Audio Processing
Overview
Section titled “Overview”EP-PatchStudio includes thorough audio processing to prepare samples for your EP device. These processing options apply when using the Batch Processor or exporting from the Sample Editor.
All processing is non-destructive in the editor, so changes are only applied when you export or process files.
Channel conversion
Section titled “Channel conversion”Mono vs stereo
Section titled “Mono vs stereo”| Mode | Use Case | File Size |
|---|---|---|
| Mono | Most drums, single-source samples | 50% smaller |
| Stereo | Pads, ambiences, stereo-recorded samples | 2× file size |
How it works
Section titled “How it works”- Mono conversion: Averages left and right channels into a single centre channel
- Stereo: Preserves both left and right channels
Recommendation
Use mono for drums and most samples (saves space, every EP device has limited storage).
Normalisation
Section titled “Normalisation”Automatically adjusts audio levels to maximise loudness without clipping.
Settings
Section titled “Settings”- Enabled/Disabled: Toggle normalisation on/off
- Headroom: -12dB to 0dB (default: -6dB)
How it works
Section titled “How it works”- Analyses peak level in audio
- Scales entire audio to reach target peak level
- Preserves dynamic range (proportional scaling)
Headroom values
Section titled “Headroom values”- -12dB: Very conservative, lots of headroom for layering/effects
- -6dB: Balanced (recommended), safe headroom for device
- -3dB: Loud, minimal headroom
- 0dB: Maximum loudness, peaks reach full scale (risk of clipping)
Example
Section titled “Example”Input: Quiet sample with -18dB peak
Normalise to -6dB headroom
→ Output: +12dB gain applied, peak at -6dB
Recommendation
Use -6dB headroom (standard mastering practice, prevents inter-sample peaks).
Trim silence
Section titled “Trim silence”Automatically removes silence from the start and end of samples.
Settings
Section titled “Settings”- Enabled/Disabled: Toggle silence trimming
- Threshold: -80dB to -30dB (default: -60dB)
How it works
Section titled “How it works”- Scans from start: finds first sample above threshold
- Scans from end: finds last sample above threshold
- Trims everything outside these points
Threshold values
Section titled “Threshold values”- -80dB: Very sensitive, keeps quiet room noise/reverb tails
- -60dB: Balanced (recommended), removes silence but keeps reverb
- -40dB: Aggressive, removes quiet content and reverb tails
- -30dB: Very aggressive, may cut wanted audio
Use cases
Section titled “Use cases”- Recorded samples with pre-roll silence
- Samples with excessive reverb tails
- Cleaning up field recordings
Recommendation
Use -60dB for most samples. Disable for samples where you want to preserve room tone or reverb.
Pitch shift
Section titled “Pitch shift”Transposes audio by semitones without changing duration (time-stretching algorithm).
Settings
Section titled “Settings”- Range: -24 to +24 semitones (-2 octaves to +2 octaves)
- Default: 0 (no shift)
Common uses
Section titled “Common uses”- -12 semitones: Octave down (deeper, darker)
- -7 semitones: Perfect fifth down
- -5 semitones: Perfect fourth down
- +12 semitones: Octave up (brighter, tighter)
- +7 semitones: Perfect fifth up
How it works
Section titled “How it works”- Uses phase vocoder time-stretching algorithm
- Preserves duration whilst changing pitch
- May introduce slight artefacts (formant shifting, metallic quality)
Example
Section titled “Example”Input: Snare drum at C3 (MIDI 48)
Pitch shift: -12 semitones
→ Output: Snare at C2 (MIDI 36), same duration, deeper/fuller
Recommendation
Use sparingly. Large shifts (>±12 semitones) can sound unnatural. For subtle tuning, use ±1-3 semitones.
Long sample handling (>40 seconds)
Section titled “Long sample handling (>40 seconds)”The EP-40 Riddim has a 40-second maximum sample duration. When batch processing encounters longer samples, it offers four handling modes:
1. ask (default)
Section titled “1. ask (default)”Behaviour:
- Pauses processing and shows modal dialogue
- User chooses: Split, Pitch Up, Trim, or Skip
- “Remember choice for this session” checkbox
Use case: When you want manual control over each long sample.
2. Auto-Split
Section titled “2. Auto-Split”Behaviour:
- Automatically splits sample into 40-second chunks
- Naming: filename_00-40s.wav, filename_40-80s.wav, etc.
- Splits at nearest zero-crossing for click-free transitions
Example:
Input: 100-second pad
→ Output:
- filename_00-40s.wav (0:00-0:40)
- filename_40-80s.wav (0:40-1:20)
- filename_80-100s.wav (1:20-1:40)
Use case: Ambient pads, loops, long textures that can be split without losing musicality.
Pros: Preserves entire audio
Cons: Creates multiple files, may split mid-phrase
3. pitch up (+12 semitones)
Section titled “3. pitch up (+12 semitones)”Behaviour:
- Pitches sample up by +12 semitones (1 octave)
- Effect: Doubles playback speed, halves duration
- Can apply multiple times: +12st → +24st if still too long
Example:
Input: 80-second drone (too long)
→ Pitch up +12 semitones
→ Output: 40-second drone, 1 octave higher
Use case: Drones, pads, loops where higher pitch is musically acceptable.
Pros: Single file, preserves continuity
Cons: Changes pitch/character
4. trim to 40 seconds
Section titled “4. trim to 40 seconds”Behaviour:
- Keeps first 40 seconds, discards the rest
- Permanent: Lost audio cannot be recovered
Example:
Input: 100-second sample
→ Output: First 40 seconds only (0:00-0:40)
Use case: When only the intro/start of the sample is needed.
Pros: Simple, single file
Cons: Loses audio data
5. skip
Section titled “5. skip”Behaviour:
- Skips file entirely (marked as error in queue)
- Result: File not processed or included in output
Use case: When long samples shouldn’t be processed (e.g., full songs, stems).
Workflow recommendations
Section titled “Workflow recommendations”| Scenario | Recommended Mode | Notes |
|---|---|---|
| Ambient pads/textures | Auto-Split | Preserves all audio, can layer chunks |
| Drones/loops | Pitch Up | Musical if octave up sounds good |
| Field recordings | Trim to 40s | Usually only need beginning |
| Full tracks/stems | Skip | Not suitable for device |
| Mixed batch | Ask | Manual control per file |
Processing order
Section titled “Processing order”The batch processor applies settings in this specific order:
- Load audio file (decode to AudioBuffer)
- Resample (if target sample rate differs from source)
- Convert channels (stereo → mono if selected)
- Trim silence (if enabled)
- Normalise (if enabled)
- Pitch shift (if not 0 semitones)
- Quantise bit depth (8/10/12/16/24-bit)
- Export as WAV (16-bit or 24-bit container)
Important
This order is optimised for quality. For example, normalisation happens after trimming silence (so levels are based on actual audio, not silence).
Related documentation
Section titled “Related documentation”Export options
Section titled “Export options”Learn about sample rates, bit depths and audio format options
View Export Options →
Batch Processor
Section titled “Batch Processor”Process multiple audio files with automated workflows
View Batch Processor Docs →
Ready to process your samples?
Section titled “Ready to process your samples?”Audio processing tools for every EP device. Free trial available with full features.
Try FreeBack to Docs
SquareWave Studio
SquareWave Studio makes grown-up software for Teenage Engineering devices.
- OP-PatchStudio
- EP-PatchStudio
Resources
Section titled “Resources”- Docs
- Changelog
- Previous Versions
- Blog
- Superfan Upgrade
Connect
Section titled “Connect”- Discord
- YouTube
- Newsletter
- Contact
- Legal & Interoperability
- Privacy Policy
- Cookie notice
- Cookie settings
- Terms of Service
- License Agreement
- Refund Policy
- Accessibility