Local audio tools for Windows. No subscription. NoDAW Liminal Stem Splitter - $29 once. NoDAW CoProducer Auto Fix: $29 once. reTUNE432 Frequency Fix - $4 once.
Joined July 2026
- Tweets257
- Following114
- Followers16
- Likes50
Solo, the kick slaps and the 808 is huge. Together: one gray thump.
A kick tuned to 54 Hz over an 808 on G1 (49 Hz) is a clash. Decide who owns each hit with arrangement, carving or sidechain before you reach for the limiter.
myaiplug.github.io/coproduce…
Name stems like this: 120bpm_Fm_01_Kick.wav
BPM and key in the name, a number for the order, same start point on every file. Your mixer shouldn't have to ask what tempo the project was. 24-bit WAV, at the session sample rate.
3
In album mode one gain covers the whole album, so the loudness steps you master in are the steps people hear.
Our synthetic six-track album had a 10.6 LU jump from ballad to closer. Nobody planned that.
Meter the record as a set, not as six files.
myaiplug.github.io/coproduce…
2
Sample lines up for one bar, then slides off your drums? Its tempo has a decimal.
In our test a 93.6 BPM loop in a 94 session was 174.6 ms late after 16 bars, more than a 16th. Trim to whole bars, do the math from the length.
myaiplug.github.io/coproduce…
ALT A black drum-pad sampler with blank pads on gold-veined black marble next to a turntable, evenly spaced cyan light lines floating above it and a row of amber light points that start on the lines and slowly drift away to the right.
3
Mix sounds muddy but every track sounds fine soloed? That's the low-mid pile-up.
In our test a master-bus cut took the vocal down 2.9 dB with the mud. Small cuts on just the keys and guitar gave the vocal 2.5 dB more room.
myaiplug.github.io/coproduce…
ALT Several upright panes of glass glowing amber on gold-veined black marble, a thick amber haze piling up where they overlap, with one clear cyan-edged pane in front, studio screens out of focus behind.
8
Vocal compressor starting point: 3:1 to 4:1, attack 10–30 ms, release 50–100 ms, aiming for 3–6 dB of gain reduction on the loud phrases.
Change one knob at a time. If you can hear it working on the quiet words, back off.
3
Your S's didn't get harsher in mastering. They got louder and brighter.
Mastering lifts the whole track and usually adds top end. Sibilance lives around 4.5–10 kHz. De-ess the worst few on the vocal before the limiter, not the whole take.
myaiplug.github.io/coproduce…
2
Want the vocals, drums or bass out of a track? Liminal splits songs into stems on your own PC. Nothing leaves your machine.
Try the free demo: nodaw.gumroad.com/l/Liminal
Unlimited, $29 once: nodaw.gumroad.com/l/LiminalP…
5
Vocal sounding thin? Check where the high-pass sits.
In our test a steep 80 Hz cut took 15.7 dB of rumble and pops off a low voice and cost it 0.3 dB. At 150 Hz the same voice lost 6.9 dB of its fundamentals.
myaiplug.github.io/coproduce…
ALT A black microphone in a shock mount on a short desk stand on gold-veined black marble, with a broad glowing band behind it that rises steeply from the left and flattens out to the right, fading from amber to cyan, studio monitors out of focus in the background.
4
Auto-Tune making your singer sound like a ringtone? It's the retune speed.
In our test 0 ms left 0% of the vibrato and turned slides into semitone steps. At 50 ms notes landed within 3 cents and 86% of the vibrato stayed.
myaiplug.github.io/coproduce…
ALT A silver condenser microphone with a round pop filter on gold-veined black marble, a glowing amber wave floating beside it that ends at a small glowing cyan staircase, studio monitors out of focus in the background.
1
32
Swing for hats: try 54–58% on the 16ths (50% is straight).
Keep the kick and snare on the grid and swing only the hats. Push it much past 60% and it starts to sound sloppy.
4
A halftime slowdown can push a bounce back over the true-peak line.
Stretching changes length and weight, not the ceiling. A file that cleared −1.0 dBTP at full tempo can sit over it after the stretch.
Meter the file you'll actually upload.
myaiplug.github.io/coproduce…
9
808 root notes in Hz:
E1 41.2
F1 43.7
G1 49.0
A1 55.0
B1 61.7
If your key puts the root near 30–35 Hz, try playing it an octave up. Most speakers won't give you anything down there anyway.
8
Your reference sounds better because it's louder.
On our synthetic test pair it was 5.4 LU hotter. Turn it down 5.4 dB first, then compare tone, vocal level and low end.
Copy the balance, not the loudness.
myaiplug.github.io/coproduce…
2
A master at −8 LUFS on a platform that normalizes to −14 gets turned down about 6 dB.
The squash stays. The loudness doesn't. Before you push the limiter another dB, ask who it's for.
4
Three times not to add another limiter:
1. Clipped samples are already non-zero.
2. LUFS is only slightly hot.
3. The real problem is arrangement.
A limiter can park true peak. It can't un-flatten a clipped transient.
myaiplug.github.io/coproduce…
2
Vocal high-pass: start around 80–100 Hz for most voices, 60–70 Hz for a deep one.
Solo the vocal, sweep up until the body changes, back off 10 Hz, then check it with the beat. If it's thin in the full mix, you went too high.
4
A 4-minute 24-bit/48 kHz stereo WAV is about 69 MB. Uploading it just to read one LUFS number is slower than measuring it.
LUFS and true peak follow a public standard (ITU-R BS.1770) any laptop runs in under a second:
myaiplug.github.io/coproduce…
22
Your mix sounds wide on headphones, then the bass vanishes on a phone speaker. Often that's a phase problem.
CoProducer checks phase, clipping, LUFS and true peak locally. Nothing gets uploaded.
Free checker → myaiplug.github.io/coproduce…
Full tool → nodaw.gumroad.com/l/coproduc…
🤖 Made with AI
13
Mix leans left but the meter says it's close? Look at the highs.
Our test mix: 2.0 dB left overall, 11.5 dB left at 8 kHz. All hats and guitar.
The balance knob just pushes your 808 off center. Re-pan instead.
myaiplug.github.io/coproduce…
ALT Cover art for the Lacquer essay on checking left/right balance band by band
14