Namespaces

Variants
Actions

MSP MIDI Tutorial 3: MIDI Sampler

From Cycling '74 Wiki
(Difference between revisions)
Jump to: navigation, search
Line 1: Line 1:
Click here to open the tutorial patch: [[Media:03iMIDISampler.maxpat]]
+
Click here to open the tutorial patches: [[Media:03iMIDISampling.zip]]
  
 
In this tutorial we talk about how to create a MIDI-controllable sampler in MSP.
 
In this tutorial we talk about how to create a MIDI-controllable sampler in MSP.

Revision as of 16:06, 6 July 2012

Click here to open the tutorial patches: Media:03iMIDISampling.zip

In this tutorial we talk about how to create a MIDI-controllable sampler in MSP. Along the way, we'll look at different features of many samplers, including looping, keymaps, and multi-timbral operation.

  • To use the tutorial patchers in this section of the tutorial, make sure you have a correctly configured MIDI controller connected to your computer. The tutorials in this section use a variety of MIDI messages as example input; if your controller lacks any of these features, you can simulate their input with user interface objects in the tutorial.

Our fun sampler

  • Take a look at the patcher for this tutorial. The two sets of patcher logic

(labeled 1 and 2 both deal with playing back samples stored in buffer~ objects by using groove~ objects under MIDI control. Look at patcher area 1. Turn on the audio by clicking the ezdac~ and turn up the gain~ slider. From the umenu object at the top of the patch, select option 1 ('bd+hh.aiff'). Play some notes on your MIDI keyboard. You should hear a the bass drum sample at different pitches depending on which key you press. Select option 2 and take a listen; you should hear the snare drum.

  • Select option 3 ('cym.aiff') from the umenu. Play a note

on the keyboard and hold it down, or select a note from the kslider object. Notice that the sample loops. Select option 4; notice that the bass guitar sound loops a swell. Play the sounds triggered by options 5 and 6. Double-click the coll file named sampler.

The umenu object in our tutorial patcher causes the coll object to dump out different data depending on which sample we select. Looking at its contents by double-clicking it reveals the entire database:

1, 24 sample1 0 0 0; (bd+hh.aiff)

2, 33 sample2 0 0 0; (snare.aiff)

3, 50 sample3 0.136054 373.106537 1; (cym.aiff)

4, 67 sample4 60.204079 70.476189 1; (bass.aiff)

5, 84 sample5 0 0 0; (epno.aiff)

6, 108 sample6 0 0 0; (ahkey.aiff)

Each entry in the file sets up the groove~ object in the patcher to play a sample. The format of each line in the coll file is:

base_key buffer~_name loop_start loop_end looping

The base key refers to which MIDI note at which the sample will play at normal speed. The second item in the coll refers to which buffer~ object will play the sample. The next three values determine the start and end points of an internal loop which the groove~ object will play, and whether or not to use it at all (the last value).

The point of a sampler is to play back recordings (audio samples), often from an acoustic instrument. If an acoustic instrumental source is used, it's usually inefficient to create a unique sample for every possible note on that instrument. Instead, every few notes are sampled, and in-between pitches are achieved by playing these sampled notes slightly fast or slow based on their base key. Similarly, a sampler is capable of playing notes that sustain for far longer than it's practical to record an instrumental sampler. Instead, a sample of modest length is used, as an area is found within the sustaining part of the sample that can be safely looped. When you play a note and hold it, the sample starts playback at the beginning; once it moves into the loop zone, it repeats that area over and over again; when you pick up the note, it plays from the loop zone through to the end of the sample.

  • Click through the umenu again and notice how the different

parameters coming from the coll object set up the groove~ object.

The base key is converted from MIDI to frequency by the mtof object and used as a divisor for the frequency we want to play. In this way, we get a ratio of the desired frequency and the base frequency which we can use to set the speed of a groove~ object. If we want the sound to come out an octave higher, we want the ratio to be 2.0; an octave lower, it should be 0.5. The second value out of the coll file is formatted with a set prefix by the prepend object and send to the groove~ to select the appropriate buffer~ to play. The loop values come out as numeric data, setting the loop start and end points and sending a loop value into an int box which is triggered each time a note plays. Note-on values (i.e. values with velocities greater than 0, passed from the stripnote object) trigger the loop message, set the value of the sig~ object, and restart the sample by sending a 0 into the groove~ object. Note-off objects set the loop state to 0, allowing the groove~ to play out the sample to the end and then stop.

Multi-timbral samplers and keymaps

  • Look at the patcher logic labeled 2 on the right of the tutorial

patcher. As in the last tutorial, a poly object is used to route MIDI data to a number of instances of a single abstraction, this time called samplervoice~. However, the values out of the poly object do more than set the voice allocation.

  • Turn down the gain~ slider in patcher area 1 and turn

up the one in patcher area 2. Using an attached MIDI keyboard, play some notes all over the range of the keyboard. Notice that, depending on which notes you play, different samples are heard. Play some chords. Notice that you can have as many as four notes playing simultaneously. Look at the message box attached to the funbuff object:

0 1,

41 2,

48 3,

53 4,

68 5,

96 6

The funbuff object is loaded with these values to use them as a key map for a multi-timbral sampler. Any MIDI notes whose pitch values are between 0 and 40 will trigger sample 1 (as defined in the coll file); pitches between 41 and 47 will trigger sample 2; and so on.

  • Double-click any of the abstractions named samplervoice~ and

look at the patcher logic inside. Notice how it resembles the patcher logic in area 1 of the main patcher, with the addition of some objects to handle MIDI velocity to scale the amplitude output of the groove~ object.

The MIDI velocity of incoming notes is divided by 127. to scale it between 0. and 1. It is then multiplied by itself, creating an exponential scaling wherein higher values on the MIDI velocity continuum yield far greater increases in volume than lower numbers. This simulates the behavior of logarithmic volume circuits (such as mixer faders) in analog audio equipment.

  • Return to the main patcher, and click the toggle box labeled 'Play a

sequence'. Double-click the patcher object named sequence as you rock out to the sequence playing out of our sampler.

The sequence subpatch contains a seq object, a midiflush, and a midiparse. These objects load a MIDI file and output the raw bytes in response to start and stop messages (seq), shut off all sounding notes in a MIDI byte stream in response to a bang (midiflush), and parse and extract the pitch/velocity pairs from a MIDI stream so that they can be used elsewhere (midiparse). This allows our awesome MIDI sequence to be played by the sampler logic in the main patch.

Summary

MIDI-controllable samplers can be created using MSP buffer~ and groove~ objects. Different parameters of sampler data (loop points, sample name, base key) can be stored in coll files for easy access so that you can easily switch samples depending on MIDI events within the same MSP patcher logic.