Streaming Audio instead of recording Audio

Have set up a microphone to capture audio for a science experiment (insects that we are farming).
We want to monitor the audio (insect sounds) from another location. The researchers are in fact in another country and need access to the insect sounds.
We could use Teamviewer to log in to our laptop. The laptop has audacity installed and is directly connected to the microphone inside the insect enclosure. If we could do this then I am concerned that the only way to access the audio in the insect enclosure is to to press the record/generate option on in audacity.
I suspect this will fill the Hard drive quickly with the audio file. This is what I want to avoid.
Can I set up Audacity to display audio and capture audio without recording the audio? This will allow us to access and stream the audio (via Teamviewer) to a second location? ie. we can monitor and share microphone access to other researchers using Teamviewer and Audacity?
I just don’t want to record the audio (even without saving) as I don’t want to fill the cache or Hard drive. We can record the audio stream if required using Audacity on a second remote laptop.
Or is it best to use dedicated streaming software such as OBS to stream the audio to Youtube etc where we can monitor from a remote location?
Thank you for your assistance.

Audacity doesn’t have streaming capabilities. You may want to try OBS, Icecast or even plain FFMPEG for this instead.

(Streaming to YouTube probably would be useful to you anyway, as YouTube has a DVR function that lets you go back up to 8 hours. So if your little insect suddenly starts making dinosaur sounds, you have a bit of time to figure out what happened there :slight_smile: )

Thank you for your reply, much appreciated. I tried OBS to Youtube but quality doesn’t seem to be that good.May have to play around with various settings in OBS.
Thank you again

Very interesting application I must say, however don’t some insects produce sounds in the 20-50Khz range?
Crickets for example, are most sensitive to frequencies in the 2-5Khz range but some species also can detect/“hear”
up to tens of Khz.

This poses some potential problems depending on the insects you are studying, namely:

  1. Type of microphone used and it’s frequency response.
    For example, using a “normal” microphone would be useless for these and also recording bats that probably feed on them.
    (The insects that is, not the microphones. :smiley: ).

  2. Sample frequency of what ever is digitizing/recording/streaming these sounds.
    Remember good old Nyquist, max frequency that can be recorded is half (actually slightly less), than sample frequency.
    So if you are recording at a sample rate of 44.1 Khz, the max frequency that can be recorded would be 22.050 Khz.
    Is this wide enough to record the little critters?

This also applies to streaming and Youtube.
Most streams are heavily compressed and bandwidth limited and thus a lot of the required information in the audio
would probably be lost.
Another example may be those 'orrible hissing cockroaches, how high do their hisses go in frequency?


Found some results of recordings done of male cockroaches “whistling” at females:
Note the frequency range.
Although a majority seem to be under the 22 KHz limit, which is OK for local recordings,
however, Youtube seems to cut-off the audio (rather sharply) above 16 KHz which would render your streams useless.

The freq response below is the “norm” for audio on Youtube.
I don’t know if there are settings that would allow this to be modified.
Keep in mind that even if you set this higher on OBS (or what ever you are using),
Youtube will compress it down, so best double check.

Thinking about this more, I would tend to use ffmpeg (like LWinterberg wrote) but use it to record locally
(not stream it) and trigger it with a Python or Bash script to record at 44.1 KHz (or higher if your hardware supports it)
in say 30 minute chunks.

These chunks of audio would automatically be named with a time and date stamp and then uploaded
to a server using wput or even FTP, where the researchers could then download them in full resolution.

One advantage to using chunks, is that the researchers can start analyzing the audio and not have to wait for one big
recording (lasting several hours) to be completed.
Worst case scenario, they would be 30 minutes behind “real time”.

Another advantage, easier to edit and search smaller recordings versus one big one.

Yet another is, if a recording is corrupted or lost, it’s only 30 minutes worth instead of several hours or a whole day.

Once the 30 minutes are up, the recording is automatically terminated, a new one started and the cycle repeats itself again.
The “change over” time from one recording to another would only take a few seconds, so I doubt your insects would start
producing dinosaur sounds in that period.
The uploading to a remote server can be done in the background so no extra time is wasted.

Once a day, the local recordings could then be purged to make space for the next days recordings.
A copy is already on the remote server.

This would be ideal on a Linux machine as you then have access to things like cron-jobs, Python, Bash, wget, wput, etc.
Perhaps there are Windows equivalents/versions, I don’t know.