Jonas Eriksson Tape Fiasco 2 [WiN]

• Product: Tape Fiasco 2
• Developer: Jonas Eriksson
• Version: 2.0.0.660
• Format: VST3
• Requirements: Windows 10 or later
• Source: erikssonjonas.com/tapefiasco-2

Tape Fiasco 2 is a creative time-manipulation plugin built around four simultaneous, freely orderable engines — granular stretch, tape-style varispeed, rhythmic stutter, and BendIt, a circuit-bent buffer mangler — feeding into a 25-effect processing chain and a 72-destination modulation system. It sits as a self-contained sound-mangling workstation on a single loop or sample, replacing what would otherwise require chaining a time-stretcher, a saturation unit, a sequencer, and a modulation matrix across separate plugins. Its differentiator is depth-per-dollar: four distinct time engines, 25 additional effects, and a host-automatable Perform surface ship in one $29 plugin rather than as separate purchases. For anyone deciding whether one consolidated time-mangling tool replaces a multi-plugin glitch chain, this is that comparison.

Key Takeaway

Activates when a loop or sample needs rhythmic re-editing, tape-style pitch-and-speed coupling, or circuit-bent glitch texture built and modulated inside one tool rather than bounced through several. Displaces a multi-plugin chain of separate time-stretch, saturation, sequencing, and modulation tools when consolidation and host-automatable performance control matter more than each individual effect’s depth. Doesn’t offer free-running, non-tempo-synced LFO modes in its modulation system, and its built-in sequencers are capped at a fixed step count rather than supporting polymetric step lengths; producers wanting unsynced, freely drifting modulation or longer non-power-of-two sequences will hit those specific ceilings.

Four Time Engines and Why Stretch Isn’t a Pitch-Shifter

Stretch and Varispeed both couple speed and pitch the way physical tape does — slow the playback rate and the pitch drops with it — rather than offering independent pitch-shifting with formant preservation or granular pitch correction; the developer states this directly rather than leaving it ambiguous, since the character of both engines comes specifically from that coupling, not from working around it. Stutter captures audio into a slice aligned to the host beat grid and loops that slice at a rate set by free milliseconds or a note division, with the captured slice carrying its own pitch shift, decay across repeats, alternating pan, and a chance of mid-cycle re-capture, giving each stutter pass variation rather than identical looped repeats.

BendIt repurposes the pitch and probability controls into two macro systems: a seven-zone WARP control running from subtle wobble through a full host-quantized DJ scratch mode, and a five-zone FRACTURE control for sub-section jumps and silence drops. The scratch zone specifically snaps its entry to the nearest eighth-note boundary and re-rolls its rate, shape, and depth on every beat, which means two scratch gestures on the same material rarely sound identical even with no parameter changes between them. All four engines can be chained in any order, so routing Stutter’s rhythmic capture into BendIt’s scratch zone, or Varispeed’s tape character ahead of Stretch’s granular freeze, changes the result structurally rather than just layering one effect’s color on top of another’s.

A 25-Effect Chain Behind the Time Engines

Past the four time engines sits a drag-to-reorder chain of 25 additional effects spanning six categories — distortion, filtering, resonance, modulation, delay, and reverb — including a 60-voice scale-tuned resonator bank (Crystal), a four-topology multimode filter (Prism, covering state-variable, Moog ladder, 303 diode ladder, and an intentionally “broken” transistor model), and a stereo tape delay with independently adjustable left and right rates. Effects in this chain can be reordered freely, so running the Sieve spectral filter before or after the Squash multiband compressor produces structurally different results rather than the same processing in a different visual position.

This chain sits downstream of the time engines specifically, meaning every color and texture effect here processes audio that’s already been stretched, stuttered, or bent rather than working on the clean input signal. The master stage closes the chain with a three-band EQ, compressor, soft limiter, and auto gain, which keeps heavily mangled output from clipping or sitting at an unpredictable level after 25 effects’ worth of potential gain changes — a practical necessity given how much level a chain combining bit-crushing, wavefolding, and resonant filtering in sequence can introduce before any final gain staging.

Modulation Depth and Where It Stops Short

The Animate page provides four modulator columns independently switchable between LFO, Random, and Envelope Follower modes, plus three bezier graph editors, for up to 42 simultaneous modulation destinations; the Perform page adds four macro knobs and an X/Y pad routable to 30 more, bringing the system total to 72 destinations across both surfaces combined. Envelope Follower mode includes a bipolar Tone control that turns the follower into either a lowpass-filtered version tracking low-frequency content like kicks and bass, or a highpass-filtered version tracking transient-heavy content like hats — a single control covering two structurally different tracking behaviors rather than one fixed follower response.

Independent hands-on testing found the LFO modulator’s 11 available rates are all host-tempo-synced, with no free-running, unsynced rate option available — a real ceiling for sound design specifically wanting modulation that drifts independently of song tempo rather than locking to it. The same testing found the built-in Stutter and Stretch sequencers fixed at a set step length with no polymetric option, and reported that activating a step in either sequencer requires a separate click before that step’s value becomes adjustable, adding a step to programming a sequence compared to a workflow where dragging directly on a step would set both its state and value in one motion.

Perform Page and Host-Automatable Live Control

The Perform page exposes four macro knobs and an X/Y pad as a deliberately smaller surface than the Animate page’s full modulation matrix, with each macro and each X/Y axis routable to up to five destinations using triangular response curves with adjustable start, end, and skew points per destination. Macro and X/Y gestures recorded during a performance land directly in the host’s own automation lane, which means a performance captured this way becomes editable, host-native automation data afterward rather than a fixed, baked-in recording that would need to be re-performed to change.

A dimming behavior locks out manual edits on any control currently assigned as a macro destination, which prevents a macro gesture from fighting against a manually adjusted knob that’s also under that macro’s control — a real usability decision that trades direct per-knob access for predictable macro behavior. This means a control assigned to a macro destination can’t be tweaked by hand independently of that macro while the assignment is active; removing the destination assignment is required to regain direct manual control over that specific parameter.

One Tool Replacing a Chain, With Its Own Fixed Tempo Grid

Tape Fiasco 2’s whole pitch is consolidation — four time engines, 25 effects, and a full performance surface in one plugin — and that consolidation comes with its modulation system locked to the host tempo rather than offering the free-running drift a standalone LFO plugin would.

FAQs