Critical listening is a learnable skill. Exercises for hearing the mix and the master, spotting compression and lossy artifacts, and trusting your own ears.
Most of us hear music the way we see wallpaper: it's there, it's pleasant, and we couldn't describe it in any detail afterwards. That's a fine way to enjoy a record. But at some point every person who cares about sound hits a wall: they read that a file is "clearly better", or a headphone is "obviously more detailed", and they can't hear it. Sometimes that's because it isn't there. Often it's because they've never learned what to listen for. Listening is a skill, it's trainable, and it takes weeks rather than years.
Hearing is what your ears do. Listening is what your attention does, and attention is narrow: you can follow one thing closely or many things vaguely. Critical listening is the practice of deliberately pointing that narrow beam at specific parts of a sound, one at a time, until you can find them at will.
The goal isn't to stop enjoying music. It's to have a second mode you can switch on when you want to answer a question: is this master better? do these headphones have a peak? can I actually hear this codec? Then switch it off and go back to the wallpaper.
You can't judge anything against a recording you half know. Choose five to ten tracks you will use for every comparison from now on, and choose them deliberately:
Then learn them. Play each one a dozen times over a week, doing the exercises below. The aim is to know exactly where the bass enters, how the reverb tail on the snare sounds, where the vocal sits in the stereo picture. That intimacy is the measuring instrument.
Each one takes five minutes. Do one or two per session, not all at once; ears tire after twenty or thirty minutes of concentrating.
Play a song and follow only the bass guitar from start to finish. Then the hi-hat. Then the backing vocal. It feels absurdly simple and it's the foundation of everything: the ability to isolate a single element in a mix and hold it. When you can do this reliably, layers you never knew existed start appearing in records you've heard a hundred times.
Close your eyes and point. Where is the singer? The guitar? Is the piano wide or narrow? Does the reverb sit behind the band or wrap around it? Headphones present the stage inside your head; speakers put it in front of you. Learning to place sounds is what lets you notice when a lossy encode collapses the stereo image, or when a headphone with poor channel matching pulls everything a little to the left.
This is the fastest exercise of all. Take a graphic EQ, boost one band by 6 dB, listen, cut it by 6 dB, listen, flatten. Repeat for every band, from 31 Hz to 16 kHz. Within twenty minutes you'll have a felt sense of what "250 Hz mud" and "3 kHz presence" and "8 kHz sibilance" actually sound like. Then have a friend change a band without telling you, and guess. The band map in the equalizer guide is the vocabulary; this exercise is the fluency.
Find two masters of the same album, an early CD and a modern remaster, and play the same song from each with the volume matched. Listen to the drums. On the dynamic master, each hit is a distinct event with air around it; on the crushed one, hits are flattened into the wall of guitars and the chorus doesn't get louder, it just gets denser. Listen for the breath before a verse, the decay of a cymbal, the difference between a loud passage and a busy one. Once you can hear the loudness war, you can't unhear it.
Don't start by trying to tell 320 kbps from FLAC; nobody learns that way. Start at 96 kbps, where the artifacts are obvious, and learn what they sound like: cymbals and hi-hats that swirl or sound like they're underwater, reverb tails that cut off early, a faint pre-echo smear before sharp transients, a stereo image that goes vague. Then go to 128, then 192, and see how high you can still find them. Most people plateau somewhere between 160 and 256 kbps depending on the music and the codec, and that number is worth knowing, because it's the honest answer to "can I hear the difference".
Whenever you compare two things, a sighted test lies to you. Knowing which file is the FLAC makes it sound better; knowing which headphone cost more makes it sound more detailed. Use an ABX tool: it plays A, plays B, then plays one of them at random and asks which. Get it right 9 times out of 10 and you can hear the difference. Score around 50% and you can't, however strongly you felt you could. The feeling of hearing a difference is not evidence; the score is.
Match the volume. A difference of half a decibel is enough to make one version sound "fuller", "clearer", "more dynamic". Louder always wins. Before comparing anything, level-match by ear, or better, by meter, and only then decide. Most gear-shop revelations evaporate at matched levels.
Change one thing. Different file, same headphones, same volume, same track. Different headphones, same file. Comparing a new DAC on new headphones with a new album tells you nothing about any of them.
After a few weeks of this, some things change. You stop buying upgrades that don't upgrade anything. You start noticing mastering quality and seek out the good editions. You can tell within seconds whether a headphone has a treble peak you'll hate in an hour. And you enjoy music more, not less, because there's simply more of it to hear. The wallpaper turns out to have a pattern.
Want to run the first exercise right now? Open the player, queue a track you know by heart, and follow one instrument through the whole thing. Then do it again with the equalizer, one band at a time.
FLACCID plays your local FLAC, ALAC, WAV and more right in the browser: no upload, no account, no streaming.
Launch the Player