The short answer
Digital audio is a long list of tiny measurements called samples, tens of thousands of them per second. When you save that audio to a file, you have a choice: keep every last sample, or throw some away to make the file smaller. That single choice is the split between lossless and lossy audio.
A lossless file such as WAV or FLAC keeps all of the data. Decode it and you get back exactly what went in, sample for sample. A lossy file such as MP3 or AAC permanently deletes the parts of the sound a listener is least likely to hear, so the file can be four to ten times smaller. The audio you get back is close, but it is not identical, and the discarded data is gone for good.
What lossless means
Lossless audio is a complete, exact record of the sound. Every sample is stored, so nothing is approximated and nothing is missing. WAV does this by storing the samples raw, with no compression at all. FLAC does it with lossless compression: it packs the same data into a smaller file using clever encoding, the way a zip file shrinks a document, and unpacks back to a bit-for-bit perfect copy. Different sizes, same guarantee of no quality lost.
The catch is size. A raw WAV file runs at about 1,411 kbps for CD-quality stereo, which is a lot of data per second. That number comes straight from the format: sample rate times bit depth times the number of channels. For the full breakdown of that math, see what is audio bitrate. A few minutes of lossless audio is tens of megabytes, which is why it is the format you master and archive in, not usually the one you stream.
What lossy means, and how it shrinks the file
Lossy audio leans on a fact about hearing: your ears do not register everything equally. A loud sound hides a quiet one right next to it, very high frequencies fade with age, and some detail is simply below the threshold of notice. A lossy encoder uses a model of hearing, called psychoacoustics, to find the data you are least likely to miss and delete it. What is left is a much smaller file that still sounds nearly the same to most listeners.
How aggressively it deletes is set by the bitrate, the amount of data kept per second. A higher bitrate keeps more and sounds closer to the original; a lower bitrate keeps less and starts to sound thin or watery. Common music streams sit around 128 to 320 kbps, so a 320 kbps MP3 carries the same song as a 1,411 kbps WAV using roughly a quarter of the data. The important part: this is a one-way trip. Once the encoder discards that data, no amount of converting back to WAV can bring it back.
When the difference actually matters
For a lot of everyday listening on earbuds or a laptop, a good lossy file is genuinely hard to tell apart from the original, and the smaller size is a real advantage for storage and streaming. The gap shows up in specific situations where either quality or size is at a premium.
- Recording and editing. Work in lossless. Every time you re-encode a lossy file you throw away more data on top of what is already gone, so masters, stems, and anything you will edit again should stay WAV or FLAC until the very end.
- Archiving. Keep the lossless copy. It is your only perfect original, and you can always export a lossy version from it later. You can never go the other way.
- Streaming and sharing. Lossy usually wins. A 320 kbps MP3 or AAC sounds excellent to almost everyone and uploads, downloads, and streams in a fraction of the time and data.
- Critical or repeated listening. On good speakers or headphones, and especially on material you will hear over and over, lossless removes any doubt that the format is holding the sound back.
Lossy vs lossless in patchd
Here is the honest boundary: patchd, a Windows virtual audio mixer, works in real time and does not compress your monitoring path. The audio moving through the engine, through your effects and out to a bus, is full uncompressed sample data the whole way. Lossy versus lossless is not something patchd applies to what you hear; it is a decision that happens later, when something else saves or streams the result.
That decision belongs to whatever app is capturing the output. Your recorder writes a file in the format and bitrate you pick there, so choose WAV or FLAC when you want the master. Your streaming or call app encodes to a lossy stream on its way out over the network. A bus in patchd is a virtual output that other apps pick up as a microphone, so what those apps receive is the clean, uncompressed signal; the format is theirs to set. If you are wiring this up for a recording, see how to record desktop audio on Windows.
The takeaway
Lossless audio like WAV and FLAC keeps every sample and is a perfect copy, at the cost of a big file. Lossy audio like MP3 and AAC deletes the least-audible data to shrink the file, at the cost of a small, permanent quality loss set by the bitrate. Keep a lossless original for anything you record, edit, or archive, and reach for a high-bitrate lossy file when you need to stream or share. The one rule that never changes: you can always make a lossy file from a lossless one, but never the reverse.
patchd is pre-launch, with a free effects rack on every channel and a real-time engine that keeps your monitoring path uncompressed so the format choice stays yours at save time. If you want a Windows mixer that keeps your audio clean on the way through, join the waitlist and get notified when it ships. The paid Studio tier is $39.99 per year, and the engine is identical on both tiers.