Free FLAC to MP3 converter — and what actually decides the quality
Free is the easy part. The part that decides whether your MP3 sounds good is which encoder the tool uses and how it is configured — two "free converters" can produce audibly different files from the same FLAC.
In 30 seconds
This converter is free, has no file-size cap, no daily limit, and no account — and it runs entirely in your browser, so your FLAC files are never uploaded anywhere. Pick 320 kbps for the best result, 192 kbps as the everyday default, 128 kbps only when storage is genuinely tight. Below is the part most free converters will not tell you: at the same bitrate, encoder choice and settings changed the measured result enough to matter — in one 2001 ABX comparison, average-bitrate encoding beat constant-bitrate encoding 15 times out of 16.
What "free" usually costs you
Most free FLAC to MP3 converters fall into two groups. Upload-based tools send your file to someone else's server, then either delete it on a timer or keep it — you cannot verify which. Installed tools are free in money but cost you a download, and sometimes bundled software you did not ask for.
A third problem is quieter: two tools that both say "128 kbps" can produce different sounding files, because they use different encoders and different settings. The rest of this page is about that difference, using measurement data nobody publishes anymore.
The settings that decided quality at 128 kbps
The recommended constant-bitrate command for Lame 3.92 (released 14 April 2002) was --alt-preset cbr 128, which expanded to:
-h --nspsytune --athtype 2 --lowpass 17.5 --ns-bass -6 --scale 0.93
An alternative form used --lowpass 16 --ns-bass -8 with the same --scale 0.93. Every one of those flags was there for a reason, and the version history shows them being tuned in public:
- 17 July 2001 —
--ns-bass -8was added. - 4 November 2001 —
--scalewas lowered from 0.98 to 0.93. - 21 December 2001 — Lame 3.90 stable packaged the main recommendation as the presets
--alt-preset cbr 128and--alt-preset 128. - Lame 3.93 — explicitly not recommended with any of these settings, despite being newer.
That last line is the useful lesson: newer is not automatically better, and a free converter that silently ships a worse encoder build will still tell you it is converting to MP3.
Encoders that were flagged as unusable
The same source kept a public do-not-use list. The reasons are specific, and they are the kind of thing you would never discover from a converter's marketing page:
- Xing at 128 kbps — clearly below the better MP3 encoders at that bitrate. The gap narrowed as bitrate rose.
- Blade at 128 kbps — worse than Xing. Eleven technical causes were listed: no lowpass filtering, incorrect transient detection, short-block switching that triggers too late, poor bit-reservoir handling, no joint stereo, non-optimal Huffman compression, incorrect absolute-threshold-of-hearing values, poor quantisation selection, no scalefactor use, no scfsi — and "a lot of bugs everywhere".
- QDesign MVP 2.1 — at or near the bottom on every file tried, and worse at high bitrates on one sample.
- A Fraunhofer build with a known bug — made files sound dry, or lacking in reverberation. Two later releases were documented as having fixed it (25 February 2001 and 9 May 2001), which means the same product name shipped both broken and fixed builds.
- An older Fraunhofer codec that sounded fine to the person running the test but produced what others described as horrible high-frequency artifacts — a reminder that "sounds fine to me" is not a quality check.
Where this data stops being valid — quoted as written (ff123, 2001)
"At the higher bitrates, my listening tests will mean less and less. At 192kbs and above, for example, I won't be able to properly tell you what codec sounds best, since most codecs are already removing the vast majority of midrange artifacts, making it mostly a contest of which codec has the best high-frequency response. For these listening tests, you'll have to rely upon somebody else."
He tightened it further on 4 February 2001: "even at 128 kbit/s, I may lack the high-frequency ability necessary to properly evaluate the quality of an mp3 codec."
How to check it yourself in ten minutes
You do not have to take any of this on faith. The organisers of those tests used a rule of thumb you can copy at home:
- Make two MP3 copies of the same FLAC at the bitrates you are choosing between.
- Play them back-to-back against the original, with the labels hidden and any equaliser or sound enhancement switched off — on many players the equaliser button only opens the window, it does not turn the effect off.
- Count how many times you can correctly identify which file is which. 12 correct out of 16 counts as genuinely hearing a difference at 95% confidence. Below that, treat them as the same and pick the smaller file.
This is the only test that matters, because it uses your ears on your equipment. It is also free here in a way it is not on an upload-based converter: there is no per-file wait, no queue, and no cap, so you can run the same track at three bitrates as often as you like.
What you get here
- Truly free — no file-size cap, no daily limit, no account, no email wall.
- Nothing is uploaded. Conversion happens in your browser on your device, so there is no server copy to delete and no privacy policy to trust.
- Your FLAC stays yours. Keep it as the master and generate as many MP3 copies as you need, at any bitrate, whenever a new device needs one.