Project

General

Profile

Bug #2514

GSM HR encoding result does not match the reference

Added by fixeria 7 months ago.

Status:
New
Priority:
Normal
Assignee:
-
Target version:
-
Start date:
09/15/2017
Due date:
% Done:

0%


Description

Implementing the automake transcoding tests, I found that GSM HR related
encoding tests are always fail, while decoding from the reference files
is ok.

Regression tests summary:

1: procqueue                                       ok
2: io/pq_file ok
3: io/pq_rtp ok
4: conv/enc/amr_efr ok
5: conv/enc/gsm ok
6: conv/enc/racal_hr FAILED (testsuite.at:49)
7: conv/enc/racal_fr ok
8: conv/enc/racal_efr ok
9: conv/enc/ti_hr FAILED (testsuite.at:79)
10: conv/enc/ti_fr ok
11: conv/enc/ti_efr ok
12: conv/enc/rtp_efr ok
13: conv/enc/rtp_hr_etsi FAILED (testsuite.at:119)
14: conv/enc/rtp_hr_ietf FAILED (testsuite.at:129)
15: conv/dec/amr_efr ok
16: conv/dec/gsm ok
17: conv/dec/racal_hr ok
18: conv/dec/racal_fr ok
19: conv/dec/racal_efr ok
20: conv/dec/ti_hr ok
21: conv/dec/ti_fr ok
22: conv/dec/ti_efr ok
23: conv/dec/rtp_efr ok
24: conv/dec/rtp_hr_etsi ok
25: conv/dec/rtp_hr_ietf ok

The same problem can be discovered using the built-in shell
script named 'test_all_formats.sh'.

I have also attempted to compare the encoding results with
the reference files and found, that the mismatching bytes are
starting from 128th until 352 bytes.

Also available in: Atom PDF