Adobe Audition CS6 - process opening MP3

[Alejandrootje]

Hello.

I have a 1000 MP3's that I like to process: matching volume and maybe some editing (fade in/out, ...). A few years ago, I read that if you do this on an MP3 and you save the file again, you would have (some) data loss.

To avoid this I could convert them all to WAV, do the necessary things and save them back as MP3. But with 1000 MP's this would be crazy.

My question is thus:

If Adobe Audition CS6 opens an MP3 (or another format) file, is it instantly opened " as a WAV file ", so one can process it (being a wave and not a data file)?

If yes, the data loss problem is avoided, as well as processing 1000 MP3's from MP3 to WAV to MP3 (...).

Thanks for an answer in a short notice.

Friendly greetings.

Oli4

[kglad]

[moved from Creative Suites to Audition CS5.5, CS6 & CC]

[ryclark]

Yes, Audition's native format is .wav. So any .mp3 file opened within it will be automatically converted to .wav as they are opened. If you have enough hard drive space it might also be a good idea to also save out all the .wav files so that you can come back to them later if you need to do any alterations. So ideally you would do all the necessary editing and topping and tailing with the .wav files and save the edited .wav files. The only convert them as a Batch Process when you are sure that you have them as you want.

[Alejandrootje]

Dear Ryclark,

I'm glad I can follow my idea to process them, based on your answer.

Some MP3's date back to more than 8 years ago, so the original WAV files I don't have them anymore. No harm done. It's not that I am a professional DJ or so.

I have a few (final) questions for you (if I may):

All my MP3's are now in (joint) stereo. In Audition I can see the option 5.1 when using matching volume's ITU-R BS.1770-2 Loudness algorithm (thanks for this invention). Is it a good choice for the future? Or should I just stick with (joint) stereo as the best choice?

It also seems that 32 bit depth is not adding much extra data to the file (in comparison to 16 bit depth). Which should I choose?

Thanks already.

Friendly greetings.

Oli4

[ryclark]

None of that is going to make any difference since your source files are already .mp3. So just stick with the basics and leave settings as they are. Nothing is actually going to improve the audio quality in any way. The idea is to preserve it as much as possible without it getting any worse.

Audition works at 32 bit depth natively so don't worry about that.

If you only want to match the levels of your .mp3 files there are some apps that can change the metadata within an .mp3 file to may it play back at a different level. This can be done without having to decode from the compressed format and re-encode with subsequent loss of quality.

How to Normalize MP3 Files to Play at the Same Volume

[Alejandrootje]

Dear Ryclark,

Could you please answer my questions so I can have the necessary information I'd like to know, because it's the technical issues that I want to understand?!

1)     Except the fact of some " deleting " of low and high " notes " of the music (these that we as humans don't hear) by converting

        a WAV to MP3 (which i.a. decreases the data of the file), I would expect that, when opening an MP3 as WAV (which is what         Audition does according to your previous answer), and re-encoding with different settings [44100 (higher/lower) Hz sample rate, 320         Kbps bitrate (higher/lower) and 16 bit bit depth (higher/lower)] would make at least " some " (technical) difference (whether you can         hear it or not)?

2)     I'll rephrase my questions in my previous post: which settings should I use to have the best audio technical vs. data management         effect (the list following is what I'm using 'till today), when converting WAV to MP3?!

        Sample rate          44100 Hz (CD quality)       for audio work

        Bitrate                   320 Kbps (CD quality)       the highest value you can set in Adobe Audition

        Bit depth               16 bit (CD quality)              24 bit would be for DVD quality? 32 bit is used for?

        (joint) stereo         does choosing 5.1 make any difference when re-encoding, for the future use of my MP3's?

                                     when not having - at this time - a 5.1 device for playing, the 5.1 MP3 would be played as stereo, right?

                                     I was thinking if you have no original 5.1 " multitrack " WAV/MP3, re-encoding a (joint) stereo MP3 to 5.1 has no                                      use?

3)     What did you mean with " Audition works at 32 bit depth natively "? How do I put that in perspective using Audition?

        I mean: when exporting a WAV to MP3, the program let's me choose settings: e.g. lower than 32 bit depth for example.

4)     I'm familiar with the normalise software but that's not how I want to fix " my problem ".

Thank you very much for the time you invest(ed) in me.

Friendly greetings and enjoy your weekend!

Oli4

[Alejandrootje]

Dear Ryclark,

I just read that it's important that the song/mix/... doesn't exceed 0 db on the levels (dBFS).

So I want to use limiting when processing the ITU-R BS.1770-2 Loudness algorithm on my songs. But with e.g. Hard limiter you can specify the value of this setting. What's the value set in this " use limiting " in the match volume effect?

Thanks again.

Friendly greetings.

Oli4

[Alejandrootje]

Dear Ryclark,

I was wondering why the next thing happens:

For the (ITU) Loudness setting, when using the ITU-R BS.1770-2 Loudness algorithm, I've chosen -10 (CD's: > 9, party e.g. between -7 and -13, ...).

But Auditions sometimes need 2 to 3 passes (each - necessary - pass, I export to an other location, to avoid the mentioning of copy in the file name) to have the loudness " arriving " at -10.

How come?

Thank you for your help.

Friendly greetings.

Oli4

[ryclark]

A lot to answer there.

1) .mp3 encoding isn't as simple as that. It is very complex, not just removing high and low frequencies. So your audio has quite a lot missing depending at what settings it was encoded. Audition won't make any difference to that. In fact it will appear completely transparent to what is supplied to it by the .mp3 file. It's when you re-encode back again to .mp3 that you will lose even more of the audio, even if you do it at a high bitrate.

2) Having already lost audio fidelity by the original encoding process it probably won't make much difference what bit depth you use for the re-encode. I would stick to 16bit at the highest bit rate you can. 5.1 doesn't come into it as that is a surround format which your original files aren't. So no point changing that setting.

3) All that the 32bit floating point format that Audition uses just means that it will be completely transparent to anything that you do to the audio when editing. It means a lot else as well but we won't over complicate the explanations here.

4) Normalizing software for .mp3 files is the only way that you are going to get changes in your playback levels to match tracks without losing more quality by having to re-encode from .wav files. Why not just keep the files a .wav files. Most things play them these days.

Yes it is always extremely important that no audio exceeds the 0dBFS because you have run out of bits to describe the audio above that. And again if encoding to .mp3 it is better not to get anywhere near that level since the encoding process can produce higher levels that are in the .wav file. Always stick to no higher than -1dBFS, I always tend to go no higher than -3dBFS myself.

As far as Matching Loudness is concerned the ITU-R BS.1770-2 Loudness algorithm is really the correct one to use in your case. It is really intended for the broadcast industry to make sure that their programmes fit with in broadcasting standards. I would try one of the other algorithms like the Perceived Loudness (Legacy) one for your purposes in trying to make all your .mp3 sound the same volume.

[Alejandrootje]

Dear Ryclark,

Thank you for your quick response to my questions.

I believe you're living in the USA (which is a +/- 12 hours difference with Belgium, where I live)?

Finally, I'd like an answer to the following 2 technical questions (asked in my 2 last posts, but yet not answered):

1)    What's the value of the " use limiting " in the " match to: IBU-R BS.1770-2 (Match Volume Settings) " effect? It seems my files peak at -0,1        dB (in the dBFS - levels - meter). Is there a way to set this level (like in the effect Hard Limiter) (the same question appeared in an

       old post I encountered on the Adobe forum, dating from 2012)?

2)    For the (ITU) Loudness setting, when using the ITU-R BS.1770-2 Loudness algorithm, I've chosen -10 LUFS,

       but Audition sometimes need 2, 3 to 4 passes (each - necessary - pass, I export to an other location, to avoid the mentioning of copy in the        file name) to have the loudness " arriving " at -10 LUFS (e.g. 1st pass: -12, 2nd pass -11, 3th pass -10). Why is this sometimes the case?

Thank you very much for your patience.

Friendly greetings.

Oli4

[ryclark]

No I am actually in the UK.

1) The limiter is just there to stop any level manipulation of the audio, ie. increasing the level to meet the standard, from clipping if the odd peak gets too high. The setting of True Peak Level actually decides what your final peak level will be.

2) I have no idea why you need to do that many passes. It may depend on how many files you are processing and what their spread of average levels are before processing. However, as indicated in my previous reply, I believe that using one of the Legacy loudness algorithms would work better on your material in your case since you are not submitting your audio for broadcasting.

[Alejandrootje]

Dear Ryclark,

Good morning.

I'm a huge fan of Fawlty Towers, A touch of Frost,

..., Britisch good stuff. 

So, the mentioned setting I can set through the menu of effects and choose True peak level (Amplitude, ...) and then set the value? Can you maybe post a screenshot where to find it in Adobe Audition CS6?

I tested the mentioned Legacy algorythms in the past, and they seemed to me, leaving to much difference in (perceived) loudness levels.

Thank you for your patience.

Friendly greetings.

Oli4

[ryclark · Accepted solution]

I had forgotten that you were using Audition Cs6. True Peak setting for the BS1770 algorithm didn't appear until a later version of Audition. So unfortunately other operations of the Match Volume also may not operate in quite the same way as the Match Loudness feature of the later versions of Audition. It might be worth your while downloading the trial version of Audition CC to see if that works differently on your audio files.

[Alejandrootje]

Dear Ryclark,

A busy week, so far.

I downloaded the trial Adobe Audition CC 2017 and there I can set the value for the TRUE PEAK limiter.

It seems that the higher the value of the ITU-R BS.1770-2 (or 3) Loudness algorithm, the more difficult the software has

to get to the target value (in my case I chose -10 LUFS with no tolerance). When e.g. I take a value of -14 and allow some tolerance (0,5 LU e.g.),

the target value is no problem. Maybe it has also to do with the combination of the Max True Peak Level value and the other settings?!

FYI

Thank you for all your help.

Friendly greetings.

Oli4

[SteveG_AudioMasters_]

alejandrootje  wrote

A few years ago, I read that if you do this on an MP3 and you save the file again, you would have (some) data loss.

To avoid this I could convert them all to WAV, do the necessary things and save them back as MP3. But with 1000 MP's this would be crazy.

The whole point here is that if you do this, you won't avoid that quality loss - that's exactly where the loss comes from; opening the MP3s and decoding them to wavs, altering them (or even not altering them, come to that) and re-encoding the file back to an MP3. Once they are compressed, there is very little at all you can do to them without damaging them - except to play them. MP3 is essentially designed to be a distribution-only format.