Stuck with a clipping track...

[SwitchBack]

Distort = pull or twist out of shape.

Any dynamic processing distorts by the pure nature of it, changing the signal irreversibly (unlike e.g. EQ'ing). It certainly doesn't have to sound bad and it can even be inconspicuous to the untrained ear, but it won't be the same signal anymore. 'Shaving off' is the basic nature of limiting. Sure you (or the chosen limiter) can be smart about it, but in the end the peaks/projections will have to be gone-ish or the signal will clip like before.

My post was about the basic process of fighting clipping, setting it apart from the pure loudness of a mix. By calling it an art I also wanted to say that the subject is deeper than just adding a simple limiter.

 

[Sebastian Holtzer]

That is not, how it is related. Any codec that alters the source material has a chance to go above 0 dBTP although you mastered at -1 dBTP. The factors are not the codec used, but the musical situation and how the codec deals with it. But you said "clip higher". I want to point out that there is no clipping. We are in the analog realm. There is no such thing as "it clips, when over 0 dBTP". In the digital realm the limit is absolute. Higher than 0 dBFS and it clips. In the analog realm that doesn't exist. Instead it is dependent on the hardware used to listen to your music, what exactly might happen. There definitely will not be clicks or something like that. On most hardware you won't even notice occasional peaks well above 0 dBTP.

If your goal is not to provide a solid, good sounding source, but to prevent the digital master to ever go beyond 0 dBTP, you have to consider the LUFS goal, you master for. For example, at -16 LUFS, -1 dBTP will guarantee that (scientifically, so the usual "with a probability bordering on certainty"). But at -6 LUFS you might need to tweak it down to -2 dBTP. that is all just a broad example, in the end the specific music with the specific LUFS goal and its specific dynamic range decides that. And you will never know it, because only on the end user's hardware is it actually happening or not happening, without the user even noticing it.

But, this is then just a master that prevents occasional peaks above 0 dBTP at all costs. It doesn't sound better, it doesn't do anything positive for your music. It just prevents a digital value from appearing, before it gets transferred into the analog realm.

OK, thank you for the clarification. I believe you understood my point: a signal mastered to a -1.0 dBTP limit is generally safe for streaming or platform-based normalization. However, when a track is purchased—on platforms like Beatport, Bandcamp, Apple, Amazon or so on—it can be downloaded in various codecs. This creates a real challenge for most artists, as the source file is typically a 24-bit/44.1 kHz WAV, yet there is usually no way to provide a separate, codec-specific master for each format offered by these platforms.

If we assume that a conversion process inadvertently generates inter-sample peaks (ISPs) that exceed the -1.0 dBTP limit set in our master, we arrive at the crux of the matter:
If I’ve understood you correctly, the quality of the listener's playback device then becomes the deciding factor in whether or not we hear artifacts caused by clipped peaks (which were due to the conversation originally above 0 dBFS but not on the original .wav). Is that correct? Please confirm—I believe this aligns with what I was assuming.

Given this, and looking at the analysis provided by my distributor—which shows that 96.2% of my streams occur on mobile devices—I have to wonder: might a -1.5 dBTP limit actually be worthwhile? Furthermore, I assume that when users download tracks for offline listening, they may end up with lower-quality codecs—at least, that is my technical understanding. Please correct me if I am mistaken. For me, this exceptionally high proportion is further evidence that purchased content in which codec ever is indeed being played back on these very devices—and I highly doubt that everyone in Germany is willing to spend between €250 and €580 on Apple AirPods. This is especially relevant given that the economy here has been in a sharp decline for quite some time—to the point where people are now looting shops during public gatherings, making off with, say, three left-footed Nike shoes. And, according to the analysis, Germany isn’t even my top-performing country.

In this context—specifically for electronic music (leaning toward trance)—would you consider it more advisable to master to -1.5dBTP, while aiming for a loudness level around -10 LUFS (rather than -6 LUFS) and attempting to maintain a dynamic range of approximately DR9?

***
>Translated with Google<
***