I'm attempting to save a short streaming audio clip to an mp3 file. I set Capture with alsamixer to 100% and then used the command.

The resultant mp3 file is 544k and identifies as mp3 using the file command. Unfortunately, playing the file with ffplay or gmplayer produces no sound.

I don't see a source there. So arecord is using the default source. What is the default source?

What are you trying to do? Capture the audio coming through your sound device? Or is this a audio file streamed from the internet?

Give more info. I suspect ffmpeg is what you are going to want to use. Or dump the output from a media player.

1 members found this post helpful.

Perhaps I don't understand your question. The source is streaming audio. I simply want to capture to a file what I'm hearing through my speaker.

I'm certainly willing to use another method, ffmpeg or other. However, as I mentioned, the arecord command produces am mp3 file.

1 members found this post helpful.

You want to capture what is coming through your sound hardware.

1st question for you to ask yourself. Does your hardware allow that or is it disabled. Some sound hardware has stream capture disabled. If it does then you'll need to make a virtual sound loopback device. Search LQ for that, it has been answered scores of times.

If your hardware allows that , or you have made a virtual loopback device in your ~/.asoundrc or using pulse.

Then it's simple enough to capture stream coming through sound device.
Example:

1 members found this post helpful.

The word "streaming" seams to signify something coming from the network, but apparently you don't mean it like that:
If you are using plain ALSA that is not so simple.
Should be easier with pulseaudio or jackaudio.
I hear it's possible with plain ALSA too, but requires the sort of superior knowledge (and temporary shifting of static configuration) I do not posses.

In other words: if it works now, good. please share your solution. If it doesn't, you'll have to ask yourself if you cannot achieve the same thing in a different manner (maybe use Audacity to layer two audio tracks?) or install one of the sound architectures mentioned.