Questions, answered

A Windows desktop app for making chip music that is yours. Four pillars: the Sounds & Instruments Generator is a real SID and AY/YM synthesiser for designing original, copyright-free sounds from scratch; Renata Studio is a chip-native DAW for composing original tracks; the Song Generator composes complete original songs on your machine; and the Player/Converter plays 300+ retro chip and tracker formats and turns any of them into multi-track MIDI, per-channel stems, audio, sample packs, or another chip's format. It reads chip-level register data — not recordings — so every voice is reconstructed independently. Fully offline.

Over 300 formats across 10 sound chips and 7 CPU emulators. Native: YM, AY, SNDH (Atari ST), SID (Commodore 64), VGM/VGZ (Sega Genesis/Master System), NSF (NES), GBS (Game Boy), MOD (Amiga ProTracker). Tracker formats via libxmp: ~70 formats including XM, S3M, IT, MED, OctaMED, Composer 669. Exotic Amiga formats via UADE: ~200 formats including TFMX, Hippel, Future Composer, AHX, JamCracker.

No, Renata works fully offline once installed. The only operations that need the internet are clicking Browse Collections links (which open archive websites in your browser) and downloading the optional 15,000-preset MilkDrop library for Light Synth.

Pick one of the four tiles on the welcome screen. Studio opens the chip-native DAW. Convert turns a chiptune into MIDI, audio, or another chip format. Play loads a tune with the visualiser and playlists. Generate opens the Sounds & Instruments Generator, which is also the door to the Song Generator. Everything is reachable again later from the menu bar and the Converter / Generator / Studio switcher at the top right.

Use the Import menu — it lists eight curated archives that open in your browser: HVSC (60,000+ C64 SID tunes), SNDH Archive (5,000+ Atari ST), Project2612 (681 Sega Genesis sets), Zophar NSF and GBS, Modland (400,000+ tracker modules), VGMRips, and The Mod Archive. Download an archive, then use Import → Import Local Archive to extract and turn it into a playlist.

Click "Browse File" to pick a file, or drag-and-drop a chiptune onto the main window. Then click Play. The visualiser above the player shows real-time spectrum or waveform output as the track plays.

Yes. Some .ay files are not AY chip music at all — they are 1-bit "beeper" tunes, where the Z80 rapidly toggles the speaker port to fake multiple voices on hardware that officially had none. Renata detects which kind an .ay file is by running a short emulation pass (many speaker edges and no AY register writes means beeper), then plays it through a free-running Z80 with the 1-bit speaker level time-integrated per sample. Beeper tunes can be transcribed to MIDI as well, via pitch detection on the rendered audio.

Some chiptune files contain multiple tunes packed together — common in SNDH, NSF, GBS, and SID. When Renata detects multiple subsongs, a "Subsong" selector appears below the playback controls. Pick a number from the dropdown to choose which tune to play.

Drag the progress bar (the slider with the orange thumb on Amiga, or the cyan thumb on Modern) to any position. Renata pre-renders the audio behind the scenes so seeking is instant — no waiting for the chip to re-emulate from the start.

If a playlist is active, the next track loads and plays automatically. Otherwise playback stops at the end. Many chiptunes loop indefinitely — Renata detects when the audio wraps back to its starting point and treats that as the end so a track does not play forever.

A lot of chiptunes are designed to loop infinitely — Game Boy and NES title-screen tunes, for example. Renata caps these at the Conversion Duration setting (default 15 minutes) for MIDI export, and uses loop wrap-around detection during normal playback. If a file does not have a recorded duration, Renata estimates one or applies the cap.

Browse a file, set the Conversion Duration in seconds, then click Convert and pick MIDI from the dropdown. The MIDI is saved next to the input file by default — tick "Custom output location" to pick a different folder. The conversion log shows progress and where the output landed.

Four things besides MIDI. Audio: WAV (16-bit), MP3 (320 kbps), and FLAC (lossless 16-bit) renders of the full song. And under "Chiptune (re-voice)", eight retro chip targets — Commodore 64 .sid, NES .nsf, Game Boy .gbs, ZX Spectrum .ay, Atari ST/AY .ym, Sega PSG .vgm, ZX/AY .vgm, and Amiga ProTracker .mod. So any of the 300+ input formats can come out as any of those.

No, and this matters. Re-voicing re-performs the notes — pitch, timing, dynamics, voice separation, folded to the target chip's channel count with voice-spread and chord-to-arpeggio — on Renata's own instruments and its own from-scratch music driver. It is a faithful cover, not a bit-exact transcode. The source's exact timbres, effects, envelopes and samples (a SID filter sweep, an Amiga digidrum) are not reproduced on the target chip. Copyright in the underlying composition still belongs to whoever wrote it.

Yes. Load an .mp3 and convert to MIDI — Renata runs a Basic Pitch neural transcription (running entirely on your machine, no cloud) to find notes, then applies the same tempo, key and time-signature detection as a chiptune conversion. It splits the result into Bass, Lead and Harmony tracks. Accuracy depends on the source: clean, sparse material transcribes far better than a dense full mix.

Collapses the transcription's polyphonic output to a single melodic line — when two notes overlap, the louder one wins. Useful when the MIDI is destined for a single chip channel or a monophonic synth. It only affects MP3 conversions; chiptune formats are unaffected.

A multi-track MIDI: one track per chip voice with notes, velocity, and pitch bend. Plus CC#7 volume automation, CC#74 filter cutoff, and General MIDI program assignments (with YM envelope routing to GM 80/81/90). Best results come from loading it in a chip-VST like chipsounds, Plogue, or Furnace VST.

How many seconds of the track to convert to MIDI. Default is 900 seconds (15 minutes). Looping tunes do not loop forever in the MIDI — they stop at this limit. Listen to the preview to figure out a sensible length, then set the value before converting.

Snaps MIDI note start times to a musical grid (1/16 or 1/32 note) for cleaner piano-roll alignment in your DAW. Uses ±30% tolerance, so notes only snap when they are obviously close to a grid point. Pick "Off" if you want every register write timed to the original frame.

Bypasses every MIDI cleanup pass: no tempo or key detection, no time-signature detection, no ghost-note pruning, no quantisation, no GM program mapping, no CC automation. The output is plain notes-only MIDI at 120 BPM. Useful for debugging, for chip-VST samplers that prefer raw data, or for getting the pre-cleanup behaviour back. Only affects YM, AY, and SNDH conversions.

Preserves CC#7 volume, CC#73 pulse-width, and CC#74 filter events even when a voice is not currently producing sound. Useful for chip music that uses silence rhythmically. Produces denser MIDI files. Ignored when "Raw register dump" is also ticked — raw mode always wins.

Chip music fakes chords by cycling 2–4 notes very fast on one channel. This option detects those staircases (C-E-G cycled at 16ths or faster) and replaces them with a sustained chord plus a CC#16 rate marker, so the piano roll shows a chord instead of a blur. Arpeggiator plug-ins that read CC#16 — Cthulhu, Ripchord, Ableton's Arpeggiator — can re-create the original cycling. On by default; affects YM, AY and SNDH conversions.

Renders each tonal voice, runs it through a bundled sample-role classifier, and names the MIDI track something readable like "Square Lead — Voice A" instead of "AY Channel A". Voices the model is not confident about (below 40%) keep their original label, so an unclear voice never gets a wrong name. On by default; affects YM, AY and SNDH conversions.

Picks a smarter General MIDI instrument per channel instead of one hardcoded default. On Sega VGM it decodes the FM algorithm — synth bass for a 2-stack, pad for a 1-to-many split, tubular bells for the bell topology, rock organ for four parallel carriers. On Amiga/Paula channels the bundled classifier listens to each channel and maps Lead, Bass and Pad roles. Drum or unrecognised channels keep the chip default. On by default.

After grid quantisation, each note is run through a bundled timing model that nudges it back toward its original position, so deliberate swing, shuffle and chip-specific micro-timing survive the quantiser instead of being flattened. Falls back to doing nothing if the trained model is not bundled. On by default.

"Detect section boundaries" re-renders the song after conversion, runs a self-similarity novelty curve to find phrase boundaries (intro, verse, chorus, bridge, outro) and inserts neutral "Section A", "Section B" … MIDI text events at each one. DAWs that display MIDI text — Reaper, Logic, Cubase — show these as arrangement markers. Currently wired for VGM and GBS conversions.

The General MIDI piano, lead, and bass instruments do not sound anything like a chip waveform. Load the MIDI in a chip-aware VST: chipsounds (commercial, very accurate), Plogue chipspeech for vocal-like leads, or Furnace VST for full multi-chip emulation. Renata embeds GM program changes that pick sensible chip-style instruments where possible.

A few possibilities. STE-flagged or DSP-flagged SNDH files use digital audio rather than YM2149 register writes, so they cannot be MIDI-converted — the conversion log will say so. Some UADE formats (about 9 of 40 tested) produce zero MIDI activity. And some files genuinely do not contain note data. Try the raw register dump option to confirm whether anything was captured.

Per-channel WAV files isolating each voice of a chiptune. A SID has 3 voices, NES has 4 (two pulse, triangle, noise), Sega Genesis VGM has up to 11 (six FM channels, DAC, four PSG). Stems are clean — no mixing or crosstalk — because each voice is rendered independently from the chip emulation.

Browse a chiptune file, then Export → Export Stems. Pick a folder. Each voice becomes its own WAV file. The conversion log shows what was created. Stems are also bundled inside Sample Pack output when you generate one.

Yes. Export menu → "Stem Rate" toggles between 44.1 kHz and 48 kHz. 48 kHz is preferred for video-tied audio work (matches most video projects); 44.1 is the chiptune default and matches CD quality.

Export menu → "Stereo Stems" toggles a subtle L/R panning per voice. Off (default) renders mono per channel. On adds a small stereo spread useful for DAW workflows where you want a wider sound out of the box.

Either no file is loaded, or the format does not support stem export yet. Most native formats and tracker modules are stem-capable; some UADE formats are still being added. Hover the menu item — the tooltip explains which case applies.

A ZIP containing per-voice stems, onset-detected one-shots, an SFZ instrument file mapping samples to MIDI notes, and a README listing every sample with its duration, file size, detected BPM, and detected key. Optionally embeds WAV loop points (smpl chunks) for samples that should loop seamlessly.

One-Shots splits each voice into individual notes via onset detection — best for sampling and DAW instruments. Stems exports full-song per-voice audio — best for remixing and layering. Both produces stems and one-shots as separate packs in the same output folder. Pick the mode in the Sample Pack Options dialog.

Controls how aggressively onset detection splits a voice into notes. Low catches only strong attacks (fewer splits). High catches soft attacks too (more splits, but may over-split sustained notes). Medium is the default and works for most material.

Filters out one-shots that are too short or too long. Default is 100 ms minimum (anything shorter is usually noise) and 5000 ms maximum (anything longer is usually a loop, not a one-shot). Adjust if you are sampling unusually short or long material.

SFZ is an open instrument format used by software samplers. Renata maps each WAV in the pack to a MIDI note: melodic samples land on their detected pitch, drum/FX samples sequence from C1 upwards. Load the .sfz file in sforzando, sfizz, ARIA, Ableton Sampler, Kontakt, or any SFZ-compatible host to play the samples from your keyboard.

A WAV header chunk that records loop start and loop end sample positions. MOD samples often loop, and Renata preserves their loop points in the smpl chunk so hardware/software samplers (Kontakt, EXS24, sforzando, Ableton Sampler) can read them and loop the sample seamlessly.

Onset autocorrelation in the 60–240 BPM range, capped at 15 seconds of analysis for speed. The detected value is embedded in stem filenames like "Voice_A_125bpm.wav", and shown in the SFZ comments and the README. If detection is uncertain, the README says "Not detected" and filenames omit the BPM.

Autocorrelation pitch detection on overlapping windows builds a 12-class chromagram, then Krumhansl-Schmuckler major and minor profiles are correlated against it to pick the best fit. The result (e.g. "C# minor") shows up in the SFZ and README. Falls back to "Not detected" when confidence is too low.

Yes. In Renata Studio, load a song and pick File → "Generate sample pack from song" — it runs the same sample-pack engine over the whole loaded track and lists the result in the Studio Browser so the samples are auditionable straight away. See the Renata Studio FAQ section for details.

Run an operation across an entire folder at once. Four batch operations live in the Tools menu: Batch MIDI (convert every supported file to MIDI), Batch Convert (render every file to WAV, MP3 or FLAC audio), Batch Sample Pack (generate a sample pack ZIP for every file), and BPM Analysis (detect tempo for every file and optionally organise by BPM range).

Tools → Batch Convert. Pick the output format — WAV, 320 kbps MP3, or lossless FLAC — then choose an input folder and an output folder. Renata scans recursively and renders every supported file. Useful for turning a whole chiptune archive into something any media player or DAW can open.

Tools → Batch MIDI → pick the input folder containing chiptune files → pick an output folder for the MIDI files. Renata scans recursively, converts every supported file, and shows progress with file count and ETA. Failed files are logged but do not stop the batch.

Tools → Batch Sample Pack → pick options (One-Shots / Stems / Both, sensitivity, min/max length) → pick the input folder → pick an output folder. Each chiptune produces its own ZIP in the output folder, named after the source file.

Non-UADE formats run in parallel using rayon, capped at half your CPU cores to manage memory. UADE formats (the exotic Amiga ones) run sequentially — the 68000 emulator has global state that is not thread-safe. A folder mixing tracker modules and TFMX files will mostly run in parallel until it hits the UADE files.

Tools → BPM Analysis. Scans a folder, detects tempo for every supported file, displays the results in a sortable table (sort by file, format, BPM, or BPM range), and offers an "Organise Files" button that copies files into BPM-range subfolders like "120-129_BPM/" — turning an unsorted archive into a tempo-organised sample library. Originals are never modified or moved.

A separate full-screen visualisation window with 428+ audio-reactive presets across 14 categories. 103 native presets are bespoke Three.js and GLSL scenes; 325 are curated MilkDrop presets running natively in WebGL via the embedded Butterchurn engine. All react in real time to bass, mid, treble, and beat detection from the Rust backend.

Click the "Light Synth" button in the player section of the main window. It opens as a separate window so you can keep it on a second monitor while you work in the main app. Close it the same way — clicking Light Synth again, or clicking the close button inside the window.

Space pauses auto-cycle. R picks a random preset. L locks the current preset (auto-cycle skips it). Arrow keys move to the previous or next preset. P opens the preset panel. F toggles fullscreen. B toggles borderless VJ mode. T toggles theme lock (preset colours stop following the app theme). S toggles the FPS overlay. Plus and minus adjust audio sensitivity. Esc closes.

Native presets are built specifically for chiptune visualisation — the Retro & Chiptune category includes Demoscene Plasma, Channel Scopes, Raster Bars, Copper Bars, Vector Balls, and 10 more. MilkDrop presets come from the legendary Winamp visualiser library, running in modern WebGL via Butterchurn — psychedelic, geometric, organic, and abstract styles.

Open the preset panel (P key or the gear button), find the MilkDrop Library section, and click Download. Renata fetches the 15,000-preset ansorre collection from GitHub and registers them as a new MilkDrop Downloads category. You can also click Import Folder to load your own .json or raw .milk files — .milk files are converted to Butterchurn JSON automatically.

Yes. Toggle borderless mode (B key) so the window has no chrome. Lock a preset you like (L key) or let auto-cycle run. Toggle theme lock (T key) so colours do not shift when the app theme changes. The FPS counter (S key) helps you spot performance dips in front of an audience.

Loads any WAV file and applies up to 15 audio effects — Arcade Cabinet hardware simulation, bitcrusher, compressor, distortion, low-pass and high-pass filters, reverb, delay, chorus, noise reduction, time stretch, pitch shift, normalise, fade in/out, reverse — plus editing tools (Trim, Cut, Copy, Paste, Delete, Crop Silence, 20-level Undo) and 16/24/32-bit float output.

Tools → Sample Editor opens it in a separate window. The editor has its own theme — it picks up whatever theme the main app has active.

Load a Reference sample (e.g. a record or film clip with a sound you want to replicate) and Renata analyses its sonic characteristics — spectral centroid, bit depth estimation, cabinet detection score, etc. Then load a Working sample, click "Apply Detected Effects", and Renata applies the same effect chain. You can also pick effects manually from the dropdown.

Yes — 20-level undo. Press Ctrl+Z, or use the Undo button in the edit toolbar. The undo stack covers Trim, Cut, Paste, Delete, and Crop Silence operations.

The Render dropdown offers Auto (matches input), 16-bit PCM, 24-bit PCM, or 32-bit float WAV. 32-bit float is recommended for further DAW processing because it has effectively unlimited headroom; 16-bit is the standard for finished samples destined for hardware samplers.

An 8-stage hardware simulator: YM3012 12-bit DAC quantisation → reconstruction LPF → 5-band speaker EQ (boxy 350–500 Hz boost, clarity dip 3–4 kHz, cascaded 24 dB/oct LPF) → dynamics compressor → amp saturation and speaker breakup → cabinet reverb with HF damping → LFSR noise plus 50/60 Hz hum → soft output limiter. Three cabinet sizes (Small/Medium/Large) plus age and volume controls.

Spectral subtraction with automatic noise profiling. Renata finds the quietest section of audio, builds a noise magnitude profile from it, then runs an STFT pass with 75% overlap subtracting that profile from every block (with a floor to avoid musical artifacts). One Strength slider goes from gentle cleanup to aggressive removal.

A built-in synthesiser for creating your own brand-new, copyright-free chiptune sounds from scratch — for commercial use or just for fun. Renata fully emulates the relevant sound chip (quirks and all), so the sounds are authentic to the hardware. You design a patch, tweak its parameters, preview it, and export it as a WAV. It is one of Renata's four pillars, alongside the Player, the Converter, and Renata Studio.

Click the Generate tile on the welcome screen, or use the Generator entry in the top navigation. It opens in the main window as its own view, switchable any time from the top-right nav (Converter & Player / Generator / Studio).

The Commodore 64 SID and the AY/YM2149 (ZX Spectrum / Atari ST / Amstrad) families, with more planned. Each is a faithful emulation of the real chip — including the quirks that give these chips their character — so a patch sounds the way it would on the original hardware.

All the chip's real parameters — for SID that includes the waveform, ADSR envelope, pulse width, and filter cutoff/resonance; for AY the tone, noise, envelope, and volume. You shape the sound directly, the way a chip musician would on the hardware.

Furnace-style parameter macros: a 50 Hz step sequence that automates a parameter over time — for example sweeping the SID pulse width or filter cutoff, or the AY volume or noise period — with optional looping. Draw the macro on the 64-step grid in the macro editor, or start from the Flat / Ramp / Saw / Zigzag presets. Macros are what turn a static patch into a moving, animated sound.

You can stack 1 to 4 voices into a single instrument and play them together — layer two detuned SID pulses for a fatter lead, or blend a tone with a noise hit. Each voice has its own parameters plus mute and solo, so you can audition layers in isolation. Renata normalises the mix so stacking voices does not clip.

Yes — use "Import .dmp…" in the Generator. SID instruments import directly. NES PSG .dmp instruments are decoded and re-voiced through Renata's SID synthesis (with a warning), since Renata's Generator targets SID/AY rather than the NES APU. The imported parameters land in the editor so you can tweak them further.

Press Preview to hear the current patch (including any macros and layered voices) rendered by the chip emulator. Preview plays through Renata's own audio path, so it reacts in the visualiser too. Adjust and preview again as often as you like — nothing is written to disk until you export.

Export WAV renders the patch to a WAV file. For a layered instrument you can export with stems — a mixed WAV plus one WAV per voice — so you can balance the layers later in a DAW. Output goes to your chosen folder, and Renata opens it for you when the render finishes.

Yes. Save any patch as a user preset, then reload it from the preset library later. The browser lets you filter by chip kind (SID / AY / sample) so you can find the right starting point quickly. Presets store the full patch — parameters, macros, and layers.

Yes. Everything the Generator produces is your own original sound, synthesised on Renata's own chip emulation — there are no sampled copyrighted recordings involved. Use the results in commercial tracks, games, videos, or anything else, with no licensing strings attached.

A window that composes complete, original chiptunes for you. You start from a named scene, Renata writes you one song, and then you tell it what to change — in plain English — until it is the song you wanted. Open it from the Song Generator button in the Sounds & Instruments Generator, or Tools → Song Generator.

The entry screen offers six starting points rather than a wall of dials: Rainy night (slow, minor-key, unhurried), Boss battle (fast, dark, relentless), Title screen 1986 (bright and grand), First level (bouncy and easy to hum), Deep space (still, wide and patient), and Demo party (fast, busy, showing off). Pick one, read the plain-English summary of what it will make, and press Compose. An Advanced door behind the templates exposes the goal dials if you want them.

Deliberate. You are not picking from four strangers — you are working on one song. One press composes one proposal; you then shape it. An earlier build ranked several candidates and showed you the "best" one, but that ranking was measured against the owner's own judgement and failed its gate badly, so it was removed rather than left in to quietly pick for you.

Tap one of the six named moves under the song: More memorable, Reduce the intensity, More hopeful, Keep everything but the bassline, 20% more emotional, or More surprising. Each one states exactly what it will keep before you press it — "a fresh bassline under the same song; melody, drums and harmony replay as heard" — and that promise is enforced in code, not hoped for. Everything not named is frozen note-for-note.

Every version you pass through is kept in a linear journal, oldest first, so a wrong turn costs one tap. Step back to any earlier version and carry on from there. "Clear journal" wipes the history — it is the one destructive action in the window, so it asks twice. The journal is also saved inside a .renata file.

Yes — save the whole session as a .renata file. That stores the song, its full journal of versions, and its provenance (the seed, the settings and the moves that produced it), so reopening it puts you back exactly where you were, including the ability to keep rewinding.

Six chip profiles: YM2149/AY (Atari ST and ZX Spectrum), SN76489 (Sega/BBC), NES, Game Boy, SID (Commodore 64), and Amiga Paula using real samples. Nine genres: Arcade/Shmup, Platformer, Title Theme, Ballad, Techno/Dance, Rock, Funk, Ambient, and Demoscene. Every song is composed within the real chip's channel budget, so what you hear is what the hardware could actually play.

No AI and no cloud — the composer is a deterministic music-theory and genre-grammar engine that runs entirely on your machine. Songs made with the synth chip profiles are original material synthesised on Renata's own chip emulation: they are yours, copyright-free, for any use. (On the Amiga profile the composition is yours too, but if the samples came from someone else's song, the samples themselves may be copyrighted — see the samples question below.)

Every generation is driven by a seed number. The same seed with the same settings reproduces the exact same song, forever — so note the seed of anything you like. The dice button clears it so the next press rolls fresh; the seed box also accepts a pasted seed from an earlier session. Seeds are stored in a saved .renata file too.

The dials, for when a template is not enough: genre, key, complexity, and per-part emphasis for lead, bass, drums, harmony and SFX, plus the Held-notes control and a form editor where you draw the song's energy arc across its sections. Everything you change is reflected back as a plain-English summary before you press Compose, so you can hold the promise against what you hear.

The short-vs-long-note mix. Slide it right and the bass holds droning whole-bar roots (the classic Amiga/Atari trick), the harmony can sustain chords on top, and the lead plays legato with long held notes. Slide it left for short, pumping chip writing. It applies to every chip and genre.

The Amiga · Paula profile composes with real instrument samples. Load a Future Composer (.fc) or ProTracker .mod song and Renata rips its instruments into a kit (kick, snare, hat, bass, lead, pad); preview any slot, swap samples from the pool, add a second song's samples, or import your own WAV files — or just hit Generate with nothing loaded and Renata builds a starter kit from its own rendered chip instruments. Note: samples ripped from someone else's song may be copyrighted — fine to experiment with, but use your own samples (or the starter kit) for anything you plan to sell.

Always: MIDI, WAV (16/24-bit), MP3, and Open in Studio for full editing. Plus the chip-native format of the profile it was composed for: .ym/.ay/.vgm for YM2149/AY, .vgm for SN76489, .nsf for NES, .gbs for Game Boy, .sid for C64, and .mod with the actual samples embedded for Amiga — including any effects, which are baked into the sample data since tracker formats carry no effect processors.

A chip-native DAW built into Renata — the fourth pillar alongside the Player, Converter, and Generator. You arrange clips on a timeline, edit notes in a piano roll, and play everything back through authentic emulated chip voices (Commodore 64 SID, NES, Game Boy, YM2149 and more) with a full mixer, effects, sends, and session view. It is a place to write and produce, not just convert a file.

Click the Studio tile on the welcome screen, use the Studio entry in the top navigation, or pick Tools → Renata Studio. It opens as its own separate window so you can work in it independently of the main player.

Use the File menu. "Open chiptune" runs any of the 300+ supported formats through the converter and drops the result in as editable MIDI clips, one track per voice. "Load stems" brings a track in as per-channel audio tracks. From there you can edit, rearrange, and re-voice freely.

Yes. The Studio File menu exports your arrangement to MIDI, WAV, and MP3, and to seven chip formats: .ym (Atari ST / AY), .vgm (Sega PSG and ZX/AY), .mod (Amiga ProTracker), and four native driver formats — .sid (Commodore 64), .nsf (NES), .gbs (Game Boy), and .ay (ZX Spectrum / Amstrad CPC). The driver formats are tiny programs with a music driver baked in, so they run on the real hardware or any emulator.

.ym, .vgm, and .mod are data formats — a register log (.ym/.vgm) or a sample-based module (.mod) that a replayer plays back. .sid, .nsf, .gbs, and .ay are native CPU driver formats: each contains a compact music driver Renata assembles for that machine’s actual processor (6502 for SID/NSF, SM83 for GBS, Z80 for AY) plus your note data. Every driver is written from scratch in-house, so exports are clean for commercial use.

Yes — load any chiptune as MIDI clips, then export to any supported chip format. Because MIDI is the hub, that is roughly 300 input formats by 8 chip outputs. Be clear on what this is: Renata re-performs the notes (pitch, timing, dynamics, voice separation, folded to the target chip’s channel count with voice-spread and chord-to-arpeggio) on its own instruments and driver. It is a faithful re-voicing, not a bit-exact transcode — the source’s exact timbres, effects, envelopes, and samples (a SID filter, an Amiga digidrum) are not reproduced on the target chip.

Load a song, then File → "Generate sample pack from song". An options dialog lets you choose contents (one-shots, stems, or both), detection sensitivity, min/max length, and loop options; pick an output folder and Renata slices the entire track into a pack. The result is registered in the left Browser rail, auto-selected, and every sample is auditionable in place — with a live progress banner while it generates and a ★ favourites filter once it is done.

Appearance menu → pick from 8 themes. Modern (the default) is a clean Neon Retro-Future look; the seven Retro Themes are original designs evoking the eras of 8-bit and 16-bit home computers, consoles and handhelds.

Mono Line (16-bit high-res desktop — green and white), Breadbin Glow (8-bit home computer look — blue/purple with green accents), Copper Dreams (16-bit multimedia desktop — blue background, white header, orange accent), Rubber Key (British 8-bit micro — black with rainbow loading stripes), Mega Blast (16-bit console — deep navy with gold accents), Grey Tray (8-bit family console — console grey with red), and Handheld Green (green dot-matrix handheld — 4-shade green LCD with magenta).

Yes — Appearance → Customise Colours. Three pickers: Background, Text, and Accent. Secondary and tertiary backgrounds are auto-derived (lighten +15 and +28); button text colour is auto-calculated for contrast (light on dark accents, dark on light). Custom colours overlay any theme and persist across sessions. Click "Reset to Theme Defaults" to revert.

18 individually-controllable screen effects simulating a vintage CRT monitor: scanlines, RGB mask, shadow mask, aperture grill, scan bar, curvature, vignette, raster wave, flicker, RGB shift, noise, interlacing, jitter, monochrome (green/amber/white phosphor), afterglow, and more. Open from Appearance → Retro Effects. Four presets (None, Arcade, DOS Terminal, 80s/90s TV) give curated combinations.

Appearance → Loading Animations toggles the themed demoscene animations that play over long operations — the matrix rain, raster bars, copper bars and tape-loading stripes described in the Import Collections section. Turn them off if you would rather have a plain progress indicator.

Appearance → "Window & Icon Animations" toggles every animation off. For finer control, open Appearance → Animation Settings: a global on/off, a speed multiplier (0.5×, 1×, 1.5×), and a reduced-motion override (Auto follows the OS setting; Always forces motion off; Never ignores the OS setting). Settings persist in theme_settings.json.

Appearance → Full Screen expands the main window to fill the entire screen. The window is also resizable normally — drag any edge to any size you like. Renata starts at 900 px wide with monitor-height (minus the taskbar) on every launch.

Import menu → pick a Browse Collections link (HVSC, SNDH Archive, Project2612, Zophar NSF or GBS, Modland, VGMRips, Mod Archive). Renata validates the link is live, then opens the download page in your browser. Download the archive yourself, then use Import → Import Local Archive to extract it and create a playlist.

Renata validates each URL is live via a HEAD check before opening. If a site is down or has moved, the conversion log shows a clear error with the site name instead of opening a broken page. Most chiptune archive URLs are stable but a few have moved over the years — get in touch through renatachiptune.com if you hit a dead link and it will be fixed in the next build.

Import → Import Local Archive → pick the .zip, .7z, or .rar → choose a destination folder for the extracted files → name the playlist via a themed dialog. Renata handles nested archives (Project2612 is a .7z containing 681 .zip files — all extracted in up to 5 recursive passes), scans for supported formats, creates the playlist, and opens the Playlists window with the new playlist selected.

Import → Import Folder → pick the folder → name the playlist. Renata scans recursively (still extracting any nested archives it finds), creates the playlist, and opens it. Use this when you have already unpacked an archive yourself.

Each theme has a demoscene-styled canvas animation: matrix rain on Modern, raster bars on Mono Line, a character-set border on Breadbin Glow, copper bars on Copper Dreams, tape-loading stripes on Rubber Key, speed lines on Mega Blast, colour bands on Grey Tray, falling tiles on Handheld Green. All use the active theme font and palette.

Export menu → Export Video. Pick a platform preset or Custom. Tick which Renata windows to record (main window, Light Synth, playlists, etc.). Click Start Recording, do whatever you want recorded, click Stop. Renata muxes audio (tapped from the playback pipeline for bit-perfect sync) and video into an MP4.

Built-in presets for YouTube (1920×1080 60 fps), YouTube 4K (3840×2160 60 fps), YouTube Shorts (1080×1920 60 fps), TikTok (1080×1920 30 fps), Instagram Reels (1080×1920 30 fps), Instagram Feed (1080×1080 30 fps), X Landscape, X Vertical, Facebook, LinkedIn, plus Custom. Each preset sets the right resolution, frame rate, bitrate, and aspect ratio for that platform.

You pick the save location at the start of the recording. Renata writes the final MP4 there, plus a diagnostic report on your Desktop covering encoder used (NVENC/AMF/QSV/software), fps achieved, bitrate, and A/V sync timing. Useful for troubleshooting any quality or sync issues.

Hardware encoders (NVIDIA NVENC, AMD AMF, Intel Quick Sync) are auto-selected based on your GPU and are very fast. Software encoding falls back when no hardware encoder is available — much slower and CPU-heavy. Long recordings produce big files; for social media, keep clips short (TikTok caps at 180 s, Reels at 90 s, Instagram Feed at 60 s).

The black canvas above the player. It reads real-time audio output from the chip emulation and draws bars or a waveform synchronised to the audio. Frequency bars and bass/mid/treble energy values come from the Rust backend so they are computed once and shared with Light Synth too.

Three: Spectrum (frequency bars), Oscilloscope (raw waveform), and Combined (both layered). Pick from the dropdown above the visualiser. The mode also affects what Light Synth sees, since both share the same audio data feed.

It only animates while audio is playing. If a track is loaded but paused, or the file has not started yet, the visualiser stays static. Click Play. If you have started preview stems audition, the visualiser also reacts to that audio.

Click "Playlists" next to the file browser button in the main window. The Playlists window opens as a separate Tauri window so you can browse your library while playback continues in the main app.

In the Playlists window, click the new-playlist button. Name it. Add tracks via "+ Add to Playlist" in the main window after loading a file, or by drag-and-dropping files (or whole folders) onto the Playlists window. You can have any number of named playlists.

Marks a track as a favourite. The Favourites filter at the top of the Playlists window shows only starred tracks across your whole library — useful for jumping straight to your go-to material no matter which playlist holds it.

Every playlist auto-exports as M3U to %APPDATA%\com.renata.app\Playlists. You can also import M3U files via the Playlists window. The auto-export means you have a portable copy of every playlist that any other media player can read.

Filters tracks by title, author, filename, or format — case-insensitive. Works across the full library or within the selected playlist. Clear the search to see everything again.

By SHA-256 content hash, not file path. Move or rename a chiptune and Renata still recognises it; the path updates automatically when the track is next encountered. Two files with identical contents are treated as the same track.

Tools → Analyze Library pre-computes track lengths for your whole library, or for a folder you pick. Working out how long a chiptune runs means actually emulating it, so the first time a track is touched it can take a moment — analysing up front means durations appear instantly everywhere afterwards, in playlists and in the player.

A store of everything Renata has already worked out about a track — its length, detected BPM, key and time signature — so none of it has to be recomputed. Tracks are keyed by content hash, not file path, so moving or renaming a file keeps its cached analysis. Tools → Analysis Cache lets you view it, export it, import one, or verify it against the files on disk.

Analysing a large archive takes real time. Exporting the cache means you can move it to another machine, keep it through a reinstall, or back it up before reorganising a library — and importing it restores every length, BPM and key without re-analysing a single file.

A developer tool. It exports per-voice audio clips and their labels from a folder of chiptunes, in the format used to retrain Renata's bundled sample-role classifier — the small on-device model that names MIDI tracks by role and picks per-patch GM instruments. It is there for anyone who wants to see or improve how that classification works; it has nothing to do with normal music-making, and nothing leaves your machine.

STE and DSP-flagged SNDH files use digital audio not YM2149 register writes — the conversion log will say so explicitly. Some UADE formats (about 9 of 40 tested) are zero-activity and produce no MIDI. And some files genuinely contain no note data. Try ticking "Export as raw register dump" to see whether any data was captured.

Some chiptunes do not record their own length. Renata uses the format-specific tag if present (FRMS for SNDH, etc.) and falls back to a 15-minute cap otherwise. Listen to the preview, decide on the song length, then set Conversion Duration before clicking Convert.

UADE formats stress CPU emulation hard — close other heavy apps and try again. If a specific file glitches consistently, file a bug report with the file attached. Hardware-accurate synthesis (logarithmic DAC tables, 2× oversampling, DC-blocking) is computationally expensive but produces cleaner output than naive emulators.

Theme settings live in %APPDATA%\com.renata.app\theme_settings.json. If that file goes missing or gets corrupted, the default theme returns. Pick a theme again and Renata recreates the file.

The file may not be a supported format, or it may be corrupted. Check the conversion log for an error message. Some exotic formats need a companion data file alongside the main file (TFMX is the famous example — usually a .tfx and a matching .smpl). Renata looks for companions automatically in the same folder.

Look for an error in the conversion log. Common causes: the chosen output folder is read-only, the chiptune file is unsupported by the sample pack pipeline, or a disk-space limit was hit (large multi-voice exports can be hundreds of MB). The diagnostic message names the failing step.

That means faqs-content.js is empty or missing — should never happen in a normal install. If you see this after updating Renata, reinstall to restore the bundled content file.

Mark Wright, on his own, over nine months of spare evenings. It is named after his wife, Renata — a name that also happens to mean "reborn" in Latin, which suited an app about giving old chip music new life. The wordmark on the welcome screen reads "Renata" with a coloured final "a"; that accent uses each theme's palette to tie the brand to whichever retro look you have active.

It is a one-time purchase — you buy it, you own it, you keep it. No subscription, no cloud account, nothing phones home. There is no cut-down free tier: it is the whole app or nothing. Renata is pre-launch, and the price has not been set yet — it will be decided with the community and announced with the Kickstarter. Join the waitlist at renatachiptune.com to hear first.

Windows 11 (64-bit) is the current shipping platform. Renata runs on standard hardware — no minimum GPU requirement for the main app. Hardware-accelerated video export benefits from a discrete NVIDIA, AMD, or Intel iGPU; software encoding works on any CPU but is slower.

On the roadmap. The Rust backend is already portable; the main work is platform-specific bits — file dialogs, window capture for video export, hardware encoder access. Tauri 2.0 supports both targets natively, so the path is open.

It depends entirely on where the material came from, and Renata is not legal advice. Anything you make from scratch — a Generator patch, a Studio composition, a Song Generator song on a synth chip profile — is original material synthesised on Renata's own chip emulation. That is yours, copyright-free, for any use including commercial. Anything derived from someone else's tune is not: a sample pack, a stem set, a MIDI conversion or a re-voiced cover of a copyrighted chiptune still belongs to whoever wrote it. Public-domain tunes, your own work and licensed material are fine. Check the licence of the original before you sell anything built on it.

Get in touch through renatachiptune.com — replies come from Mark, the developer, not a support desk. Include the conversion log text and, if a specific file is involved, the file itself; that is usually enough to reproduce a problem straight away.

Yes. Help → Tutorial Videos opens the video series, and there is a growing set on YouTube at @RenataChiptune covering every part of the app — playing, converting, stems, sample packs, batch tools, the Light Synth, the Sample Editor, the Generator and Renata Studio — plus a weekly developer diary. This FAQ list is the fastest reference; the videos are the walkthrough.

Help → Check for Updates. Renata does not phone home on its own — the check only happens when you ask for it, which keeps the "fully offline" promise honest.

Help → About Renata logs the version to the conversion log. The current build is v1.3.0.