Obviously, it would be illogical to use a non-blocking platform and then sabotage its main function using a process that is better left to the client processor for analysis - why pay 100,000 audio users when 100,000 users have already paid for the last processor that does nothing on the actual deed ...
HOWEVER, it may be useful to store intellectual property elements on your backend - waveforms and algorithms of certain audio nodes that were developed at high cost, but they are very easy to copy if they were on the client side .... then it will have it makes sense to keep this component on the server side (but whether it will be a node solution or not, this is another argument)
source share