CHOW Tape Model Explained: Every Control, Which Hysteresis Mode to Use, and a Settings Chart by Job
CHOW Tape Model is a free, open-source physical model of a reel-to-reel tape machine, originally based on the Sony TC-260. What every control does, from Drive, Saturation and Bias to the playhead, Degrade, Chew and Wow and Flutter sections, which hysteresis mode to pick, a settings chart for the mix bus, master, drums, bass, vocals and lo-fi, setup steps for a mix bus, and the mistakes that turn it into mud.
The short answer: CHOW Tape Model (ChowTape in its manual) is a free, open-source plugin from Chowdhury DSP that physically models an analog reel-to-reel tape machine, originally the Sony TC-260. The tape sound lives in the Hysteresis section: Drive pushes the signal into the tape, Saturation sets how early it distorts, and Bias sets how clean it stays, with low bias giving the underbiased "deadzone" sound. Around it sit input filters, a compressor, tone controls, a playhead section with tape Speed (which can snap to 3.75, 7.5, 15 or 30 ips), and three damage sections: Degrade, Chew, and Wow and Flutter. On a mix bus or master, pick a higher-order hysteresis mode, keep the drive low and the speed fast, switch the damage sections off, and if the low end swells, let the sub bypass the tape with Low Cut and Makeup. It comes in VST, VST3, AU, AAX, CLAP, LV2 and standalone versions across Windows, macOS and Linux, plus AUv3 on iOS, so it loads in every major DAW, Pro Tools included.
I use free saturators next to paid ones on my own releases (Sony, Universal, Ultra, Dim Mak, 50M+ streams), and at low drive the difference is mostly flavor, not quality. CHOW Tape Model is free and there are no affiliate links on this page. I sell a mastering chain whose color comes from tube saturation, not tape, disclosed at the end.
What CHOW Tape Model is
CHOW Tape Model started as a class project for Music 420 at Stanford's CCRMA, and the physical model behind it was published as a paper, "Real-time Physical Modelling for Analog Tape Machines," at the 2019 DAFx conference. It was first modeled on the Sony TC-260; the manual says the current version can emulate "a wide variety of reel-to-reel tape machines," and that several controls go further than a physical machine could.
The heart of it is magnetic hysteresis, the way tape magnetizes under a strong field. The plugin solves the Jiles-Atherton model of hysteresis in real time, and the manual calls hysteresis "largely responsible for the warm sound often associated with analog tape distortion."
It's open source under the GPLv3, which the manual describes as free "as in free beer" and "as in free speech." The developer's products page lists version 2.11.4, with installers for Windows, macOS and Linux and an iOS version on the App Store. Download it from chowdsp.com or the project's GitHub releases only.
The signal flow, and why the order matters
The manual draws the signal path in this order: Input Gain, the input filters, Tone (pre-emphasis), Compression, Hysteresis, Tone again (post-emphasis), Chew, Degrade, Wow and Flutter, then the playhead losses, before Output Gain and the Dry/Wet mix.
Three things follow from that order:
- Compression comes before the tape, so it shapes the dynamics going into the hysteresis, the way a hot or a careful record level would.
- The input filters' Makeup is added back after the tape. Whatever Low Cut or High Cut removes on the way in skips the whole tape process and returns at the output, untouched.
- Dry/Wet mixes the untouched input in at the very end, so parallel tape is built in.
Every control, explained
Most sections have their own power button, so you can switch a whole section off instead of zeroing its knobs.
| Section | Control | What it does | How to use it |
|---|---|---|---|
| Main | Input Gain | Sets the level going into the rest of the plugin | The manual recommends feeding the plugin at or below unity and adding any extra gain here, because very large input levels can make it unstable |
| Main | Dry/Wet | Blends the processed signal with the dry input | Pull it back for parallel tape on a master or drum bus |
| Main | Output Gain | Sets the level coming out | Match the bypassed level before you judge the sound |
| Main | Oversampling | Internal oversampling; more means fewer aliasing artifacts and better noise behavior, at more CPU | The manual says to use as much as your CPU allows |
| Main | Mix Group | Instances in the same group share the same parameters | Tape on several tracks, set once |
| Input filters | Low Cut and High Cut | Filter the input before the tape | Keep the sub, or the extreme top, out of the saturation |
| Input filters | Makeup | Adds what the filters removed back at the output | With Low Cut, the manual's example: sub-bass passes through unaffected |
| Stereo | Stereo or Mid/Side | Processes left and right, or encodes to mid and side, processes, and decodes back | Mid/Side to treat the center and the sides differently |
| Stereo | Balance | Makes one channel louder than the other going in | In Mid/Side mode, push toward the side to saturate the width more than the center |
| Stereo | Makeup | Restores the level Balance took away, at the output | On, so Balance changes how hard each channel hits the tape rather than the image |
| Hysteresis | Drive | The amplification inside the hysteresis process, which changes its nonlinear character (unlike Input Gain) | The main "how much tape" control |
| Hysteresis | Saturation | The level where the hysteresis saturates; higher values mean a lower saturation point and more distortion | Low for glue, higher for obvious color |
| Hysteresis | Bias | The amount of tape bias; lower bias widens the hysteresis curve and creates the "deadzone" of underbiased tape | Higher for clean tape, lower for the gritty, broken-machine sound |
| Hysteresis | Hysteresis Mode | The equation solver: RK2, RK4, NR4, NR8, STN or V1 | See the next section |
| Compression | Amount | Where the compression curve takes effect; at 0 dB it does nothing, and at 9 dB anything above -9 dB starts to be compressed, per the manual | Off for a clean bus; a little on drums to even out what hits the tape |
| Compression | Attack and Release | How fast compression starts and backs off | Slower attack keeps more of the transient going into the tape |
| Tone | Bass and Treble | Shape a pre-emphasis filter before the tape; the post-emphasis filter after it mirrors the move automatically | Not an EQ: they choose which part of the spectrum hits the tape harder |
| Tone | Frequency | The transition frequency between the bass and treble sections | Move it to decide where "bass" ends and "treble" begins |
| Playhead | Spacing | The space between the playhead and the tape | One of the physical causes of tape's high-frequency loss |
| Playhead | Thickness | The thickness of the tape | Another loss effect; small moves |
| Playhead | Gap | The width of the playhead gap | Another loss effect; small moves |
| Playhead | Azimuth | The playhead alignment angle; misalignment offsets the two stereo channels in time, which widens the image | A width effect; check mono |
| Playhead | Speed | Tape speed in inches per second, which changes all the loss effects; continuous, but it can snap to 3.75, 7.5, 15 and 30 ips | Faster for clean and open, slower for dark and colored |
| Degrade | Depth (with a 0.1x option) | The intensity of the wear on the tape; 0.1x makes the control more subtle | Use 0.1x for a hint of age |
| Degrade | Amount | How much wear, typically the age of the tape | More reads as older tape; raise it slowly |
| Degrade | Variance | Time-varying randomness in the degradation | Movement, so the wear doesn't sound static |
| Degrade | Envelope | An amplitude envelope on the tape noise | Set by ear on a loop with gaps, where the noise shows |
| Chew | Depth, Frequency, Variance | How deep the tape is chewed, how much space between chewed sections, and how random that spacing is | Dropouts and damage for lo-fi; off on anything clean |
| Wow and Flutter | Flutter Depth and Rate | Fast speed wobble, captured from a real Sony TC-260; Rate can sync to the tape speed or the song's tempo | Off on a bus or master |
| Wow and Flutter | Wow Depth, Rate, Variance, Drift | Slower wobble, with random irregularities from Variance and Drift | Lo-fi keys, pads and samples |
Because Tone's post-emphasis mirrors its pre-emphasis, turning Treble down doesn't darken the result the way an EQ cut would. It means the highs hit the tape less, so they saturate less; turning it up makes the top end saturate more. The playhead losses all depend on the tape speed, and per the plugin's source code the loss section also models a head bump, a low-end lift whose frequency moves with Speed and Gap.
Which hysteresis mode to use
| Mode | What it is | Use it on |
|---|---|---|
| RK2 | 2nd-order Runge-Kutta solver | Background tracks, when CPU is tight |
| RK4 | 4th-order Runge-Kutta solver | Background tracks with a little more accuracy |
| NR4 | 4-iteration Newton-Raphson solver | Key tracks |
| NR8 | 8-iteration Newton-Raphson solver | The mix bus, the master and the key tracks |
| STN | A State Transition Network, a cheaper approximation of NR8 | The mix bus when NR8 costs too much CPU; the manual says it distorts more harshly at extreme settings |
| V1 | The parameterization from earlier versions of the plugin | Older sessions, or if you prefer the original sound |
The manual's rule: the Runge-Kutta solvers are cheaper but less accurate than the Newton-Raphson ones, higher orders are more accurate and use more CPU, and the higher-order solvers belong on mix buses and key tracks.
CHOW Tape Model settings chart, by job
The manual doesn't publish knob ranges or defaults, so these are relative starting points. Set the final amount in context, at matched level.
| Job | Speed | Drive and Saturation | Bias | Sections on | Dry/Wet | Watch for |
|---|---|---|---|---|---|---|
| Mix bus glue | 15 or 30 ips | Low: you should only notice it when you bypass it | Higher, for clean tape | Hysteresis (NR8) and Playhead; Wow and Flutter, Degrade and Chew off | Full, output matched | The top end going dull over a whole song |
| Master | 30 ips | Very low | Higher | Hysteresis (NR8) and Playhead; Low Cut with Makeup if the sub swells | Pulled back | Low-end balance shifting; check against a reference |
| Drum bus | 15 ips | Low to medium | Higher | Hysteresis, Playhead, and a little Compression | Pulled back if the snap goes | Cymbals softening, the snare losing its crack |
| Bass and 808 | 15 ips | Low to medium | Higher | Input filters: Low Cut with Makeup, so the sub skips the tape and only the harmonics above it saturate | Full | The sub swelling or losing focus in mono |
| Vocals | 15 ips | Low | Higher | Tone with Treble turned down, so the esses hit the tape less | Full or slightly back | Spitty esses; de-ess before the tape |
| Wider sides on a bus | 15 or 30 ips | Low | Higher | Stereo: Mid/Side mode, Balance toward the side, Makeup on | Full | The center thinning or the image collapsing in mono |
| Keys, pads and lo-fi | 7.5 or 3.75 ips | Medium to high | Lower, if you want grit | Wow and Flutter on, Degrade on (0.1x for subtle), Chew for dropouts, Azimuth for width | To taste | Pitch wobble clashing with the vocal or the bass |
The general rule on speed holds here: faster speeds keep the top end flatter and more extended, slower speeds darken the sound and move where the low-end bump lands. Switch speeds while the loudest section plays and listen to the kick and the bass. The same jobs with every other free, stock and paid tape plugin are in the best tape saturation plugins.
How to set up CHOW Tape Model on a mix bus
- Download it from the developer only: chowdsp.com's products page or the GitHub releases, or the App Store on iOS. Free-plugin mirrors are how adware gets installed; the Camel Crusher story is the cautionary tale. FL Studio users: the developer suggests enabling the "use fixed size buffers" setting if you hear pops and clicks.
- Insert it on the mix bus, or before the limiter on a master; what a mastering chain is shows where saturation sits. Feed it at or below unity, as the manual recommends.
- Set Oversampling as high as your CPU allows, at least for the bounce.
- Choose NR8, or STN if NR8 is too heavy for your session.
- Switch off Wow and Flutter, Degrade and Chew, and leave Compression off for now. On a bus you want saturation, not damage.
- Snap Speed to 15 or 30 ips.
- Raise Drive while the loudest section plays until you hear the tape, then back off. Use Saturation to decide how early it distorts and Bias to keep it clean.
- Protect the low end if it swells: turn on Low Cut with Makeup, so the sub passes through untouched.
- Match the level with Output Gain and compare against bypass. Saturation adds level, and louder sounds better at first; how to use reference tracks covers loudness matching.
- Save it as a user preset. If you use it on several tracks, put them in one Mix Group so one change moves them all.
CHOW Tape Model vs other tape plugins
| Plugin | Price | Formats and notes | Best for |
|---|---|---|---|
| CHOW Tape Model | Free, open source | VST, VST3, AU, AAX, CLAP, LV2, standalone; AUv3 on iOS | The full tape toolkit, in every DAW |
| Airwindows ToTape9 | Free, open source | No AAX; the standalone build uses your DAW's generic sliders | Simpler, colored tape with fewer controls |
| HoRNet TapeLite | Free | VST3, AU and AAX | Five controls, Pro Tools friendly |
| Pro Tools Reel Tape Saturation | Stock (check your tier) | Pro Tools only | Pro Tools users on the mix bus |
| Logic ChromaGlow (Magnetic model) | Stock (Logic Pro 11 and later, Apple Silicon) | Logic only | Logic users who want tape with the lows left clean |
| UA Ampex ATR-102 | Paid | UADx Native and UAD DSP | A modeled mastering deck |
CHOW Tape Model is the free pick when you want to control the machine, not just hear it. For non-tape free saturators, from Softube Saturation Knob to Klanghelm IVGI, see the best free saturation plugins, and for how tape differs from tube and console color, what saturation is.
Mistakes that make CHOW Tape Model sound bad
- Feeding it hot. The manual warns that very large input levels can make the plugin unstable. Bring the level in at or below unity and add gain with Input Gain or Drive.
- Bouncing with low oversampling. Aliasing and noise are what oversampling fixes; raise it before the final export.
- A low-order solver on the mix bus. The manual recommends the higher-order modes for mix buses and key tracks.
- Wow, flutter, Degrade or Chew left on a master. Pitch drift and dropouts on a full mix make everything sound out of tune or broken.
- Low bias by accident. Underbiasing creates the deadzone sound. Great on a lo-fi loop, wrong on a mix bus.
- Tape on the sub. Saturation and the head bump can swell an 808 and blur it in mono. Use Low Cut with Makeup.
- Judging it louder. Match the output to the bypassed level every time.
Where I land (and the disclosure)
CHOW Tape Model is the free tape plugin to install first. Most people only need the Hysteresis section, a fast tape speed and a pulled-back mix; the rest is there when you want the machine to misbehave on purpose.
My own analog color chain doesn't use tape. The Ultimate Master Chain for Plugin Alliance gets its color from tube saturation: its SAT macro blends in parallel tube saturation, and its DENSITY macro adds harmonic richness and body. The stages are explained in the Plugin Alliance mastering chain. I'm biased; if tape is the sound you want, CHOW Tape Model used lightly before the limiter does the job for free.
Analog color before the limiter, already set
THE ULTIMATE MASTER CHAIN FOR PLUGIN ALLIANCE
The analog color chain: 100% Plugin Alliance, with a SAT macro for parallel tube saturation (tube, not tape) and a DENSITY macro for harmonic richness, among 10 macros (LO-CUT, LOWS, MIDS, AIR, DENSITY, SAT, GLUE, MONO-MKR, WIDE, LIMITER). Ableton Live 10 or newer; Logic build sold separately. $29, free updates.
See the Plugin Alliance ChainQuick answers
Is CHOW Tape Model free? Yes. It's open source under the GPLv3, from Chowdhury DSP, with installers for Windows, macOS and Linux on chowdsp.com and an iOS version on the App Store. The manual calls it free as in free beer and free as in free speech.
What are the best CHOW Tape Model settings for a mix bus? A higher-order hysteresis mode (NR8, or STN to save CPU), Speed snapped to 15 or 30 ips, low Drive, Bias toward clean, Wow and Flutter, Degrade and Chew off, and the output matched to bypass. Add Low Cut with Makeup if the low end swells.
Does CHOW Tape Model work in Pro Tools? Yes. It ships in AAX alongside VST, VST3, AU, CLAP, LV2 and a standalone app, so it loads in Pro Tools, Logic, Ableton Live, FL Studio and the rest.
What does Bias do in CHOW Tape Model? It sets the amount of tape bias, the inaudible high-frequency signal real recorders add to the audio. Lower bias widens the hysteresis curve and creates the "deadzone" of underbiased tape; higher bias keeps the tape cleaner.
Written by
CHARLIE CROWN
Founder of Born Creative Audio. Independent artist and engineer: 50M+ streams, 100% owned, never signed a record deal. Also founder of Born Creative Records. Work-for-hire remixes for Sony, Ultra, and Dim Mak. FabFilter featured artist; endorsed by iZotope, Sonarworks, and McDSP. Everything on this site comes from running a real independent music business, not theory.
Stay in the loop
KEEP READING
Production insights, release breakdowns, and independent artist strategies — straight to your inbox.