The Sound Engine
Crunchy, warm, a little dusty around the edges. Here is where that sound comes from, stage by stage, and how to push it further or pull it back.
What makes the SP-1200 sound
The SP-1200's sound comes from its circuits more than from any effect. Its input amplifier, a 12-bit converter running at about 26 kHz, its filters and its output stage all colour the sound in small ways that add up. eSPi models that whole signal path, and every sound you play goes through it by default.
The path comes in two halves:
- When a sound is loaded or sampled. The input stage, the converter and the bit depth are printed into the sample at that moment. What is on the pad already sounds like an SP-1200 sample.
- Every time the sound plays. Pitch, the three filters, the envelope and the output stage are applied live, so you can change them at any time.
When a sound is loaded
Every sound that lands on a pad, whether you drop a file, load from the library or sample your input, takes this route once:
- Input stage. A model of the SP-1200's input amplifier: gentle saturation, slew, a little preamp hiss, a slight droop at the top end, and the steep anti-alias filter in front of the converter.
- Sample rate. The sound is converted to the engine sample rate, 26,040 Hz by default.
- The 12-bit converter. A model of the converter chip itself: sample-and-hold, converter non-linearity, carry glitches, comparator noise and settling.
- Bit depth. The sound is reduced to the Engine Bitrate, 12-bit by default.
- DC removal. Any DC offset is taken out.
The result is the classic sound: gritty, slightly noisy and rolled off at the top. Because this all happens at load time, changing these settings affects sounds you load from then on, not the ones already on your pads. Reload a sound to hear it through new settings.
Every time a sound plays
Each voice then runs through these stages, in this order:
- Pitch
- Pitch is a change of speed, as on the hardware: higher notes play faster and shorter, lower notes slower and longer. In SP 1200 pitch mode, eSPi uses the original's slightly-off pitch steps and its raw resampling; in Clean mode, exact tuning and smooth interpolation. See Pitch mode.
- SSM2044 filter
- The analog low-pass filter chip, with cutoff, resonance and envelope amount per pad. See Filters.
- SP Channel filter
- The fixed filters on the SP-1200's output channels, chosen per pad from Ch3 to Ch6.
- Multi filter
- A resonant low-pass, high-pass or notch filter per pad, off by default. See Multi page.
- Envelope
- Attack, decay, sustain and release. See Envelope.
Effects, groups and the mixer are covered in Effects and Mixer & Outputs.
The output stage
After the voices are summed, the whole mix passes once through a model of the SP-1200's output electronics. This is where much of the machine's weight and glue comes from.
- The real machine's stepped digital-to-analog output.
- Saturation and slew from the output op-amps.
- Small converter errors, different for each voice.
- VCA Zipper: the stepped, 8-bit volume control of the original.
- Low-Mid Voicing: the SP-1200's warmth in the low mids.
- Presence Dip: its slight dip around 10 kHz.
- Mix Phase Invert: an optional polarity flip of the whole mix, off by default.
Clean edges
Grit is the goal; clicks are not. Every note fades in and out over 1 to 2 ms, voices that are stolen fade out instead of cutting off, and start and end points you edit snap to zero crossings while Zero-Crossing Snap is on.
Change the character
Everything above is on by default. When you want more crunch, less noise, or a clean modern sampler, the switches are in Settings › Engine. The ones that change the sound are summarised here.
Engine Bitrate
Choose 6-bit, 8-bit, 12-bit (SP-1200) (the default), 16-bit or 24-bit. Lower is crunchier; 16 and 24 bits are nearly clean.
Engine Sample Rate
Choose 13020 Hz, 19530 Hz, 26040 Hz (SP-1200) (the default), 27050 Hz (SP-12), 32000 Hz or 44100 Hz, or type any rate from 8000 to 48000. Lower rates lose top end and gain grit; higher rates sound more open.
Bitrate and sample rate both apply to sounds loaded from then on.
Input Stage (SP-1200)
Switches the whole input model on or off. Its Advanced... button reveals each part on its own switch:
- Saturation, Slew Limiting, Preamp Noise, HF Droop
- Phase Inversion (off by default)
- Anti-Alias Filter
- Sample & Hold
- DAC Non-Linearity, Carry Glitch, LM311 Comparator, Settling
- Noise Floor, Clock Bleed
Output Stage (SP-1200)
Switches the output model on or off. Its Advanced... button reveals:
- Saturation, Slew Limiting
- DAC Nonlinearity (INL), VCA Zipper (8-bit)
- Mix Phase Invert (180 deg) (off by default)
- Low-Mid Voicing, Presence Dip (10k)
Turning a master switch on turns all of its parts on, except the phase-invert switches. Turning it off turns them all off.
Other switches
- 7bit Semitones
- Uses the SP-1200's 7-bit pitch steps when a sound is pitched. On by default.
- Zero-Crossing Snap
- Snaps edited start and end points to zero crossings so edits don't click. On by default.
- DC Removal
- Removes DC offset from loaded sounds. On by default.
SP-1200 Mode
The sound engine is SP-1200 all the time. SP-1200 Mode goes further and brings back the hardware's limits, so that what you make will fit and behave on a real machine. Turn it on with Setup 21 or in Settings › Engine, choosing Original or Reissue.
- A memory budget. Original has 4 banks of 2.5 seconds of sample memory; Reissue has more (8 banks). The header shows a memory meter, and a load that doesn't fit shows MEMORY FULL!.
- Hardware ranges. 8 channels, 24 PPQ sequencer resolution, 40 to 240 BPM, 1/8 swing only, and pitch always locked when recording.
- No stock kit.
Turning the mode on starts a new project, and eSPi offers to save the current one first. Everything about the mode, from the memory meter to exporting a disk for real hardware, is in SP-1200 Disks.