you dont have an integer number of frames in both files
which can lead to adding trailing junk in last frame of each file
Example of files stored by mms :
-rw-r--r-- 1 root root 44 2012-10-09 16:18
0af09760121c11e2ae28c46e53df2efe.w av
-rw-r--r-- 1 root root 81324 2012-10-22 20:22
630353d01c7511e29aabb66c8cd35abb.w av
-rw-r--r-- 1 root root 44 2012-10-09 16:30
c54a0410121d11e2a1b43fdb989df6fc.w av
Two files are empty , one is 81324 in legth. 81324-44=81280
81280/320=254 complete frames
Therefore its a file problem
Original comment by oifa.yul...@gmail.com
on 23 Oct 2012 at 8:34
As it appears while using alaw , ulaw files which have incomplete frames due to
decoding techniques used padding instead of 0 becomes very high negative values
which in turn causes clicking.
Need to add different padding depending on codec used in wav file
Original comment by oifa.yul...@gmail.com
on 23 Oct 2012 at 4:05
Hi Yulian,
I apologize for the delay in the feedback. I took the actions that you and I
discussed earlier today but the peak is still in the the stream. I have
attached the file with changes so you can verify that I didn't misunderstand
you and a tshark capture of the call. Please let me know if there is anything
else I can do to help.
Best Regards,
Tom
Original comment by quintana...@gmail.com
on 24 Oct 2012 at 2:47
Attachments:
[deleted comment]
1) Please verify that you are testing against correct version ( with updates )
2) Try this one
Original comment by oifa.yul...@gmail.com
on 24 Oct 2012 at 7:55
Attachments:
Hi Yulian,
I re-compiled using the provided WavTrackImpl.java and when I run the tests the
peak is still present. I have attached the capture file.
Best Regards,
Tom
Original comment by quintana...@gmail.com
on 24 Oct 2012 at 2:39
Attachments:
Hi Yulian,
The time of dip between the old capture and the new capture is significantly
smaller. I have taken screenshots of the sine waves where you can also find the
timing on the top ruler (time is measured in seconds, the old capture has a
10ms dip the new capture has much less but still perceivable). Also, it seems
there is a dip in the beginning, in the picture titled captures-front-dip.png I
have zoomed in to the sine wave so you can see the dip. I have attached all the
relevant files. I hope this helps.
Best Regards,
Tom
Original comment by quintana...@gmail.com
on 25 Oct 2012 at 12:04
Attachments:
Ok lets summarize
Indeed the problem occured due to 2 problems
1) Padding was invalid in case of G711
2) There may be additional bytes in header for any other coded then pcm.
Please confirm that currently it works.
Original comment by oifa.yul...@gmail.com
on 25 Oct 2012 at 9:50
Attachments:
Please check PCM , ALAW and ULAW file formats , 10x
Original comment by oifa.yul...@gmail.com
on 25 Oct 2012 at 9:51
Hi Yulian,
I have attached the pictures with the results. It seems the only thing left is
the dip at the beginning of each file.
Best Regards,
Tom
Original comment by quintana...@gmail.com
on 25 Oct 2012 at 2:32
Attachments:
Hi Yulian,
After the suggested fixes everything is working perfect for wav files
containing alaw, ulaw, and PCM. I have attached all the relevant files.
Best Regards,
Tom
Original comment by quintana...@gmail.com
on 26 Oct 2012 at 2:05
Attachments:
Modified by revision cd6f061a87a7
Waiting for mss team approval
Original comment by oifa.yul...@gmail.com
on 27 Oct 2012 at 7:26
Original comment by oifa.yul...@gmail.com
on 27 Oct 2012 at 10:04
Closed
Original comment by oifa.yul...@gmail.com
on 29 Oct 2012 at 2:14
Original issue reported on code.google.com by
quintana...@gmail.com
on 22 Oct 2012 at 6:21Attachments: