Rendering to AAC questions

Hi, I got some questions about rendering to AAC format, and one question about the “bitrate” properties of a wavfile.

  1. When rendering an AAC file, shall i choose “Itunes standard (AAC)”. If so, why not the other "Itunes + (AAC) - See PIC 1

  2. In the “Audio file format”-dialog, shall I choose m4a or mp4, and whats the difference? - See PIC 2

  3. I choosed “Itunes standard (AAC)”, m4a, and rendered. When the file got imported to wavelab, I read in the info at the bottom right corner that the file is “stereo 32 bit float, 44 100 hz” I dont know much about this topic. I didnt made any choices about that when rendering. But the file I rendered from was 32 bit floating. Shall the AAC file be 32 bit float and 44 100 hz? - See PIC 3

  4. When right clicking at the file itself (not in wavelab), and choosing properties, then going to the details tab, it says that the bit rate of the file is “127 kbps”. What does that mean and shall it be 127? :S - See PIC 4

  5. Here is a question about a 16 bit wave file of the same song that Í rendered. When going to that files properties - details, it says that the bit rate of the file is “1411 kbs”. I havent made any choices about it in wavelab, but I wonder, is it the correct value? - See PIC 5

And here is the two last pics.

  1. If you look closely at the details in the preset settings, iTunes standard AAC is 128kbps (the old iTunes standard really). iTunes+ is 256kbps and Apple has been using this for iTunes Store purchases (and now Apple Music/iTunes Radio) for many years now. I would call it the current standard.

  2. Google it for more, but mp4 can contain audio and video, m4a is audio only.

  3. WaveLab has to decode AAC and mp3 files back to WAV to load into the audio editor, so this is why you see 32-bit floating displayed. WaveLab is not the best place to analyze all aspects of lossy formats because it has to decode them to WAV. Try Jaikoz, Sonnox ProCodec/Codec Toolbox, JRiver Media Center, or many others to analyze the actual AAC files you have made.

  4. This can be related to the actual bitrate with a variable bitrate encoded file. PG (or somebody else) can probably explain it better. Some things changed here I think with a recent WL update.

  5. This is correct. 1411 kbps is the value you will see for 16-bit WAV.

The calculation is this:
Number of Channel * word length * sample rate
Playing CD quality is playing 2 channels with 16 bit word length and 44100 samples per second or 21644100=1411200 bit/s or 1411 kbps.

@ 4): I see this too from time to time, and in different places in Windows: sometimes mp3s show 1 bit less from what they really are. 320 CBR shows as 319 kbps etc… No clue how it occurs, but never noticed an actual difference in the audio either. I’m not sure if it’s related to WL (calls for an experiment) but they’re certainly not VBR mp3s.

@5): Unfortunately Windows since version 7 cannot show sample rate and bit depth anymore, but shows bitrate for everything. For my old brain I created a desktop picture with the ‘translations’ in it. 1411 kbps is stereo 44k1/16b, 2304 kbps is mono 96k/24b etc, etc. It’s a pain, but after a while you start recognizing the numbers…

Justin, maybe I’m wrong, but I thought all decoders decoded lossy to PCM, in order to analyze, or even play lossy files. If that’s the case I don’t see the difference in Wavelab presenting that PCM as a WAV file, vs the other programs not mentioning the PCM. But like I said, maybe I’m wrong.

I think Wavelab could be the best place to analyze lossy files since it now decodes to 32 bit float like Apple and Sonnox, if some small changes were made to Global Analysis to remove or lessen the artificial, less compatble limits (minimum time between reported points, total number of points reported, points below 0.0 reported) that are not imposed by the other programs, Sonnox, Apple, Audacity.

Wavelab has Encoder Checker with blind test, and Global Analysis supporting all lossy formats, AAC, MP3, Ogg. Maybe some of the others don’t support Ogg.

If the simple built-in Mac MFiT roundtrip routine (built in src, encode, decode, analysis, and wav/clip output) was added in Wavelab Mac (making Wavelab MFiT compliant), and a clip safe script was added on Mac and Win, I think Wavelab would be the most comprehensive lossy analysis available on any platform, and the only one with MFiT in a mastering program.

I forgot about WMA. I think that’s also there in Wavelab, at least on Windows.

EDIT: sorry, apparently not. Don’t see WMA in the Encoder Checker on Windows.