qualitymanifest / timestampsdr Goto Github PK

Hello, maybe I am wrong but I found that timestampSDR works using the following algorithm.
When timestampSDR receives sound it creates file somename.wav and appends transmission based on timeout settings. When timeout exceed it creates another file somename2.wav and appends transmission.
I am writing tool for sync somename.wav with my storage. The idea is to search all *.wav in folder, sync them and delete. How can i know that somename.wav is already recorded and not in process of recording. On my opinion good idea would be to write file with some other name somename_in_process.wav and after timeout exceeds rename to somename.wav. In this case I would exactly know which file is already recorded.

Or maybe there is some other method to find which file already recorded ?

Hello,
I want to run rtl_fm with 8kHz, but timestampSDR needed 48kHz input. You advice to resample with -s <whatever> -r 48k options. But in this case I catch an error. Moreover keenerd wrote using -r option is also a very poor quality resampler. Use sox for resampling.

Have you tried to use sox to resample ? Can you give an advice how to make proper pipeline?

There should be an way to specify maxFiles in a way that removes the limit, perhaps -1

Since 100% of zero-padded messages are thrown out, fragments of separate transmissions are recorded right up next to each other, which can make them hard to understand sometimes.

It would be good to have a timeoutDelay option which allows the user to set a delay in milliseconds where zero-padded (silent) messages are briefly recorded after a transmission stops. After this timeout delay completes, it should call the main timeout. Of course, if a message comes in during the timeout delay, it should cancel the timeout.

While maxFiles can be used to limit how many files are recorded before ending the program, there currently is not an option to ensure that recordings don't take up too much storage space.

There should be an optional maxStorage setting that accepts a number in megabytes, which checks on some arbitrary intervals (maybe once a minute?) to make sure that the size of the recordings directory isn't too large. If so it should save the current file if there is one, and close the UDP listener.

https://www.npmjs.com/package/pre-commit

Didn't know this existed. Seems more useful than the current init-dev script since you can't forget to install it.

package.json script, /dev, and contributing.md should be updated accordingly.