Does anyone have experience so far using Antescofo~ with muted brass instruments? I’ve been doing some tests and find the suivi is perfect with open brass but sometimes less good when a mute is used. It makes sense I guess that the fundamental might be obscured and harder to track with a mute… But there’s probably also a simple solution to improve the performance…
If any special effect (such as muted brass) does not work with an “out of the box” Antescofo, my suggestion is to look at its spectrum (for example in AudioSculpt) and see what are the harmonics present. You can then set the harmonic lists used by Antescofo to detect pitch patterns using the “harmlist” message to Antescofo.
By default, Antescofo uses the first 10 harmonics. So the default “harmlist” is “harmlist 1 2 3 … 10”.
If you send a “harmlist 2 3 4 5” to Antescofo, then it’ll ignore the first fundamental during recognition.
Note: To do this dynamically inside a score, before you had to send a message (as an action) to a [receive] object to the Antescofo object itself (like the [r antescofo-mess] in the help file). Since the last version, you can do this dynamically inside a score using the antescofo::YYY where YYY is any method you can use inside the patcher for Antescofo! For example:
NOTE C4 1.0 Event1
antescofo::harmlist 2 3 4 5
NOTE D4 1.0 Event2
NOTE E4 1.0 Event3
antescofo::harmlist 1 2 3 4 5
In the above score, upon the recognition of Event1, Antescofo will automatically set the harmlist as indicated and on Event3 it’ll set another value for harmlist.
Just a follow up on this after some tests on files sent by Christopher, since I believe it’ll be of general interest:
The test was on a collection of Trumpet recordings of the same score using the following mutes: Cup mute, Straight Mute, Harmon Mute, Cleantone Mute, Buzz wow mute.
All examples would work using factory Antescofo parameters in HELP. No need to tweak any parameter.
Keep in mind: Always start your scores with a fake explicit silence! Something like
Note 0 1.0
This will reduce confusions of the recognition engine in the very beginning of a performance. I’ll try to address this internally in the next release.
Thank you Christopher and ofcourse, ping me here if anything goes wrong!! 
Thanks, Arshia. It’s funny — I always instinctive assumed the opposite: that Antescofo text files SHOULDN’T start with an explicit silence, that the follower would somehow wait for the first attack… Thanks for clearing this up.