In general you should not start a batch analysis while performing live. Running batch analysis tasks is supposed to be done offline. I remember that Serato entirely disabled batch analysis while a controller was connected.
The new multi-threaded analysis pauses batch analysis while an ad-hoc analysis for a track loaded into a deck is performed:
Currently all cores are utilized for batch analysis if no ad-hoc analysis is running. As an extension we might add an option to permanently exclude a fixed amount of the available cores (e.g. 1/4, 1/2 or at least 1) from batch analysis on a multi-core system. The default setting will use all available cores. Changing this setting will not affect a running batch analysis. It just serves as a stopgap for those who didn't do their homework ;)
Adding an adaptive re-scheduling algorithm for batch analysis depending on processor utilization would be too complex and expensive for this purpose.
Please note, that I don't plan any extensions or changes on the current implementation until it has been merged to master! This issue might track such an extension for future development.
In general you should not start a batch analysis while performing live. Running batch analysis tasks is supposed to be done offline. I remember that Serato entirely disabled batch analysis while a controller was connected.
The new multi-threaded analysis pauses batch analysis while an ad-hoc analysis for a track loaded into a deck is performed:
https:/ /bugs.launchpad .net/mixxx/ +bug/1641153 /github. com/mixxxdj/ mixxx/pull/ 1624
https:/
Currently all cores are utilized for batch analysis if no ad-hoc analysis is running. As an extension we might add an option to permanently exclude a fixed amount of the available cores (e.g. 1/4, 1/2 or at least 1) from batch analysis on a multi-core system. The default setting will use all available cores. Changing this setting will not affect a running batch analysis. It just serves as a stopgap for those who didn't do their homework ;)
Adding an adaptive re-scheduling algorithm for batch analysis depending on processor utilization would be too complex and expensive for this purpose.
Please note, that I don't plan any extensions or changes on the current implementation until it has been merged to master! This issue might track such an extension for future development.