Load and edit a sample

Tempera supports many common audio formats as samples, stores them internally as 16-bit 48kHz, and processes them with 32-bit floating point math.

The maximum length of a sample loaded into track is about 11 seconds, however when a longer audio file (up to 5 minutes) is loaded, it is possible to select which slice of audio to load.

When the waveform is now shown across all 4 displays, gray vertical bars are hinting at where the cell boundaries will be. If you’re loading a percussive sample, you will likely want to line the markers with sample transients.

To load a sample into a track:

  1. Go to Tracks

  2. Choose the track you want to load a sample into and press Edit

  3. Press Load to go to the file browser

  4. Choose a sample and press Load

  5. Adjust the Start and End points and press Load. Scroll faster by holding the Round button.

  6. The sample is now loaded into the track, don’t forget to adjust its Base tuning

To edit a sample when it’s already loaded into a track:

  1. Go to Tracks

  2. Choose the track you want to load a sample into and press Edit

  3. Press Trim

Then, on the Trim page:

  • Turn the first and last knobs to adjust the start and end points

  • Listen to the whole sample or just its tail

  • Normalize it, and Trim everything before or after the start and end points, respectively.

  • Rev to reverse the sample

Frequently Asked Questions

  • Why does Tempera export audio files as .flac?

    FLAC is an open lossless compression algorithm, which makes audio around 50% smaller at zero loss of quality. Think of it like a zip file, and not like an mp3.

  • After saving a canvas, can I move or delete the original audio files used as samples?

    Yes, the saved .canvas file is fully self-contained.

  • When previewing/auditioning audio samples, the volume is too loud.

    Adjust Audition Volume in Settings Settings.


While Tempera will accept and play any audio, here are a few general recommendations for making your own samples:

  • If you’re making a harmonic sound (such as a synth waveform), try keeping the base sample to a lower tuning, such as 110Hz or 220Hz.

  • When making sequenced sounds, it’s nice if the audio sample is neatly divisible in 8 equal length slices. This will make it align well with the touchgrid cells.

  • When making a sample with an embedded melody that’s meant to be played with a keyboard, it’s usually a good choice to have the melody run in “safe” notes, such as octaves and fifths. For a bonus challenge: try making a sample melody more with timbre and less with pitch.