Hi,
I believe there is an issue with cont_seg.c when using it to segment audio file. All the split audio files are of zero size. I am using the sphinxbase code from the trunk (probably from 2 weeks ago). The issue is as below I believe. Correct me if I am wrong.
Inside segment_audio, we make a call to fe_process_frame_ext. This function has a parameter for voiced_buf and voiced_nsamps. The zero file is because the voice_nsamps is always zero. It is zero because voiced_buf is set to NULL in segment_audio and if it is then fe_write_frame does not copy the pcm data internally.
May be the issue is known or I am missing something. Kindly let me know.
Thanks
Ram
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I believe there is an issue with cont_seg.c when using it to segment audio file. All the split audio files are of zero size. I am using the sphinxbase code from the trunk (probably from 2 weeks ago). The issue is as below I believe. Correct me if I am wrong.
Inside segment_audio, we make a call to fe_process_frame_ext. This function has a parameter for voiced_buf and voiced_nsamps. The zero file is because the voice_nsamps is always zero. It is zero because voiced_buf is set to NULL in segment_audio and if it is then fe_write_frame does not copy the pcm data internally.
May be the issue is known or I am missing something. Kindly let me know.
Thanks
Ram
Hi,
Is this a known issue or am I missing something. Kindly let me know
Thanks
Ram
Sorry, this is a known issue.
Thanks for the reply. Is there an ETA (main or branch release) for the fix? Kindly let me know.
Tracked at https://sourceforge.net/p/cmusphinx/bugs/425/