"Format not recognized" when input is a stream

[OptimisticLock]

Hi,

I am getting this on Intel Mac:

$ wget https://patrickdearteaga.com/audio/Chiptronical.ogg 
$ cat Chiptronical.ogg |  docker run --rm -v $PWD/:/data audiowmark add - b.ogg 1     
audiowmark: error opening -: Format not recognised.

This works fine:

docker run --rm -v $PWD/:/data audiowmark add Chiptronical.ogg   b.ogg 1 

[OptimisticLock]

Silly me. I forgot to do docker run -i. Please disregard.

[swesterfeld]

I can reproduce the problem that you cannot use stdout in this case, while it does work when you provide an output filename:

$ wget https://patrickdearteaga.com/audio/Chiptronical.ogg
$ cat Chiptronical.ogg | docker run -v $PWD:/data --rm -i audiowmark add - - f0 > /dev/null
audiowmark: error writing to -: unable to write wav format to standard out without input length information
$ cat Chiptronical.ogg | docker run -v $PWD:/data --rm -i audiowmark add - x.wav f0
Input:        -
Output:       x.wav
Message:      f0f0f0f0f0f0f0f0f0f0f0f0f0f0f0f0
Strength:     10

Time:         unknown
Sample Rate:  44100
Channels:     2
Data Blocks:  4

The problem is that this (pulled by Dockerfile) version of libsndfile returns SF_COUNT_MAX as number of frames for the Chiptronical.ogg file if the file is provided stdin, and audiowmark cannot really handle this case when writing to stdout.