[Christina S]

Also, when I run make_initial_phone_labs, Festival raises the following error about 100 utterances in… What could the problem be?

SIOD ERROR: not a number

BACKTRACE:
0: (#<SUBR(5) Unilex_Word> (quote #<Utterance 0x7ff4f2703230>))
1: (apply_method (quote Word_Method) utt)
2: (Word utt)
3: (utt.synth_toSegment_text
(eval (list (quote Utterance) (quote Text) text)))
4: (if
(eq? (cadr l) (quote apml))
(do_synth_for_htk_apml (car l) ofd)
…)
5: (f (car l2))
6: (cons (f (car l2)) r)
7: (set! r (cons (f (car l2)) r))
8: (while l2 (set! r (cons (f (car l2)) r)) (set! l2 (cdr l2)))
9: (mapcar
(lambda
(l)
(format t “%s
” (car l))
…)
p)
10: (make_initial_phone_labs
“christina_utts.data”
“utts_coverage.mlf”

I’ve tried to add words to my lexicon so I think it might be something to do with the formatting of it but I can’t see any issues.

Thanks for your help

[Christina S]

This is the output from soxi:

Input File : ‘recordings/arctic_a0001.wav’
Channels : 2
Sample Rate : 48000
Precision : 24-bit
Duration : 00:00:03.84 = 184320 samples ~ 288 CDDA sectors
File Size : 1.11M
Bit Rate : 2.31M
Sample Encoding: 24-bit Signed Integer PCM

I’ve attached a sample file too.

Attachments:

You must be logged in to view attached files.

[Christina S]

Thanks Simon. I realised I had also recorded with 2 channels so needed to discard 1. Now I’m running something like:

bash$ sox recordings/arctic_a0001.wav -b16 -r 16k wav/arctic_a0001.wav remix 1

Now it seems like the files in the wav directory are much bigger (68.7 MB each) compared to in the recordings directory (1.1 MB each). Where am I going wrong? I expected them to be much smaller in the wav dir.

This is causing problems when I get to do_alignment. It’s throwing an AllocBlock error.

[Author]

Posts