## Preparing the Performance, Rehearsals¶

### Controlling Antescofo from AscoGraph¶

AscoGraph is very useful in a concert context, when you have to follow the follower to control the following part of Antescofo and to control what is launched in the Max patch.

The top of the Ascograph window has five buttons that correspond respectively to the Antescofo function previous event, stop, play, start and next event that you can launch also directly in your patch.

The play button simply sequences from the beginning of the score to the end, using given event timing and internal tempi, performing actions wherever available.

The Play string function (Menu/Transport/Play string) plays (evaluates) only the selected part of your score (restricted to a sequence of actions).

When Antescofo is in follower mode, and AscoGraph is up, the current position infered by the listening machine is outlined both in the piano roll and in the AscoGraph text editor.

### Moving in the Augmented Score¶

They are several commands (Max/PD messages that can be sent to Antescofo) that are very useful in rehearsals. The idea is to be able to take over the score deterministically (with correct electronics) if you are rehearsing for example from "Measure10" (a label in score) onwards and assuring that things are prepared (or not) once you start from there as if you have performed everything before. Here is a list of these commands:

#### Methods with no argument¶

• start: Sends initialization actions (before first event) and waits for follower

• play: Simulates the score (instrumental+electronics) from the beginning until the end or until STOP

#### Methods with label/symbol control¶

• startfromlabel label : Executes the score from current position to position corresponding to $label in accelerated more WITHOUT sending messages. Waits for follower (or user input) right before this position. • scrubtolabel label : Executes the score from current position to position corresponding to$label in accelerated more WITH sending messages up to (and not including) $label. And waits for follower (or user input). • playfromlabel label : Executes the score from current position to position corresponding to$label in accelerated more WITHOUT sending messages, then PLAYs (simulates) the score from thereon.

• playtolabel label : PLAY (simulate) score from current position up to $label. • gotolabel label : Position yourself on$label without doing anything else! (backward compatibility)

NOTE: Most above commands would use "current position" and not "from the beginning". In order to reset, one would precede one of the above with the STOP method.

#### Methods with float/beat-position control¶

Same as above, but taking a float value corresponding to cumulative beat-position in score (useful for interaction via notation editors such as AscoGraph and NoteAbility PRO). Namely: startfrombeat, scrubtobeat, playfrombeat, playtobeat, and gotobeat

#### Non-standard use of score navigation command¶

Like the others commands understood by Antescofo, these navigation messages can be issued from the Antescofo score itself. They can be used to develop open pieces, or to alter the flow of the performance on purpose.

### Tuning the Listening Machine¶

In the default audio observation mode, you need to define the analysis parameters which are window size, and hop size in sample numbers. Antescofo automatically buffers audio and does the required analysis. Your choice of window size and hopsize could alter the performance and computation of the system! Default value is [2048 1024]. For example, when following piano music, you might want to set window size to 4096 (for better spectral resolution). If you want more exact timing, reduce the hopsize (and pay more CPU).

#### Calibration¶

The quality of audio input is of utmost importance for any recognition system and Antescofo is no exception! If the audio input is too low or corrupted, Antescofo can no longer correctly detect events.

Antescofo comes with a built-in Calibration mode to help make sure there is enough audio level for the recognition system to work. You can use it by sending the "calibrate 1" message to Antescofo, start sending audio to Antescofo (no need to "start") and watch the values sent from the left-most outlet with a prefix "calibration".

Antescofo calibration values are always between 0.0 and 1.0 and can be interpreted as follows: Whenever there are "events" present in audio, these values should be relatively high (>0.75) and in the contrary (non-events such as silence, hall noise etc.) it should be relatively low (<0.5). It is best practice to test these conditions before any use. If these conditions do not meet, you have to tweak the input level (in Max or audio console) to Antescofo.

It is highly preferable in practice to separate live processing input from that of Antescofo to keep separate level.

#### Tuning¶

For pitch and polyphonic detection, Antescofo is tuned by default to 440.Hz (on A4). This can be changed with the message tune $1, where $1 is the new tuning.

#### Harmonic Analysis Control¶

##### Tracing the Evaluation of Functions¶

Function calls can be traced, see the description of the functions @Tracing and @UnTracing.