Drop in almost any audio file and get a 128 kbps MP3 back — the small-file rate, and the right one for voice. No signup, no watermark, no limit on how many.
128 kbps gives the encoder 16 kilobytes per second, and that number is the whole character of the rate. For a single voice in a quiet room it is plenty: speech occupies a narrow band and compresses well, which is why podcast feeds and audiobooks have sat here for two decades. For music it is the rate at which an MP3 starts showing its work.
Use it for anything that is mostly talking — podcasts, interviews, lectures, sermons, audiobooks, voice memos, call recordings. Use it when the file has to move over a slow or metered connection, or land on a device with no room to spare. And use it when the number of files matters more than any one of them: at this rate an eight-hour audiobook is a download, not a project.
128 kbps is 16 KB every second: 960 KB a minute, 57.6 MB an hour. A four-minute song is about 3.8 MB. An eight-hour audiobook lands near 460 MB — against roughly 5 GB for the same recording as 16-bit stereo WAV, which is why nobody ships audiobooks as WAV.
The thing to know is what this rate does to music, and where. 128 kbps MP3 fails first on high-frequency noise: cymbals and hi-hats pick up a metallic shimmer, applause and rain turn into a swirling wash, and long reverb tails go grainy as they fade. A voice recording has almost none of that, which is exactly why the same rate is transparent for a podcast and obvious on a live album. If the material is music you intend to keep, encode the keeper at a higher rate and make the 128 kbps file a copy of it, not the other way round.
For speech, almost always — a podcast, an interview or an audiobook at 128 kbps is hard to separate from its source on any normal listening setup. For music it is the rate where the compression becomes audible on the difficult parts, so it is a good travel copy and a poor master.
Yes, to some degree — MP3 is lossy at every rate and 128 kbps discards the most. How much you notice depends entirely on the material: a spoken word file, very little; a cymbal-heavy live recording, quite a lot.
No. What the encoder removed is gone, and a later conversion to a higher rate just produces a bigger file holding the same audio. Keep the original if you might want a better copy one day.
At 960 KB a minute, an hour of audio is 57.6 MB and an eight-hour audiobook about 460 MB. The same eight hours as 44.1 kHz 16-bit stereo WAV would be roughly 5 GB, and at 320 kbps about 1.15 GB.
Any of MP3, WAV, FLAC, OGG, AAC, M4A, WMA, Opus and AIFF, up to 200 MB per file. The output is always an MP3 at a constant 128 kbps.