FAQ

Crackling usually means the CPU cannot fill the audio buffer in time, or a plugin is misbehaving. Raise the buffer slightly, freeze heavy instruments, and quit background tasks. If one channel causes it, bypass inserts until the offender appears.

Share reverbs and delays on sends, convert finished MIDI to audio, and close plugin windows you are not tweaking. Treat CPU like a budget: spend it on elements that define the record.

Export a lossless master first. Many catalogs normalize loudness; aim for musical dynamics and control true peak below 0 dBTP. Generate lossy formats only from the lossless master.

Use both if you can. Speakers reveal room interaction; headphones expose detail and clicks. Check mono compatibility and translate to a car or phone speaker before calling a mix finished.

Layer a midrange transient so the brain hears punch even when subs disappear. High-pass non-bass elements and avoid stereo widening on sub frequencies.

Patterns are reusable musical loops; the arrangement timeline places those patterns in song order. Lean on patterns for hooks and variations, then assemble structure like building blocks.

Save iterative versions, collect samples into the project folder when possible, and mirror to cloud storage weekly. Disasters are rare until they are absolute; backups are professionalism.

Compression raises the noise floor of resonances. De-ess before heavy compression, notch problem frequencies surgically, and use serial gentle EQ rather than one extreme boost.

Leave headroom on the master bus: typically several dB of peak room unless you are mastering in the same project. The mastering stage needs space to shape tone and loudness.

You can, but studio projects are often too heavy. Create a performance version with bounced stems, simplified chains, and tested buffer settings on the exact machine you will use live.

Level-match references to your mix in perceived loudness, not peak meters. AB quickly; long exposure biases your ears. Swap genres if you chase a specific drum texture rather than copying a whole chain.

Only from sellers and URLs you verify yourself. When in doubt, cross-check URLs, checksums if provided, and keep a copy of what you downloaded with its version number.

No: great records have been made with stock tools. Plugins help when they solve a specific problem faster or with a sound you cannot get otherwise. Buy or install new tools when a limitation is real, not when a forum says you are “supposed” to own a stack of compressors.

Use a consistent folder naming scheme, collect external audio into project folders before archiving, and sync or back up to at least two locations. Version major mixes with dates in the filename so you can return to a known-good bounce without guessing.

Use one sample rate per project (often 44.1 kHz or 48 kHz) and match imported files or convert deliberately. Buffer size is a trade-off: lower for tracking with monitoring through the DAW, higher for mixing when stability matters more than round-trip latency.

When changes no longer improve translation (car, phone, second room), not when you are merely tired. Set a time box, bounce, and listen tomorrow. If you keep “fixing” the same band, you are circling, not finishing.

A gentle limiter for loudness reference is fine if you are honest about it. Heavy limiting while balancing masks problems and trains you to mix into a squashed ceiling. For learning gain staging, keep the master chain minimal until the mix stabilizes.

dBFS measures digital peak level relative to full scale. LUFS approximates how loud humans perceive program material over time: streaming platforms care more about LUFS-style loudness than raw peaks. Use both: peaks for clipping, LUFS for comparing musical loudness.

Meters in the DAW may show sample peaks, while inter-sample peaks can exceed 0 dBFS after conversion to lossy formats or on certain DACs. Leave headroom, watch true-peak aware metering if available, and export a lossless master before making MP3 or AAC derivatives.

For fun and rough balance, yes. For final EQ moves and low-end trust, use wired monitors or headphones you know well, ideally checked against speakers. Bluetooth adds latency and codec coloration that skews detail and punch.

Wide stereo bass, phasey 808 samples, or stereo imagers on sub frequencies often cancel when clubs, phones, or radios fold to mono. High-pass non-bass elements, check mono while balancing kick and bass, and consider keeping sub content mono or minimally wide.

Export stems at the project sample rate with identical start times so they line up on import. Label clearly (drums, bass, music, vocals, FX). Remove the mastering limiter from stems unless the engineer asked for a “loud ref”: always include a separate stereo rough mix for context.

Real-time playback may include monitoring-only FX, unprinted automation, or a master chain you forgot was bypassed for the render. Offline exports can also treat plugin oversampling and PDC differently. Solo the master, confirm the exact chain is active, match export length and tail options, and level-match before you chase “tone.”

Dither is very low-level noise added when you reduce bit depth (for example to 16-bit for CD or some delivery specs). It masks the harsh distortion that truncation causes. Stay in 32-bit float inside the DAW; apply dither on the final step when you commit to 16-bit. For 24-bit archival masters, dither is usually unnecessary.

24-bit PCM is the usual choice for tracking: quieter noise floor than 16-bit without the huge file sizes of multitrack 32-bit float unless you have a specific pipeline reason. Pick one depth per project and stick to it; if you import 16-bit stems, leave headroom rather than normalizing them hot. For final delivery, bit depth is a separate conversation from your tracking depth: see the dither FAQ for 16-bit masters.

“Stepped” often means automation points are too sparse, snap-to-grid is fighting musical moves, or MIDI velocities are identical on every hit. Try smoother curves, record automation in real time for one pass, and loosen quantization on fills. If you hear zipper noise when a parameter moves, increase automation resolution where your DAW allows or spread the move across more time. For MIDI, small timing and velocity variation usually beat another humanization plugin.

Think of it as keeping levels honest at every step: trim a screaming virtual instrument before it hits the mixer, avoid stacking ten plugins all pushed into the red, and leave the master bus breathing room. You are not chasing a single “correct” dB value on every meter: you are preventing noise, distortion, and surprise clipping so EQ and compression behave predictably. If one fader lives at −18 dB while another sits at +6 dB, fix the source or the clip gain first.

Layered drums sum like any other audio: if transients fight, you lose punch in mono and on small speakers. Solo the stack, fold to mono, and nudge one layer a few samples until the body returns; zoom the waveform if your ears are tired. High-pass the “click” layer so only one source owns the sub. Try inverting polarity on one layer briefly; if it gets worse, undo and adjust time instead. Stereo wideners on low-frequency layers are a common mono killer.

Normalizing raises everything to a target peak, which can be useful for very quiet field recordings or wildly mismatched stems, but as a default habit it fights gain staging: you push noise and breath along with the note you cared about. Prefer trimming clip gain or faders while watching headroom; use normalize when you have a specific problem (“this take is 20 dB under the rest”) rather than “because the button exists.” For mastering loudness, normalization is not a substitute for intentional dynamics and limiter choices.

Mud usually stacks from long decay + full-range returns + too many different verbs. Share one or two reverbs on sends, high-pass the return so kick and bass are not smearing into a cloud, and add pre-delay so the dry hit speaks before the wash. Automate send levels down in dense sections instead of leaving every channel equally wet. If the mix clears when you mute reverbs, the arrangement may also be too busy: reverb rarely fixes overcrowding.

You duplicate (or bus) the signal, smash one path with heavy compression, then mix that wet compressed layer under the dry track. Drums get punch and length; vocals get density without losing every consonant. It is not magic; if the dry part is already wrong, parallel compression just makes a louder wrong part. Start with modest blend ratios; solo the compressed path only to hear artifacts, then judge in the full mix.

Treat delay like reverb: high-pass the return, shorten feedback when the arrangement fills in, and use pre-delay or ducking so echoes do not step on the next word or snare hit. One send with a filtered tape-style delay often beats three ping-pong presets at once. Automate send level down in choruses; clutter is often “too much at the same time,” not the wrong plugin brand.

No. It is an independent educational project. Product names on these pages are used in a descriptive sense only.