Just a few questions…
I’ve read that when you import an MP3 file into Audacity, there will be quality loss in reencoding. However, does that affect the original file, or the soundtrack in the project?
Also, when an AAC or M4A file is imported, does any quality loss occur?
The quality loss is on the export side (encoding), not on the import one (decoding).
The imported track exhibits the same quality as the playback in any audio player does.
Thus, if you hear problems while playing the file, those will be passed to Audacity as well.
The goal is to keep the quality level when exporting from Audacity, either by a lossless codec such as wav or flac or by high quality ogg or mp3 settings.
With your first statement, I’m still not certain that the import part is true, because the Audacity Wiki has a quote that says “If you import an MP3 into Audacity, edit it then export it as an MP3, you are thus losing quality twice - once in the original MP3 encoding of the imported audio, then again when you export it from Audacity as MP3.”
Which looks to me like importing a file with a lossy codec always loses quality.
Also, does the second statement mean that even after an imported track is decoded, the original file is unaffected?
I was typing the same time as Robert so I’ll just post the extra detail in case you want it.
When you import an MP3 Audacity by default makes a full resolution (32-bit float) copy of it expanded to lossless PCM. This does not make the audio in the track better than it was. It makes a copy that is of the same quality as before rather than being worse.
The original MP3 is not affected unless you choose to overwrite it when you export. Whether you export over the original MP3 or export a new one, the quality of that exported file will be worse than the original. The lower the bit rate you choose for export (see the “Options…” button when you export) the greater the re-encoding loss.
You can make simple cut/paste/join and volume edits to MP3’s without re-encoding, using other tools. Or as Robert said you can export as WAV or FLAC which are lossless but this makes the files much larger than MP3. WAV can be 10 times larger, FLAC 5 times larger.
It works the same way. The quality loss occurs on export. Not all M4A files are lossy, though. Some may contain Apple’s lossless ALAC codec which is of similar size to FLAC.
Now, if you change Audacity’s Quality Preferences to 16-bit (not generally recommended) then the copy Audacity makes of the imported file is not of such high resolution. It is still of the same quality when you import it, but due to rounding errors as a result of the lower, whole number resolution, effect edits can technically degrade the sound. By default this is mitigated by “dithering” the errors. They become extremely quiet “whoosh” noise that you won’t hear in most cases rather than what sounds like harmonic distortion.
Gale
No. On import, even at 16-bit resolution, the audio has the same quality as before. The first loss the Wiki refers to is the loss caused by whoever created the MP3 in the first place.
The original file is unaffected unless you export over it. Audacity is making a copy of the file, not editing it directly.
Gale
Okay then let me get it straight.
Importing a file with a lossy codec does not affect the quality of the soundtrack or the original file.
Exporting to file with a lossy codec will affect the quality.
The AAC CODEC was designed to minimize accumulated quality loss from re-encoding. So, it shouldn’t be as bad as re-encoding MP3. But, I’m not sure if going from MP3-to-AAC is any better than going from MP3-to-MP3… It might be worse…
Sometimes you are stuck with what you’ve got and you have no choice. The quality loss isn’t always noticeable, but it’s something you should be aware of and something you should avoid whenever possible (or whenever practical).
All “normal” audio editors have to decode compressed files to edit the raw PCM data.
There are some special MP3 editors such as MP3directCut that can do some limited MP3 editing without decoding. MP3directCut can cut, past, splice, paste, change the volume, fade-in or fade-out. But, you simply can’t do more advanced audio processing like EQ, mixing, or reverb without decompressing to get the actual audio data.
I’ve got a video editor that only re-encodes the video where you change it. For example, if you combine two video clips with a cross fade (or other transition), only the cross-faded part is re-encoded. But, I don’t know of any audio editor that works like that.
DVDDoug
Thanks for the extra detail. Even though it’s off topic, out of curiosity, what is this video editor that you mention?
I think what the Wiki means is quality loss in the file when it was originally saved as an mp3, not loss from importing, “once in the original MP3 encoding of the imported audio.” So you’ve already got a lossy file to start with and if you export it again as mp3 you get further loss. 
That’s right. Importing a file does not change the sound quality. Data compression when exporting in a lossy format (such as MP3, Ogg, WMA, …) does reduce the sound quality (always).
Okay then. Thank you all for your assistance.
I want to convert a certain mp4 video to mp3 or aac, whose file profile is in the attached image below.
*I opened the file on another software (fre:ac), which showed its Sample Resolution to be 16 bits.
I changed Audacity’s Default Sample Rate to 48kHz and Default Sample Format to 16 bits and Dither under High Quality Conversion to None. But when I imported the video(ffmpeg installed in advance), the track control panel showed its profile to be in the 32-bit float format. I want to extract the audio to its original form, if not the highest quality(which I believe cannot be achieved as the audio is already so lossy). How can I achieve that?
I have attached images of:
Thanks in advance!
Audacity uses 32-bit float format internally, and is beneficial when performing any additional processing.
Conversion from 16-bit to 32-bit float is entirely lossless.
Processing in 32-bit float is higher quality that processing in 16-bit.
If you are only importing and then exporting again (without any processing) then it is OK to turn off dither (If you do any processing, then better to leave dither set to “shaped”).
My recommendation, leave it at 32-bit float.
How do I save it in .aac? The only option shown is m4a right?
AAC and M4A are both the same format, just different file extensions. You can simply rename an M4A file as AAC or vice versa.
For best quality, export as WAV or FLAC.