Author: Fei Y (Page 2 of 19)

Every audio converter promises the same thing: drop a file in, pick a format, get it back. What none of them tell you is what they quietly changed on the way through — the bitrate they chose for you, the sample rate they resampled to, whether a “lossless” conversion really was lossless. So we ran the same audio file through the most-used free online audio converters and measured the files that came back.

Disclosure: this blog is published by RedPanda Compress, which is one of the tools tested. It went through the same procedure as the others, and the numbers are reported the same way.

The short answer

• Best defaults for MP3: CloudConvert. It is the only server-side tool here whose out-of-the-box MP3 is high quality — 232 kbps VBR, and it measured very slightly cleaner in the treble than our reference. Free use is capped at 10 conversions a day.
• Fastest, and nothing leaves your machine: RedPanda Compress runs the conversion in the browser, so there is no upload at all. It wrote a 192 kbps MP3 in 15.8 seconds and did a lossless FLAC → WAV conversion in 0.7 seconds.
• Watch the default bitrate: FreeConvert, HappyScribe and Online Audio Converter all default to 128 kbps MP3 — the lowest of anything we tested, and measurably thinner at the top end. All three let you raise it; most people never open the panel.
• Watch the sample rate too: Online Audio Converter’s default converts a 48 kHz file to 44.1 kHz — even when you ask for a lossless WAV. It was the only tool that did not hand back the original audio unchanged.

How we tested

• The file: three minutes of the Big Buck Bunny soundtrack (CC-BY, Blender Foundation) — the same source as our video compressor comparison — decoded to a 48 kHz, 16-bit stereo WAV of 32.96 MB, plus a FLAC copy of the identical audio at 10.62 MB. Anyone can download the source and repeat this.
• Two jobs, not one. First a lossy job — WAV to MP3, which is what most people actually do. Then a lossless job — FLAC to WAV, where there is a single correct answer: the audio that comes back should be identical, sample for sample, to the audio that went in.
• Defaults only. We changed nothing. Every one of these tools has an options panel; the point of the test is what happens when you don’t open it.
• What we measured: the format, bitrate, sample rate and size of the returned file; the energy above 15 kHz compared with the original (this is where low-bitrate MP3 encoders cut first); for the lossless job, the MD5 of the decoded audio and a null test — playing the returned file against the original with the polarity flipped, so that anything left over is the difference between them.
• Which tools: the free web converters ranking for “audio converter” in US organic search (roughly 12K searches a month) — Online Audio Converter, FreeConvert, CloudConvert and HappyScribe — plus RedPanda Compress. Sites that only offer a desktop installer were left out.
• The machine: an 8-core Apple Silicon Mac in Chrome. For everything except RedPanda, most of the elapsed time is the upload, so it depends on your connection, not on the tool.

Test 1: WAV to MP3, at each tool’s default

Tool	Default MP3	Sample rate	Treble vs original	Time
CloudConvert	232 kbps VBR 5.23 MB	48 kHz unchanged	+0.2 dB no loss	~20 s
RedPanda Compress	192 kbps 4.32 MB	48 kHz unchanged	−0.02 dB no loss	15.8 s no upload
FreeConvert	128 kbps 2.88 MB	48 kHz unchanged	−1.1 dB	minutes upload-bound; see below
HappyScribe	128 kbps 2.88 MB	48 kHz unchanged	−1.1 dB	9–12 min then an ad
Online Audio Converter	128 kbps 2.88 MB	44.1 kHz resampled	−1.5 dB	25–65 s mostly upload

Two tools returned an MP3 with no measurable loss in the top end, and three returned one that was measurably thinner. That difference is entirely the default bitrate — and it is the one setting almost nobody changes.

Test 2: FLAC to WAV, where “lossless” has to mean lossless

Converting FLAC to WAV should not be a judgement call. FLAC is lossless compression; WAV is the raw audio. Decode one into the other and the samples should come out exactly as they went in. So we compared the MD5 of the returned audio against the MD5 of the original, which passes only on a perfect, sample-for-sample match.

Tool	Sample rate	Identical to the original?	Time
RedPanda Compress	48 kHz	Yes — null test is digital silence	0.7 s no upload
CloudConvert	48 kHz	Yes — null test is digital silence	under 8 s
FreeConvert	48 kHz	Yes — null test is digital silence	~1 min mostly upload
HappyScribe	48 kHz	Yes — null test is digital silence	~7 min then an ad
Online Audio Converter	44.1 kHz	No — resampled	10–25 s

Four out of five passed. The one that didn’t is worth explaining carefully, because it is not a quality problem in the way you might expect: Online Audio Converter’s WAV setting has a quality slider that defaults to “CD Quality 44.1 kHz”, and our file was 48 kHz. Its resampler is a good one — the leftover difference measured about −100 dBFS, which nobody can hear. But the file you get back is not the file you put in, its sample rate has silently changed, and if you are feeding it into a video edit or a DAW session running at 48 kHz, that is a real problem. The slider does have a 48 kHz position. It just isn’t the default.

Tool by tool

CloudConvert: the best defaults of any server-side converter

CloudConvert was the only uploading tool that got both jobs right without being asked. Its MP3 default is a variable bitrate averaging 232 kbps, which measured a touch cleaner in the treble than the reference, and its FLAC to WAV came back bit-identical in under eight seconds. The options panel is also the most complete of the group: bitrate, VBR quality, sample rate, channels, volume, trimming, and a “copy” mode that rewraps the existing audio without re-encoding it. The catch is the free tier — 10 conversions per day.

RedPanda Compress: no upload, and the fastest by a wide margin

Because the conversion runs inside the browser, the file never goes anywhere: no upload, no queue, no server-side copy to delete afterwards. That is also why the FLAC to WAV job finished in 0.7 seconds while the uploading tools were still sending the file. The MP3 default is 192 kbps, which measured within 0.02 dB of the original in the treble, and the WAV output was bit-identical. There is no account, no daily cap and no watermark. The trade-off is that the work happens on your own hardware, so an older laptop will be slower, and there are only three output formats — MP3, M4A and WAV — rather than the fifty-odd a server-side converter can offer.

Online Audio Converter: the most popular one, and the only one that changed the audio

This is the highest-traffic free audio converter on the web, and the interface is genuinely good — formats across the top, a quality slider underneath, Google Drive and Dropbox as sources, and a batch mode. But both of its defaults move away from the original: 128 kbps for MP3, and 44.1 kHz for WAV. If you use it, move the slider before you convert; for WAV that means pushing it to the “DVD 48 kHz” position when your source is 48 kHz.

FreeConvert: correct on lossless, low on MP3

FreeConvert handled the lossless job perfectly — 48 kHz, bit-identical — but its MP3 default is 128 kbps, and our measurement put it 1.1 dB down in the treble. The advanced options do let you set the bitrate, sample rate and channels before converting. Uploads dominate the clock: the 10.6 MB FLAC job took about a minute end to end on our connection, and the 33 MB WAV job never got past the upload stage on two separate attempts, so we are not quoting a figure for it. On a faster upstream connection than ours this is likely a non-issue; on a slow one it is the whole experience.

HappyScribe: accurate, extremely slow, and there’s an ad in the way

HappyScribe is a transcription company with a converter attached, and the converter is honest about the audio — its FLAC to WAV came back bit-identical. It is just very slow: nine to twelve minutes for the three-minute WAV to MP3 job, and about seven minutes for the FLAC, on a file the other tools finished in seconds. When it does finish, the download button is replaced by “Watch a short ad to download”. There are no bitrate options at all, and the MP3 it writes is 128 kbps.

What to check before you convert

• Find the bitrate before you press convert. Three of the five tools here default to 128 kbps. If the panel says 128 and your source is better than that, raise it to 192 or higher — the file is bigger and nothing else about it is worse.
• If the job is lossless, check the sample rate. FLAC to WAV, WAV to FLAC, ALAC to WAV — none of these should change the sample rate. If the tool offers you a “quality” choice for a lossless format, that is a resampler, and the safe answer is whatever matches your source.
• Don’t re-encode what is already fine. Converting an MP3 to another MP3 loses a little every time. If you only need a different container, look for a “copy” or “remux” option.
• Think about where the file goes. Every tool here except RedPanda uploads your audio to a server and deletes it later on its own schedule. For a music file that is a non-issue; for an interview, a client’s recording or anything under NDA, it is the whole question.

Frequently asked questions

What is the best free audio file converter?

For MP3 at the best out-of-the-box quality, CloudConvert — its default is 232 kbps VBR and it measured no loss against our reference. For speed and privacy, RedPanda Compress, which converts in the browser with no upload and did a lossless FLAC to WAV in under a second. Both returned the original audio unchanged on the lossless test.

Does converting audio always lose quality?

No. Converting between lossless formats — FLAC, WAV, ALAC, AIFF — is exact, and four of the five tools we tested proved it by returning audio identical to the original. Quality is only lost when the destination is a lossy format such as MP3, AAC or Opus, or when a tool resamples the audio on the way through.

What bitrate should I pick for MP3?

192 kbps is a sensible floor for music and is where our own converter sits; 256 or 320 kbps if you are archiving and the file size doesn’t matter. 128 kbps is fine for speech — a podcast, a lecture, a voice memo — and that is really what it exists for. The gap this test measured between 128 and 192 kbps was about 1.1 dB of energy above 15 kHz.

Is it safe to upload audio to an online converter?

It depends entirely on the file. The reputable services here state a deletion window — a few hours to a day — and honour it. But the file does sit on somebody else’s server in the meantime. If the recording is confidential, use a converter that runs in your browser, where nothing is uploaded in the first place.

Can I convert FLAC, OGG, WMA, M4A and AAC too?

Yes — every tool in this comparison reads all of those. The server-side converters offer the longest format lists (CloudConvert and FreeConvert both run to hundreds of conversions, including video containers). A browser-based converter covers the common audio formats in and MP3, M4A or WAV out, which is most of what people actually need.

Conclusion

The converters that get talked about as interchangeable are not. On the lossy job the spread between them was entirely down to a default bitrate that three of the five set to 128 kbps. On the lossless job, where there is only one right answer, four got it right and the most popular tool on the list silently resampled the file. If you want the best MP3 without touching a setting, use CloudConvert. If you want it fast, unlimited, and without your audio leaving your computer, use RedPanda Compress. And whichever you pick, look at the bitrate before you press the button.

Pulling the audio out of a video looks like the simplest job on the internet, and most tools make it look identical: drop in an MP4, get an MP3. But the file you get back is not the same from tool to tool. We ran the same video through the most popular video to audio converters and measured what actually came out — bitrate, sample rate, how much of the original sound survived, and how long each one took.

Disclosure: this blog is published by RedPanda Compress, which is one of the tools tested. It went through the same procedure as the others, and the numbers are reported the same way.

The short answer

• If you want an MP3 and the best quality: CloudConvert. Its default is a 234 kbps VBR MP3, which kept everything that was in the original track, and it finished in 96 seconds. Free use is capped at 10 conversions a day.
• If you don’t want to upload the video: RedPanda Compress converts in your browser, so there’s no upload at all. It writes a 192 kbps MP3 in about 80 seconds with no measurable loss.
• If you want zero quality loss: don’t convert to MP3 at all. Keep the audio exactly as it already is inside the video (an M4A file). In our test that took 2.2 seconds and produced a bit-for-bit copy of the original track. More on that below.
• Watch the default bitrate: FreeConvert and HappyScribe both default to 128 kbps, the lowest setting of anything we tested, and both measurably thinned out the top end.

How we tested

• The file: the same video we used for our video compressor comparison — Big Buck Bunny (CC-BY, Blender Foundation), 9 min 56 s, 158.6 MB, with a 130 kbps AAC stereo soundtrack at 48 kHz. Anyone can download it and repeat these tests.
• What we measured: time (upload and processing separately where the tool shows both), the size and format of the file that came back, its bitrate, sample rate and channel count, whether the full 9:56 survived, and the free-tier limits we hit.
• Quality: MP3 encoders throw away the highest frequencies first, so we measured how much energy above 15 kHz survived in each output compared with the original track. A result within a few tenths of a decibel is effectively a clean copy; more than a decibel means audible detail has gone.
• Which tools: the pages that rank on the first page of Google for “video to mp3” and “mp4 to mp3”, ordered by estimated US search traffic (SEMrush, September 2026). Downloader sites aimed at YouTube links rather than your own files were left out.
• Setup: an 8-core Apple Silicon Mac, Chrome, about 40 Mbps down and 15 Mbps up. For tools that upload, the upload depends on your own connection.

Results

Tool	Time	Output	Treble kept*	Free limits
RedPanda Compress M4A, no re-encoding	2.2 s	9.80 MB AAC 130 kbps	identical	none; in browser
RedPanda Compress MP3	~80 s no upload	14.3 MB MP3 192 kbps	−0.08 dB	none; in browser
CloudConvert	96 s	17.5 MB MP3 234 kbps VBR	+0.19 dB	10 conversions/day, 1 GB, 5 min processing
ezgif	102 s 82 s upload + 20 s	14.3 MB MP3 192 kbps	−0.04 dB	200 MB per file
Zamzar	under 1 min 35 MB clip only	MP3 192 kbps	−0.11 dB	50 MB per file
FreeConvert	~4 min mostly upload	9.5 MB MP3 128 kbps	−1.42 dB	1 GB, daily processing allowance
HappyScribe	~17 min 43 s upload + ~16 min	9.5 MB MP3 128 kbps	−1.42 dB	watch an ad to download

The fastest, cleanest option isn’t an MP3 at all

The audio inside an MP4 is already a compressed, finished file — usually AAC. Converting it to MP3 means decoding it and compressing it a second time, which always loses a little and takes real processing time. But you can also just lift that audio out and put it in its own container, without touching the audio data. The result is an M4A file.

We tried this with RedPanda Compress’s audio converter set to M4A. It finished in 2.2 seconds and produced a 9.8 MB file. We then compared the audio stream inside it with the audio stream inside the original video: the checksums were identical. Not “almost the same” — the same bytes, with nothing re-encoded. CloudConvert exposes the same idea in its options, where the audio codec can be set to “copy” instead of MP3.

M4A plays on iPhones, Macs, Android, Windows and every browser. The one place it can still be awkward is older car stereos and cheap MP3 players, which is the main reason to pick MP3 anyway. If your destination takes M4A, this is the option to use: it’s instant and it’s lossless.

Tool by tool

CloudConvert: the best MP3 defaults

CloudConvert was the only tool that defaulted to a genuinely high-quality MP3: variable bitrate averaging 234 kbps, which measured slightly cleaner in the treble than our reference. It took 96 seconds from drop to download, and its options panel is the most complete of the group — bitrate, VBR quality, sample rate, volume, trimming, and the “copy” codec mentioned above. The free tier allows 10 conversions per day with a 1 GB file limit and a 5-minute processing cap.

RedPanda Compress: nothing is uploaded

Because the conversion runs inside the browser, the 158 MB video never went anywhere: the MP3 appeared after about 80 seconds of local processing (69 to 104 seconds across three runs) at 192 kbps, with no measurable treble loss. There’s no account, no daily limit and no watermark, and the M4A route above finished in seconds. The trade-off is that the work happens on your machine, so an older laptop will take longer, and a very long file depends on your available memory.

ezgif: plain, fast, and it tells you what the bitrates mean

ezgif’s page looks like it was built in 2009 and works better than most of the modern ones. Upload took 82 seconds, conversion 20, and the default 192 kbps output matched the original’s treble almost exactly. The page even includes a small table explaining what 128, 192, 256 and 320 kbps are each good for. The limit is 200 MB per file, and there are ads down both sides.

Zamzar: fine, if your file is under 50 MB

Zamzar’s free tier stops at 50 MB, so our 158 MB test file was never an option. We gave it a 35 MB clip instead: it converted in well under a minute to a 192 kbps MP3 with no measurable loss, and the download needed no account. For a short clip it’s perfectly good; for a full-length video the size cap rules it out unless you pay.

FreeConvert: 128 kbps by default

FreeConvert handled the file without complaint, but its default output was a 128 kbps MP3 — the same nominal bitrate as the original AAC track, which is exactly the situation where a second round of compression does visible damage. Our measurement showed 1.4 dB less energy above 15 kHz than the original, and the spectrogram shows the cut-off clearly. You can raise the bitrate in the advanced options, but most people won’t. The whole job took about four minutes on our connection, nearly all of it upload.

HappyScribe: slow, and an ad stands between you and the file

HappyScribe uploaded the file quickly — 43 seconds, the fastest upload in the group — and then spent roughly sixteen minutes converting it, by far the longest of anything we tested. The result was another 128 kbps MP3 with the same 1.4 dB of lost treble. Before you can download it, the page asks you to watch a short ad. There are no bitrate options at all.

What to check before you convert

• Find the bitrate setting. It’s the single thing that decides quality. If a tool won’t tell you what bitrate it writes, assume 128 kbps. For speech — lectures, interviews, podcasts — 128 is genuinely fine. For music, ask for 192 or higher.
• Don’t convert twice. The audio in your video has already been compressed once. Every MP3 conversion is a second pass. Going straight to M4A skips that entirely.
• Check the size limit before you start. Zamzar stops at 50 MB free, ezgif at 200 MB, CloudConvert and FreeConvert at 1 GB. In-browser tools are limited by your device instead.
• Remember the upload. On a normal home connection, sending a 158 MB video takes longer than converting it. If you only need the audio, the tools that never upload finish while the others are still transferring.

Frequently asked questions

What is the best free video to audio converter?

For an MP3 with good default settings, CloudConvert produced the best result in our test. If you’d rather not upload the video, RedPanda Compress reached almost the same quality with nothing leaving your device. If your player accepts M4A, extracting without re-encoding beats both, and takes seconds.

Does converting video to MP3 lose quality?

Yes, a little, because the audio is compressed a second time. How much depends on the bitrate: at 192 kbps and above we could not measure a meaningful difference from the original; at 128 kbps we could. Extracting to M4A loses nothing at all.

What bitrate should I choose?

192 kbps is the sensible default for music and general use — it was enough to keep everything measurable in our test. 128 kbps is fine for speech and halves the file size. Going above 256 kbps mostly adds file size, since it cannot restore detail the original video didn’t have.

Can I extract audio from MOV, MKV or WebM too?

Yes. Every tool here accepts the common video containers. We have dedicated pages for MOV, MKV and WebM, and a step-by-step guide to extracting audio from video.

Conclusion

Every tool here will give you an audio file, and for a lecture recording any of them is fine. The differences show up when the soundtrack is music: a 128 kbps default quietly costs you the top of the range, while 192 kbps or better keeps what the video had. And if you don’t specifically need an MP3, the fastest and cleanest answer is to take the audio out as it is — no second compression, no waiting.

Related reading: Best Free Online Video Compressor 2026 · MP3, AAC, WAV, FLAC and Opus compared · How an in-browser audio converter actually works