I like the way Synesthesia analyzes audio to audio react glsl and how it uses media to blend both (I couldn't get that look importing the Synesthesia signal and playing with blend modes), but magic allows me to such many things I really could't do my show nowadays without it!
For example, in a DJ-VJ show, I'll start the video while the dj stills crossfading, changing the tempo of the original song. When the song reaches bar X, for example, or any moment where I need the video to sync the audio, and tempo is already the one for which I made the video, he sends me a midi signal to switch (input selector) between VideoFile A to VideoFIleA with a specific start time. Maybe its not the best approach but it lets me edit sync videos for each song and then adapt them for the live show:

- Captura de Pantalla 2022-10-26 a las 13.36.56.png (358.06 KiB) Viewed 1751 times
On the other hand, I tried playing with NDI outputs, but since I use Magic Window to project the final signal I can't modify it without affecting what I send to Synesthesia.
As I said, maybe the way to go is to input both signals (Magic and Magic+Synesthesia) to a Resolume or anything that could merge both and let me control the output signal but can't afford it yet for this personal project. Maybe with several instances of Magic? Is it even posible?
Anyway many many thanks for your support. Maybe the solution is to rethink the setup instead of trying to introduce synesthesia 'by force'.
Hope I explained myself

!!