DSP Robotics Support

This is isolated part of the schematic.

Instead of live input (sure, works fine in reaper as a vst plugin) - I'd like to load a wave file.

Next - I'd like to select one of these drop list items (play carefully - sometimes there is a lag) and save fft processed wave into a new audio file.

Two more things. I'd like to trim some samples from the beginning of audio to save, and add some samples at the end, to make the processing as if the input file was c.a. 2 seconds longer. (I guess the same or similar problem with delay will be here).

Basically I'd like to make it as a batch process for all items, but first important thing is - how to make a single item.

Puff... this one is too heavy for me to make on my own.

Let me know if it works... it should

It seems to work fine, thank you.
I will need to do some more exact tests too.

What surprised me a little bit is the fact, that while savewave uses int16 format by default - the float32 format is saved faster. I guess it's because internal audio processing is 32bit and otherwise - it converts to 16bit additionally?

A quick question. Where is the "green node" that would indicate, the processing is done and file can be saved?

Ok, I think I found it.

Next open question.

Is there any way to indicate, that wave file was saved? So that next process can be started automatically after previous one was saved? Because I see there may be a conflict: while wave file is saved to disk, next process starts at the same time.

Or maybne it does not matter?

I think you can think of saving as instant process. Trigger that triggers saving may start next cycle. Test it...

Checked, everything works perfect. Thanks.
Years ago it would take me a lot of manual work using fft filter in cool edit.
Now it splits the file within minutes - with just a single click.

KG, one more question.

I'd like to connect that gain related part from other topic, after a single fft processing is done, so that each time (before before next fft process starts) I also get these gain values as well. I thought it should be somewhere like in the schematic, but now I'm confused (I'm still not sure how to work with this mem thing). How to make it? (use on 25+ seconds long audio)

Otherwise I would have to reload all files in separate thread.

//edit: if I connect it to analyzer to get the array directly - I'm getting values different than they are in DAW (gain module through separate file loading shows the same values as DAW, so the detection itself - works correctly).