SIMLAF

Blind listening test · A/B/X

Can you hear it?

6 short trials. In each one, one of our processors is off (A) or on (B), and both play at the same loudness. X is one of the two: say which. About three minutes. Headphones help.

6 trials. One question each: is X off or on?

  • 6 trials, about 3 minutes
  • Lossless FLAC, loudness matched (AAC where FLAC won't play)
  • Nothing plays until you press play
  • Your answers stay in this browser

How the audio was made

Same take. Same loudness. Only the processor changes.

Every trial is one of our own processors, SIMLAF Compressor, SIMLAF Exciter, SIMLAF Imager, SIMLAF Spectral, SIMLAF Vintage EQ, switched off and on over the same section of the same take. No other maker's product is in this test.

Loudness is EBU R128 integrated (LUFS), measured on the loop as it repeats: before → after matching. True peak in dBTP, the higher of the two versions. Null residual: the level of OFF minus ON after matching, relative to the signal; the lower it is, the less difference is left (Easy above −10 dB, Medium to −20 dB, Hard below). Each run draws 6 of these 8 at random.
MaterialProcessor and settingOFF, LUFSON, LUFSTrue peakNull residualLoop and files
Bright synth leadLUMEN, preset Fifth Element SIMLAF Spectralstock preset Vocal Harshness -12.1 → -18.0 -14.3 → -18.0 -9.2 -6.7 dBEasy 4.00 sOFF · ON
Amapiano groove, bright hatsNGOMA SIMLAF Spectralstock preset Overheads Harsh Hits -20.4 → -18.0 -21.0 → -18.0 -2.1 -15.2 dBMedium 4.29 sOFF · ON
Amapiano groove with log drumNGOMA, kit Midnight Amapiano SIMLAF Vintage EQstock preset Kick Thump -23.3 → -18.0 -20.8 → -18.0 -2.0 -3.7 dBEasy 4.29 sOFF · ON
Amapiano groove with log drumNGOMA, kit Midnight Amapiano SIMLAF Compressorstock preset Parallel Smash -23.3 → -18.0 -24.4 → -18.0 -2.0 -16.2 dBMedium 4.29 sOFF · ON
Amapiano groove, bright hatsNGOMA SIMLAF Compressorstock preset Drum Bus Glue -20.4 → -18.0 -19.1 → -18.0 -1.8 -18.1 dBMedium 4.29 sOFF · ON
Bright synth leadLUMEN, preset Fifth Element SIMLAF Exciterstock preset Synth Bite -12.1 → -18.0 -12.0 → -18.0 -9.2 -23.2 dBHard 4.00 sOFF · ON
Bright synth leadLUMEN, preset Fifth Element SIMLAF Vintage EQstock preset Silky Top -12.1 → -18.0 -11.4 → -18.0 -8.2 -3.7 dBEasy 4.00 sOFF · ON
Bright synth leadLUMEN, preset Fifth Element SIMLAF Imagerstock preset Gentle Widen -12.1 → -18.0 -12.0 → -18.0 -9.2 -28.6 dBHard 4.00 sOFF · ON

Check it yourself: every file in the test is linked in the table (FLAC, 6.9 MB in all), and every measurement is in measurements.json. Null them in any DAW.

The method

  1. The dry take and the processed take come from our own offline renderers, sample for sample: the same notes, the same timing, one with the processor on.
  2. The same loop is cut from both, whole bars or whole phrases. Its last 20 ms are blended into its first 20 ms so it repeats without a click, identically in both.
  3. Each version gets one gain value so both measure -18 LUFS. Nothing else is added: no limiter, no EQ. True peaks stay at or below -1 dBTP.
  4. The files are lossless FLAC, 16-bit with dither, about 882 KB a trial. A browser that can't play FLAC gets AAC at 256 kbps instead (about 266 KB a trial), measured the same way.
  5. In your browser, A, B and X play from the same start time and switch with a 12 ms crossfade, so you keep your place in the loop and no switch clicks. X has its own copy, so switching to X sounds the same whether it is A or B.

What the score means

  • Someone who hears no difference gets each trial right half the time. The page tells you how often pure guessing would match your score or beat it.
  • Equal loudness is the point. Louder tends to sound better, so a processor that wins only by being louder wins nothing here.
  • Laptop and phone speakers hide much of this; headphones are the fair test. Each trial is marked Easy, Medium or Hard by its null residual.
  • We don't collect answers, so there is no leaderboard, no average and no "top 10%". The only scores anyone sees are the ones people choose to share.