Hi fox_m, thank you for your long reply.
If you can't get rid of the clipping, your original audio is clipped. What DAW are you using? (Audacity?) If you experience clipping in FCPX, your audio is already clipped.
My main DAW is Ableton Live (Audacity is not a DAW lol). But as far as I know this problem exists with any DAW.
You should never (never, ever) normalize audio. Did I mention you should never normalize? Also, you shouldn't be using MP3 - uncompressed, either AIFF or WAV, 24bit/48K if possible (although 16bit and 32bit float are usually acceptable.)
In the music business, "Normalization" is key. One of the goals of mastering is to raise the volume (make it sound as loud as possible), without the audio ever reaching 0db. Especially in Popular music (that's my field), if a track is not loud enough, that's considered a bad thing.
For video editing, I always use uncompressed formats, the one I use is AIFF. What I meant by saying MP3, is if you take an officially released MP3 by a renowned artist (which is always normalized, depending on the Genre, of course you don't want classical music to be as loud as possible ) and import it into Final Cut Pro X, the audio clips, if you don't lower the volume of given track.
That means, clipping ONLY occurs AFTER importing the music into Final Cut, before importing it of course does not clip.
Normalization artificially raises (or lowers) the gain of all the *entire* signal to reach an AVERAGE peak level (therefore, already "loud" segments will be clipped if there are very quiet segments in the music if you're normalizing louder.) Normalizing does not change the shape of the overal audio signal - it just moves it up or down. Going up you run the risk of clipping. Going down you run the risk of cutting out low (quiet) signals completely. Either way, your audio can become damaged. ("Normalized -2dB" or "normalized to -0.30dB for YouTube" are meaningless statements... there's no indication what the peaks are, the average RMS or anything else since the final audio depends entirely on the original source.)
OK, so I actually never have come across this type of normalization, I know that's a thing, but the normalization process I'm talking about (and if you say normalization you normally mean only that), is peak normalization. What you described is RMS or Loudness normalization, which is very rare and pretty much is not a thing in the music business, I could be wrong, but I've honestly never seen that with music. What I'm talking about is looking for the highest peak in the sound file and just raising the overall volume, so that this peak is at 0db.
Here's a paste from Wikipedia of what peak normalization is:
One type of normalization is peak normalization, wherein the gain is changed to bring the highest PCM sample value or analog signal peak to a given level — usually 0 dBFS, the loudest level allowed in a digital system.[1]
Since it only searches for the highest level, it does not account for the apparent loudness of the content. As such, peak normalization is generally used to change the volume in such a way to ensure optimal use of the distribution medium in the mastering stage of a recording.
One example is music videos: if your music videos are too quiet, because you wanna work "safe" and stay at like -6 db, if the viewer watches a playlist and your video starts, they have to reach out to the volume setting and adjust it. Maybe they even can not if they're watching on a mobile device or quiet headphones for example. In that case you'll just have a quiet music video, which statistically speaking makes the music sound "worse" for the modern listener. Today, louder music is perceived as "not as good" music, unfortunately, the days of great dynamics and very little compression are sadly over in popular music
Let's stick to the playlist example: if the viewer CAN adjust the volume to make it louder on his playback device, the next video will probably be louder so they have to turn down the volume again. Imagine if that music is playing in the background, that would be pretty bad...
Get some decent output from your DAW that isn't normalized to bring into FCPX. It's all digital, so it can be quieter... just not clipped. It *should* be "well-balanced". All instruments should have good, positive correlation. Make boost adjustments (compression/limiting) at the end of the process (in FCPX).
As I said in my original post, that's not an option if you're working with music. The compressing and limiting is done in the DAW, and ONLY in the daw. You can't just lower the overall volume and throw a compressor in FCPX on it, that would alter the music (pretty much destroy the mastering process)!!
In a scenario of a movie, that's a different story, since you don't only have music, but lots of audio tracks that should be compressed and processed in the video editing software. In my situation it's often JUST music that I'm working with in my videos.
I pretty much know my stuff with limiters and compressors and how to use them, but I'm sure your advice and tricks are great for beginners that read that topic
I myself have my own plugins that I use for audio processing (like waves, PSP, mAudio and so on), I haven't really tried using Waves for Final cut so I'm not sure if they work, but PSP plugins are my choice for FCPX.
So yeah, the problem is still there, and I still don't know why or if Final Cut is raising the levels of audio tracks that are clearly not peaked.
I'll maybe run some tests with comparing the output before and after importing into FCPX, they might be interesting to look at for others, too.