bit depth changes when reopening mp3 file

[mikegee63]

OK, here is the scenario:

I record audiobooks for audible in a small home studio, so, files have to be saved and uploaded to them with very specific parameters, basically they are:

==========================

.MP3

CBR

16 bit-depth

192khz bit-rate

mono

==========================

I record in .wav 44.1 16bit....

==========================

So, I record in WAV and save to MP3 in the parameters stated above, the PROBLEM is when I re-open the saved mp3 file the bit depth always changes to 32-bit !!

Does anyone know why this is happening or even better, how to fix it?

[ryclark]

That is because 32bit .wav is Audition's native audio file type and any file that is opened into Audition is always converted to that inside the programme. So there is no way to fix it in Audition. However if you have saved the .mp3 files with the correct settings then they will be correct for the end user. You can check that your .mp3s conform to the Audiobook's company's specifications using MediaInfo

[mikegee63]

You have got to be kidding, I almost find that hard to believe..... so, just to be clear:

You are telling me that when I save an MP3 with a 16bit bit depth and then re-open it later, Adobe will then go ahead and re-open it with a 32bit bit depth, because 32bit wav is its native file type?

and this holds true even though we are talking about MP3's bit depth being changed, not wav files being changed?

I am flabbergasted !

You seem very confident in your answer, although I would definitely like a couple of other opinions on the matter.

Thanks!

[SteveG_AudioMasters_ · Accepted solution]

mikegee63  wrote

You have got to be kidding, I almost find that hard to believe..... so, just to be clear:

You are telling me that when I save an MP3 with a 16bit bit depth and then re-open it later, Adobe will then go ahead and re-open it with a 32bit bit depth, because 32bit wav is its native file type?

and this holds true even though we are talking about MP3's bit depth being changed, not wav files being changed?

I am flabbergasted !

You seem very confident in your answer, although I would definitely like a couple of other opinions on the matter.

What ryclark said wasn't an opinion, it was a valid explanation - that's what we do here.  But let's alter the terminology slightly, as you may be getting slightly confused as to what's going on:

Audition's not opening your MP3 as such - it's decoding it.

Audition's working format is wav. Whenever any file recorded in a different format is opened in it, it gets decoded and that decode is displayed as a wav file. So this doesn't affect the original file in the slightest - it's just what Audition does to re-display it.

So no, your MP3 file hasn't had its bit depth changed at all - this is just what happens in Audition when it's decoded and displayed.

You have to bear in mind here that MP3 is only supposed to be a distribution format, not a working one. What audible want is a file that essentially they have to do nothing to, except put into a book format. It would be better if they took wav files, processed them themselves and saved all these problems. But that would increase their overheads, so it's unlikely that they will ever do this. Also they'd take quite a bit longer to upload, which a lot of providers wouldn't like.

Oh, and none of this is opinion either - it's all well-established facts.

[mikegee63]

I'm sorry if I offended anyone, I just wanted another person to tell me the same thing before I gave up trying to find a solution, like wanting a second opinion from a doctor, you just want to make sure a person knows what they are talking about before you take or don't take action, and one of the best ways to accomplish that is to get a second opinion.... OR a second confirmation.

[mj89683611]

Just to clarify something here—

The MP3 format has no bit depth; there's no such thing as a 16-bit MP3. The MP3 file just stores a set of coefficients derived from selected frequency components of short overlapping segments of the input waveform. When the MP3 is decoded, the coefficients are used to mathematically reverse the process and construct a new stream of PCM samples representing a waveform which is perceptually similar, if not identical, to the original input. These output sample values, i.e. their amplitude, will have whatever precision the decoder supports. In other words, the decoder determines the bit depth.

When saving a file in Audition, when you choose the sample type (e.g. 16-bit, 44100 Hz, no dither), Audition will convert the 32-bit working copy to that first, and then it will do whatever it needs to do to write the target format, in this case feeding the converted samples to the Fraunhofer MP3 encoder. Although this encoder accepts 32-bit input, even at 320 kbps it is very selective about what it attempts to preserve below -96 dB, so any input above 16-bit is pointless. It's also quite poor with input under -72 dB. Much better quality is achieved if you save to 32-bit or 24-bit WAV, and then convert to MP3 using the LAME encoder, outside of Audition.