Tape / FIELD GUIDE

A little tape machine.

Put a sound in. Let it go around. Change something, and listen to what comes back.

Tape records audio through a magnetic model, then feeds the playback back into the recorder. Each trip can pick up a little more of the machine: softened attacks, wandering pitch, a disappearing top end. These docs follow the controls on the Deck tab, then open up the components in Advanced.

The loop stores returning audio. Feedback sets how much of that audio goes around again; it is not a physical erase-head current. The recording itself still goes through the record electronics, magnetic material, and playback head. For the longer story, see how it works.

The deck

Start with a drum loop or a few held notes. Keep Speed at 1×, Source and Record on, Feedback around −6 dB, and Dry/Wet at 100%. Turn Input up a little to hear the recording work harder. Bring Volume down to compare at a similar listening level. Now turn Source off and let the repeats find their way out.

Drag a knob to change it, hold Shift for a finer adjustment, type into its value box for an exact setting, or double-click to return to its default. The machine preset menu changes a collection of physical settings and keeps your recorded loop. The presets are imagined machines with different personalities, rather than calibrated copies of particular decks.

Deck / Advanced switches between the performance controls and the components inside the machine. It does not switch the audio processing on or off. In the standalone app, Audio settings selects your audio devices; in a plugin, your DAW handles that.

Transport

Speed 0–2×

This is a multiplier of the machine’s Nominal speed. At 1×, tape moves at that speed; 0.5× takes twice as long to go around, and 2× takes half as long. Changing speed bends the pitch of an existing recording. Slower tape also gives short wavelengths less room, so the high end can become softer.

Your incoming audio keeps its own pace. Slow the tape under a playing drum loop and the drummer keeps going while the tape catches a different pattern. The motor’s response and any wow or flutter also affect the actual speed.

Stop tape / Play tape

The button names the action: Stop tape means the tape is currently running. Press it to bring travel to a stop; it then reads Play tape. The transport smooths the speed change. A stopped tape stops carrying new parts of the loop past the heads. The source clock and powered record head keep going, so this is not a pause button for the entire instrument.

Reverse

Plays the saved loop backwards. Recording and feedback updates are suspended while reverse is on, so you can turn the loop around without overwriting it. Fresh source time keeps advancing. Return to forward playback and the Source, Record, and Feedback settings take over again.

Source

Input dB

The level of fresh audio entering the recorder. 0 dB leaves its level unchanged; higher values push it harder into the record electronics and tape. You may hear thicker notes, rounded transients, or obvious breakup. Lower values give that fresh audio more room. Input does not turn up the returning loop.

Source on / Source off

This button names the current state. Source on admits new audio from your DAW, audio device, or loaded source file. Source off closes that input, while the returning loop can still be recorded again. Use it to stop adding new layers and listen to the ones already there.

Record on / Record off

Record powers both the audio recording signal and the high-frequency bias current. With Record on, fresh input and feedback pass through the magnetic recording path. Record off stops that write path. The saved loop can still play, and Feedback can still fade it each pass.

To hold a loop, set Record off and Feedback to 0 dB. Source off alone leaves the recorder active: the sound can keep changing each time around.

Loop

Feedback −∞ to 0 dB

How much of each channel’s playback returns for another pass. The knob links the two levels; left and right audio stay separate. At −∞, nothing returns. Around −6 dB, the return multiplier is about one half. At 0 dB, the full level returns.

With Record on, that returning audio also passes through the recording and playback model. A −6 dB setting does not promise repeats exactly half as loud, and 0 dB does not promise an unchanged loop. The machine’s gain, saturation, filtering, and noise all take part. Higher feedback makes those accumulated changes easier to hear.

Volume −∞ to +18.1 dB

The knob below Feedback sets your listening level after the tape path. 0 dB is unity and −∞ mutes the tape output. Changing Volume does not change recording, feedback, or the dry source in the Dry/Wet blend. Use it to compare a louder, more driven setting with a quieter one.

Loop length 5–100 cm

Changes the length of the whole recorded loop. A longer loop gives the sound more time before it returns; a shorter one brings it back sooner. Existing audio is stretched to fit with pitch preservation. Listen for stretching texture with large changes or sharp transients. The current loop keeps playing while the new length is prepared.

Length and Speed are different ways to change the timing: length stretches the recording while trying to keep its pitch; speed changes how fast that recording passes the head. The Loop period readout shows the time around at the current travel speed. Head separation is already part of that trip.

Head transit shows the time from recording to pickup at the current forward speed. It is shorter than the whole trip around the loop. When stopped, both timing readouts show “--”; reverse still shows a loop period, but no forward head transit.

Dropping audio onto the tape

Drop an audio file onto the tape picture to replace the recorded loop. Its duration sets the length, with pitch-preserving stretching if it needs to fit within 5–100 cm. Import turns Source and Record off, sets Feedback to 0 dB, and starts the tape. Turn Source and Record back on when you want to overdub.

To keep a file playing into the recorder instead, use Advanced → Source input.

Playback

Dry/Wet 0–100%

Blends the original source with the tape playback. 0% is fully dry and serves as bypass; 100% is tape only, the default. Between them, the blend is linear: 50% combines half the dry source with half the tape playback after Volume. Mix changes smooth over 5 ms.

The dry source plays at its original level, independent of Input, Source on/off, Record, and Volume. The tape keeps moving and recording according to its other controls, even at 0%. Turn Dry/Wet back up to hear where it has gone.

If a source file is loaded, the dry side plays that file. Your DAW’s own plugin bypass passes the host input instead. With Align to DAW on, compensation delay also applies to the dry signal.

Align to DAW

Reports the processing delay and the record-to-playback head travel to your DAW so it can compensate. Turn it on before playback when you want the tape output to line up with other tracks. With it off, only numerical processing latency is reported, leaving the head travel as part of the tape timing.

Alignment uses the selected speed and keeps the last nonzero speed during a stop. It does not undo pitch bends, wow, flutter, or the motor’s changing speed, and it cannot remove the wait when monitoring a live input. In the standalone app there is no DAW to compensate the wet playback.

If the display says Too slow to align (>10 s), the head-travel delay is beyond the supported compensation range. Only processing latency is reported until you return to a supported speed. Some DAWs need playback stopped and restarted to pick up a changed delay.

Switches, together

These are useful starting points with the tape running forward and Dry/Wet at 100%.

To hear…SourceRecordFeedback
Fresh audio through tape, onceOnOn−∞
New layers with fading repeatsOnOnTry −6 dB
The loop changing without new inputOffOnTry −6 dB, then raise it
The saved loop heldOffOff0 dB
The saved loop fading without another magnetic writeOffOffBelow 0 dB

Reverse temporarily suspends the writes and feedback changes in this table. Dry/Wet changes what you hear while the tape continues underneath.

Advanced: inside the machine

This is where you change the tape and the machine around it. A gap gets wider. A winding gets more current. The motor gets a little less steady. The sound follows those changes through the model.

Choose a group from the Advanced menu. The descriptions below follow the knobs in each group; the expandable notes cover additional parameters available through DAW automation. Hover or focus a knob in the app to highlight its component. The drawings are magnified explanations, not measurements of your audio.

Try one control at a time with Record on, forward playback, and Dry/Wet at 100%. A held loop with Record off plays saved audio, so it will not demonstrate a new magnetic write or a newly adjusted pickup in the same way. Physical changes keep the loop and prepare in the background; let the change settle before judging it. Some combinations of dimensions cannot be prepared, so check the status if a requested change does not take.

These are things to listen for, rather than promises that turning a knob up always makes the same sound. Bias, drive, tape, and head geometry interact. For a first experiment, raise Record current a little, then compare it with lowering Driver current limit. Both can make a mess of a drum hit, but at different places in the machine.

Material

The coating holds a magnetic history. These controls describe how much magnetization it can carry, how it switches, and how uneven it is along the tape.

Coating thickness µm

The depth of magnetic coating, sampled at multiple depths in the model. A thicker coating puts more material under the head, but its deeper layers are harder to reach and contribute differently to playback. It can change the level and low-to-high balance. Thinner coating brings the layers closer to the surface; thickness is not a simple volume or warmth control.

Saturation magnetization A/m

The magnetization reached when the modeled domains are fully aligned. Raising it increases the available magnetic output scale for the same switching history; lowering it reduces that scale. This changes playback level and, with feedback, how hard the next pass is driven. It does not by itself raise the field needed to reach saturation.

Coercivity A/m

How much field it takes to reverse magnetic domains. Higher coercivity makes the tape harder to write with the same current and bias. Lower coercivity lets that same head push the material further. Listen for changes in sensitivity, distortion, and the point where loud attacks start to flatten; harder tape may need the bias and record current adjusted together.

Coercivity spread %

Spreads the switching thresholds among the domain families at each location. A wider spread mixes easier and harder domains, distributing switching over a broader range of fields. A narrower spread makes their thresholds more alike. This changes the shape of compression and harmonic texture, especially on sustained sounds.

Switching center width A/m

The width of the distribution of switching centers. Larger values spread the transition over a wider field range, giving a gentler change in magnetization for a given field change. Smaller values concentrate it. Listen for differences between rounded compression and more abrupt, grainy transitions, particularly with low bias.

Coercivity variation %

Makes some stretches of tape harder to write than others. This unevenness belongs to positions on the material as it moves through the writer. Turning it up can make a steady tone less even, with extra roughness and changing distortion. At zero, this particular source of material variation is removed.

Remanence noise A/m

Adds a fixed random magnetic contribution to material leaving the record region. More gives the playback head more magnetic grain to read, including during quiet passages. Less makes the coating quieter. Its audible color depends on tape speed, the noise’s spatial correlation, and the pickup response.

Modulation noise %

Scales the recorded magnetization by a small, material-fixed random variation. Unlike the additive remanence noise, this contribution grows with the recording. Raise it to hear more grain around notes and roughness in their tails; lower it for a steadier recorded level.

Additional Material parameters in DAW automation

These stay the same across the factory presets, so the current Advanced panel leaves them out.

Coercivity correlation µm

The distance over which neighboring coercivity variations remain related. Larger values make longer patches of similar hardness; smaller values make finer irregularity. At a given speed, longer patches tend toward slower changes in the recording. This needs nonzero Coercivity variation to have an effect.

Noise correlation µm

The spatial scale shared by remanence and modulation noise. Longer correlation shifts the material’s variation toward longer wavelengths; shorter correlation makes finer grain. The resulting hiss or texture also depends on speed and playback filtering. With both noise amounts at zero, there is no noise here to reshape.

Heads

The record head turns current into a field. The playback head senses the magnetic pattern farther down the tape. Gap means the opening between a head’s poles; spacing means the clearance between that head and the coating.

Head separation mm

The distance from record to playback. Greater separation makes fresh sound take longer to reach the pickup at the same speed; less separation makes it arrive sooner. The loop return accounts for this transit, so it is not an extra delay added on top of the loop’s length. It also changes the delay reported by Align to DAW.

Record gap µm

Changes the shape and extent of the recording field. A wider gap spreads the field over more tape while reducing the field scale for a given winding current. A narrower gap concentrates it. Both signal and bias see this geometry, so the effect on saturation and high-frequency recording depends on their levels as well as speed.

Record spacing µm

The clearance from the record head to the tape surface. More clearance weakens and spreads the field reaching the coating. This can soften detail, reduce recording level, or leave the tape effectively underbiased. Less clearance strengthens the coupling. Try changing it with a quiet tone as well as a loud one.

Playback gap µm

The pickup aperture averages the recording over a small length of tape. A wider gap has more trouble resolving short wavelengths, so treble can fall away or cancel at particular frequencies. A smaller gap retains finer detail. The same gap matters more at slow tape speeds, where an audio cycle occupies less tape.

Playback spacing µm

The distance from the coating to the pickup. Increasing it loses short-wavelength detail especially quickly, often making the sound quieter and darker. Decreasing it brings back more of what is recorded. This is a useful place to start when you want the sound of poor tape contact.

Track width mm

The width of each stereo track included in the pickup flux. Wider tracks increase the pickup’s output scale. When azimuth is nonzero, they also span more timing skew across each track, which can cancel more high-frequency content. Narrower tracks reduce both effects. The width must fit within Track pitch.

Track pitch mm

The distance between the centers of the left and right tracks. With a tilted playback head, larger pitch means a larger timing difference between channels. At zero azimuth, changing this distance does not create stereo widening or crosstalk. Listen on a centered sound while changing Azimuth to hear the relationship.

Azimuth mrad

The tilt of the playback gap across the tracks. Moving away from zero smears timing across each track and offsets left against right. Expect treble loss and a change in stereo phase, particularly when summed to mono. Changing the sign swaps which channel is ahead; zero gives the aligned geometry.

Additional Heads parameters in DAW automation

These winding and efficiency settings are available to the host, but do not appear as knobs in the current Advanced panel.

Record turns turns

The effective number of record winding turns. More turns produce a stronger modeled field from the same current, scaling both signal and bias. Fewer turns weaken both. The resulting sound depends on whether the material needed more excitation or was already being pushed too hard.

Record efficiency %

The fraction of ideal record-head coupling used by the model. Raising it strengthens signal and bias fields together; lowering it weakens them. It can move the recording toward or away from useful bias and saturation without changing the driver’s current limit.

Playback turns turns

Scales the voltage induced in the pickup winding. More turns make the magnetic contribution louder before playback electronics and feedback; fewer make it quieter. In this model, changing turns does not add a separate winding-resonance effect.

Playback efficiency %

Scales how strongly the playback winding couples to the recorded material. Higher values raise the magnetic signal reaching the electronics and the return path. Lower values reduce it relative to noise added by the playback electronics. Use Volume instead if you only want to change listening volume.

Electronics

The circuits before and after the magnetic part. Recording controls affect the combined fresh source and returning audio. Playback electronics also feed the return path, so their gain and filtering can build up over repeats.

Record current mA

How much signal current the record circuit asks for per unit of shaped input, before the driver limit. Raising it drives both fresh sound and feedback harder. Listen for rounded attacks, extra harmonics, and eventually strong breakup. Lower it for a less driven recording. The separate bias current is not scaled by this knob.

Driver current limit mA

The ceiling of the record amplifier’s smooth current limiting. Lowering it makes the electronics round off the signal sooner, often giving obvious compressed or fuzzy attacks. Raising it lets stronger peaks reach the magnetic material, where saturation can still occur. Bias is added after this signal-current limiter.

Bias current mA

The amplitude of the rapid alternating current added at the record head. Bias repeatedly moves the domains while the slower audio signal guides the recording. Too little can give rough, nonlinear quiet notes. Increasing it can smooth that response; going further can reduce sensitivity and treble. There is no universal sweet spot: it changes with the material, speed, and head geometry.

Record time constant µs

Sets the high-frequency roll-off before the recording driver. A longer time constant lowers the corner frequency, sending a darker signal to tape. A shorter one lets more high-frequency detail and sharp attacks reach the driver, which may also change how those peaks distort.

Compensation zero µs

Controls the rising part of the record compensation response. Increasing this time constant brings that rise lower in frequency and raises the upper-frequency drive relative to a fixed compensation pole. It can restore brightness or push cymbals and consonants into saturation. At zero, the rising contribution is removed; the compensation pole remains.

Playback time constant µs

Shapes the equalizer that turns the pickup voltage into a useful audio response. At a fixed Playback EQ gain, increasing this time constant raises the upper part of that response relative to the low end; decreasing it reduces it. Listen for a change in brightness and level, which can become stronger with each repeat.

Playback low cut Hz

Sets the low-frequency corner of playback equalization. It also sets the matching low-frequency pre-emphasis in the record circuit. Raising it tends to reduce very low playback content while changing how hard bass drives the recorder. The result can be leaner, but the linked record shaping means it is more than a bass filter after the sound.

Playback EQ gain 1/s

The scale of the playback equalizer, expressed in its physical filter units. Increasing it makes playback louder before the feedback is taken. That can make repeats persist or hit the recorder harder on their next pass. Lowering it reduces that gain. Volume controls the tape listening level without changing the return or the dry source.

Playback bandwidth kHz

The upper corner of the two playback low-pass stages. Lower values darken the output and remove more high-frequency noise; higher values let more treble through. With feedback and Record on, a restricted bandwidth can make each trip around lose a little more edge.

Input low cut Hz

Controls low-frequency coupling in the recording electronics after fresh source and feedback are added. Raising it removes more bass before the driver and tape, changing both the tonal balance and what drives the distortion. Lower it to admit deeper notes. It works together with the record pre-emphasis set by Playback low cut.

Playback noise nV/√Hz

The input noise density of the playback electronics. More adds hiss before playback equalization and bandwidth filtering. Less gives quieter electronics; zero removes this contribution. This noise is generated in time, separately from the material’s fixed magnetic grain, and it can join the feedback on later passes.

Additional Electronics parameters in DAW automation

Bias frequency and Compensation pole are fixed across the factory presets. Record drive and Record bias are additional host controls; they currently have no visible group in the Advanced menu.

Bias frequency kHz

The rate of the alternating bias carrier. Higher frequency packs more reversals into a given distance at the same tape speed; lower frequency gives fewer. It changes how the material is worked by the head and where carrier-related residue may land. It is not an audio tone control, and raising it is not a guarantee of cleaner sound.

Compensation pole µs

The falling part of record compensation. A larger time constant brings the pole lower, reducing upper-frequency drive relative to a fixed Compensation zero. A smaller value leaves the rise extending higher. Together, the zero and pole determine the shelf; equal time constants cancel this compensation stage.

Record drive × current

A multiplier of the combined signal current before the driver limit. Raising it drives the fresh source and feedback harder, much like increasing Record current. It leaves the bias multiplier alone. Setting it to zero removes the audio drive while bias can still excite the material.

Record bias × current

A multiplier of Bias current. At 1× it uses the physical setting; at zero it removes the carrier. Raising it moves toward stronger bias, with the same material-dependent trade between smoother recording and overbias loss. The carrier’s phase continues through changes and stops.

Motion

These controls change tape travel, pickup contact, or the bias supply. A movement in speed affects the writing and reading of the tape together. Fresh audio keeps its original clock.

Nominal speed ips

The machine’s tape speed in inches per second, before the Deck’s Speed multiplier. Faster tape gives an audio cycle more physical length and shortens both the trip between heads and the loop period. Slower tape does the opposite and tends to expose more high-frequency loss. Changing speed also changes the pitch of saved audio.

Wow depth %

The amount of periodic speed variation from the wow oscillator. More depth produces a larger pitch wander and changes repeat timing. Small amounts can loosen a sustained chord; large amounts make an obvious wobble. Zero removes this oscillator’s contribution.

Wow frequency Hz

How quickly that periodic speed variation cycles. Lower rates give a slow drift; higher rates move toward vibrato. The depth controls how far it moves, and the motor response can soften faster movement. With Wow depth at zero, changing the rate has no effect.

Flutter depth %

The amount of a second periodic speed variation, usually set faster than wow. Increase it for a more restless edge, pitch shimmer, or audible roughness around a steady note. Decrease it for steadier travel. It comes from transport modulation, so its sound also depends on the transit between heads.

Flutter frequency Hz

The cycling rate of periodic flutter. At lower rates you may hear individual pitch movements; faster rates can make a note sound buzzy or spread energy around it. The servo response limits how much of a fast requested movement reaches the tape. This requires nonzero Flutter depth.

Random flutter %

Adds an irregular, correlated disturbance to the transport instead of a repeating sine wave. More makes pitch less predictable; less makes the periodic wow and flutter easier to pick out. It is useful when a perfectly repeating wobble starts to sound too tidy.

Contact modulation µm

The amount of periodic change in playback head clearance. In the audio model this moves the pickup spacing around its base value, with a minimum positive clearance. More can make the high end and level breathe as coupling changes. Less gives steadier contact. It changes pickup coupling rather than tape speed.

Contact frequency Hz

The rate of that playback-spacing cycle. Slow settings can sound like swells or intermittent dulling; faster settings can add a tremulous texture. The envelope runs on the engine’s time clock. With Contact modulation at zero, there is no clearance movement to speed up.

Bias modulation %

Varies the bias amplitude above and below its base value. More depth makes the recording move between different bias conditions, which can vary its sensitivity, treble, and distortion. This works at the record head and keeps running through a tape stop when Record is on. It needs a nonzero bias current.

Bias modulation frequency Hz

The rate of the bias-amplitude envelope, separate from the much faster bias carrier. Lower values give a slower change in the recording’s character; higher values make those changes more rapid and can add roughness. Its effect depends on Bias modulation depth and the material’s response to the chosen bias range.

Additional Motion parameters in DAW automation

These response settings are available to the host but stay fixed across the factory presets.

Servo time constant ms

How quickly the modeled motor follows its requested nominal motion and disturbances. A larger value makes that response more gradual and reduces faster wow or flutter; a smaller value follows more quickly. The Deck’s Speed changes also have their own smoothing, so this is not a single control for the entire stop gesture.

Flutter correlation ms

How long the random speed disturbance remains related to its recent values. Longer correlation makes broader, slower wandering; shorter correlation gives faster irregularity, which the servo can smooth away. This needs nonzero Random flutter to change the sound.

Source input

A repeating file is useful when you want to keep sending the same sound through different settings. This group chooses what enters the recorder; it does not replace the saved tape loop.

Load file / Replace file

Choose an audio file to play continuously instead of the live device or DAW input. Replace file loads another clip from its beginning. Loading leaves the loop length and controls alone, and the current source keeps playing while its replacement loads. A failed manual load leaves that source in place.

The file repeats at its original speed, including while tape is stopped or reversed, or Source and Record are off. Turn those switches back on and you catch the file where it is now. Its clock advances whenever the host or device processes audio, independently of the DAW’s transport button; without audio processing, its position pauses.

WAV, AIFF, FLAC, Ogg, and other registered platform formats are supported. Use mono or stereo audio at 22.05–192 kHz, within the 8,388,608-frame source and converted-file limits. Mono feeds both channels; a mono plugin averages a stereo file. The player does not normalize levels or crossfade the seam, so a discontinuity between the file’s ends can click.

Clear

Removes the file source and returns to live device or DAW input. It does not clear the recorded tape. Source, Input, and Record still decide what gets recorded.

Filename, duration, and status

The panel shows the selected file, its original duration, and loading or error messages. Sessions remember the source file’s path, so keep that file available when reopening. It reloads from the beginning; a missing file supplies silence until you replace it or press Clear. Audio dropped directly onto the tape is different: that recorded loop is stored in the session.

Try a short drum break as the source, choose a longer tape loop, and turn Source off after a few passes. Now the file keeps going quietly behind the scenes while you work with what the tape has caught.