Clipping in OC remixes and modern soundtracks - loudness war versus loudness norms

[Master Mi]

I noticed these phenomenons like partial frequency clipping or even peak clipping in lots of remixes from OCR and also in many official modern audio productions quite a while ago when I loaded a few audio files into my DAW to compare certain parameters and audio measurements for different soundtracks.

And I think that such technically obviously not quite clean masterings of today's era are above all the direct consequence of the constraints that the loudness war has brought with it since the beginning of the 90s with the dawn of the era of digital music production.

And once the music industry sets such rather counterproductive trends for mainstream music in terms of music quality, many musicians seem to feel compelled to follow suit in order to keep up in terms of loudness or to be noticed at all.

But just as - at least according to their aggressive endless advertising - the arms industry stands for peace and security in the world or the pharmaceutical industry for people's health (while the reality often speaks a completely different language), the same is obviously true of the modern music industry and music, especially the quality of many of today's music productions.

And by that I don't mean the often first-class compositions, but rather these often radically messed up, because far too loud, undynamic and therefore usually not very lively masterings and remasters (if you want to hear more dynamic and lively music from certain bands, always get the original works if possible and never any remasters or even remasters of remasters).

Even in today's music industry, making a quick buck is often more important than the actual thing - and then it ends up being crap again.


I will show this below with corresponding negative and positive examples of soundtracks from different times in the music industry and certain forms of individual approaches to mastering and music production in general.

…


1) David Guetta Feat. Sia – Titanium
---------------------------------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:
 

While David Guetta is literally a guarantor for really groovy and impressive compositions, when it comes to mastering he is rather a real negative example of how not to do it.

Here you can find massive clipping in the bass range of up to over 9 dB above the zero point (see Spectroscope/Spectrum Analyzer in the middle right of the picture), while the measurement displays are apparently no longer sufficient to capture the true extent of the mastering excesses.

The true peak measurement shows a value of a whole decibel above the zero mark, which should never be exceeded in a technically clean mastering, as this leads to audible clipping in some cases.

And the loudness of this mastering was driven so close to or even above the maximum limit by excessive use of compressors and limiters with a loudness value of around -8 dBFS (the exact loudness value within the complete loudness measurement is only obtained at the end of the soundtrack, but the loudness has certainly not changed much after the first half of the soundtrack and the heavy compression) that any efforts to achieve dynamics and sound quality in this soundtrack were literally beaten to death with the digital mastering club.

But this is probably how it must be when the music producer sees mastering as the final boss and then shouts to the mastering engineer: "FINISH HIM!"
(with the iconic "Techno Syndrome" theme from the Mortal Kombat movie playing in the background)


2) Donkey Kong Country 2 - Prismatic (OC Remix)
---------------------------------------------------------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:
 

This is definitely one of my many favorite soundtracks from OCRemix, which combines a really beautiful, energetic composition with impressive sound design.

The mastering here is also very loud with a measured loudness value of around -11 dbFS and technically not entirely flawless with similarly strong clipping in the bass range and a slight overshoot of the level peaks at 0,3 dB above zero.

But at least the mastering here doesn't seem to have been completely fucked up compared to the mastering of the previous David Guetta soundtrack and offers a little more dynamics and sound quality.


3) Final Fantasy 7 Remake - Let The Battles Begin! - A Merc's Job
---------------------------------------------------------------------------------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:
 

This is an original soundtrack from the video game Final Fantasy 7 Remake and at the same time a really powerful, modern new composition of the standard battle theme from the old Final Fantasy 7 for the Playstation.

In such a world-class production (and a Japanese one at that), I would never have guessed that there were apparently minimal technical flaws in the mastering (slight clipping in the bass frequency range and minimal exceeding of the 0 dB limit at the peak levels with a value of 0.1 dB, even if this value remained at 0 dB for a long time during the loudness measurement) and that Japanese sound engineers would let them get carried away with something like this.

But I could imagine that they really had a lot of time pressure in this mammoth project and perhaps couldn't subject every soundtrack to an exact loudness measurement again.

And perhaps the mastering engineers, as is often the case in the music industry, were given exactly the same guidelines by the leading developers and producers, contrary to their own aspirations and rules, and were told: "As loud as possible, but also with a sense of power and dynamics, with gentle clipping if necessary."


4) Bridges - Miss Eerie
--------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:

Many people probably don't know it.

But the band Bridges, which was founded back in 1978, was the immediate predecessor band of the well-known Norwegian synth-pop band A-ha, while the soundtrack "Miss Eerie" was virtually the early version of the cult hit "Take On Me".

Although “Take On Me” has more impressive vocals, “Miss Eerie” has some really creative, effective compositional interludes and variations to offer.
For example, I really like the melodic, multi-octave playing of the steelpan synthesizer from minute 1:36 onwards in the soundtrack, which gives the whole thing a very refreshing individual touch away from the later mainstream productions.

The mastering (which may not even be the original mastering, but a remaster) is also convincing in terms of production quality - technically absolutely clean, with no critical values (neither in the spectrometer nor in the measurement display of the maximum level peaks) above the permissible limits.

The maximum level peaks may still have been -3.2 dB at the time of the screenshot.
But towards the end they were also just below the zero mark (which is still within the limits for a technically clean production).


5) Michael Jackson - They Don't Care About Us (Brazil Version)
-------------------------------------------------------------------------------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:
 

Throughout the history of his musical career, Michael Jackson has not only stood for unique world-class compositions, but also for excellent music production at the mixing and mastering level, which has not so easily bowed to the bad trends of the modern music industry, but has gone its own way, which did not aim to win the small-minded race in the loudness war.

According to my memory, Michael Jackson once mentioned somewhere that music should always be able to breathe a little.
He was probably alluding to the desire to make his musical works as dynamic, natural and lively as possible.

And this is only possible if the loudness is not set to such a high level during mastering, so that the level peaks can rise up freely without the forced use of compressors and limiters, which also preserves the transients and maintains the sound quality.

As you can see here in the measurements for the soundtrack, the loudness measured from the beginning is only around -20 dBFS, even shortly before the end of the soundtrack (see the position of the marker at the top of the waveform after the 120th bar).
This is only slightly louder than the loudness standards I use to master my own soundtracks and remixes (around -23 dBFS, or more like -23,1 dBFS).

And despite the low loudness, you can clearly see from the measurements of the maximum level peaks with a value of -0.2 dB that the level peaks in the soundtrack obviously need this low loudness in order to be able to level up freely and unhindered without exceeding the 0 dB mark.

Apart from the more than sufficient headroom in the spectroscope measurement across the entire frequency range, the waveform of the soundtrack also clearly shows that this is obviously a very dynamic and lively example of a sophisticated music production.


6) Treasures Of The Deep - Shipwrecked
----------------------------------------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:
 

“Shipwrecked” is an original soundtrack from the underwater exploration action-adventure video game Treasures of the Deep for the Playstation.

Apart from the truly atmospheric, ambient composition, this soundtrack is also a positive example of successful music production in the field of video game music in terms of mastering.

The spectroscope measurement makes a relatively relaxed impression along the frequency range shown.
The maximum peak levels of -0.4 dB measured also appear to be within the bounds of technically clean mastering.

And with the measured loudness of around -18 dBFS, no unnecessary excesses appear to have been made here either.


7) Crisis Core - Final Fantasy 7 - Wings Of Freedom (Master Mi Remix)
-------------------------------------------------------------------------------------------------------------------------

Soundtrack link:
 

Visualization of the measured values for the soundtrack:
 

For comparison, these are the measured values from one of my soundtracks and remixes, which I regularly mix according to the loudness standard of the EBU R 128 norm with a target value of -23 dBFS.

It's possible that I will develop my higher compositional potential over the coming years.

But I believe that I'm already on a really good, future-oriented path in the area of mixing and mastering, which might be able to solve numerous problems in today's music production.

As you can see in the picture with the measurements, none of the measured values shown here are even dangerously close to any limit value.

The maximum peak level values show a completely relaxed peak level control at the end of the measurement with a value of -7 dB.
Even the measurements in the spectrometer still show more than enough headroom over the entire frequency range, although this is already a very dynamic remix in the classical-orchestral and rock ballad music genres.

I believe that the far-sighted developers and sound engineers behind the EBU R 128 loudness standard had exactly this in mind, namely to develop a uniform, universally applicable loudness standard that works for all music genres and other listening programs and at the same time allows maximum dynamics, naturalness of sound and the highest sound quality without making the listener's eardrums bleed.

...

And yes, I think that such a generously designed, uniform loudness standard can solve almost all the problems in the field of music and audio production that the soulless trend of the loudness war has brought with it in the course of the modern music industry.

These problems include, for example, serious loudness jumps between different musical works and other audio programs from different music producers - or sometimes even loudness jumps between different musical works and audio programs from one and the same music producer (often caused by a lack of good loudness concepts in the field of music production, where the primary focus is still on maximum peak levels).

What's more, a generously designed, generally applicable and, at best, globally enforced loudness standard can also guarantee the return of high dynamics, liveliness and sound quality in musical works, regardless of the music genre or type of audio program, the respective TV and radio station, the streaming platform, etc.

In the near future, it may be possible to develop playback devices that use an integrated software solution to bring music and audio programs mastered at different volumes to a uniform loudness level and play them back accordingly.

If this is approached wisely, all music and audio programs in the world could be brought to a uniform loudness level (so that there are no more annoying loudness jumps between different music works and audio programs and the volume on the playback device only has to be set correctly once according to your own listening habits and loudness preferences), and at the same time the best possible dynamics, liveliness and sound quality could be guaranteed for music and all other audio programs.

…

And since we're in a very large, multifaceted and international musician and remixer forum here, I'm particularly keen to hear your opinions and experiences on this topic. ))

But I would also be happy if you just wanted to share something about the phenomenon of clipping in certain frequency ranges or peak clipping, which I have noticed in numerous remixes from OCR as well as in many modern music productions and remasters in the field of modern mainstream music.

Edited 9 hours ago by Master Mi

[Unknown Pseudoartist]

Hey! Honestly I didn't read 100% of what you wrote yet, but I seen you posting similar walls of text regarding concerns on music production, and the kind of crussade you are going is interesting, to say the least. People may agree or disagree with your thoughs, and may like or dislike your own mixes and stuff. But at least you are questioning these things and going your own way with them, rather than simply pretending to fit the current trends on whatever works or sells best, and that's an attitude I personally applaude.

My own concern with the so-called loudness war as a musician now is more about a simple fact: to make mixes that sounds good and clear can be hard enough, but to make it to sound good and clear AND also loud as heck... definitively not a comfy thing. I'm not fond of mixes that are too much dynamic either, since it can be annoying (as a listener) to listen to music that can be totally quiet at a moment and extremely loud at another moment. It's just not a pleasant experience when you have to keep fingers at the volume buttons to adjust it. But everything sounding hella overcompressed and loud can be a headache for real, and so I totally understand and even relate to the special charm on older productions, even though oftentimes that might be a bit too raw.

Funnily enough I see the exact opposite trends on other media forms, like movies, series, animes, etc. When people is talking you have to rise volume because you can't hear the words right, but when action happens your ears are screwed and have to decrease it. It's like they assume everyone watches stuff in a home cinema, but that's a pain when you just use an ordinary computer in a small bedroom with regular 2.1 speakers and living with more humans at home in near rooms.

[Master Mi]

@Unknown Pseudoartist

I think the desire for more dynamic and consistent loudness levels between different soundtracks and other audio programs began back in the 1990s.

Perhaps some of you still remember those days when you would frantically reach for the remote control as soon as a commercial break suddenly interrupted your favorite show, often mixed much louder than the regular TV program.

You couldn't just stroll over to the refrigerator and grab a portion of lovingly prepared sashimi along with the rest of the hearty 7-course meal to make the most of the commercial break in a witty and culinary way, taking on the battle against continuous advertising stupidity with copious amounts of brain food.

Sometimes you really had to hold down the volume down button for a few seconds or, better yet, press the mute button on the remote control so that the stork wouldn't fall out of the neighboring tree in fright 'cause of the commercial sound pressure level.


With the development of methods and metering devices for measuring the loudness of sound events, soundtracks and all kinds of other audio programs, which ultimately led to the EBU R 128 guidelines created by far-sighted sound engineers, it was possible to satisfactorily solve not only the problem of loudness jumps between successive audio content.

At the same time, a standardized basis was created to allow for much more dynamics, especially in the field of music.

If you are interested in the smaller details and background information, I recommend the compact Wikipedia link on this topic:
https://en.wikipedia.org/wiki/EBU_R_128

…

Incidentally, EBU R 128 is not only a basis that allows much more dynamics in music and other audio contributions.

EBU R 128 also limits the dynamics within a soundtrack or other audio program in order to ensure an enjoyable listening experience without causing permanent hearing damage.
For example, you won't be startled by the extreme sound level of a realistic gunshot at close range in the middle of the gentle rustling of leaves in a film.

This is achieved by additional parameters within this loudness standard, which, in addition to the program loudness main parameter (where the entire soundtrack or audio program is measured from start to finish and a target loudness value of -23 dBFS is aimed for after complete measurement), would primarily be the following parameters:


1) Maximum Short-term Loudness
------------------------------------------------------------

This refers to the continuously measured loudness of the last 3 seconds throughout the entire soundtrack or audio program, which must not exceed a value of -18 dBFs at any time.
For example, -17.9 dBFS over a recorded period of 3 seconds would no longer be permissible.


2) Maximum Momentary Loudness
-------------------------------------------------------------

This refers to the continuously measured loudness of the last 400 milliseconds throughout the entire soundtrack or audio program, which must not exceed a value of -15 dBFs at any time.
Accordingly, sporadic sound effects such as a gunshot with a maximum level of -14.9 dBFS or higher over a measured period of 400 milliseconds would no longer be permissible.


It is not necessary to comply with the maximum values of both additional parameters together.
It is sufficient if either the Maximum Short-term Loudness value or the Maximum Momentary Loudness value is complied with during a loudness measurement for a soundtrack or other audio program such as a podcast or commercial.

In conjunction with the primary program loudness target value of -23 dBFS (with a permitted tolerance of +/- 0.5 dB), this limits the dynamics to a level that is tolerable for human hearing and at the same time provides an enjoyable listening experience.

You can read all about this in detail here:
https://tech.ebu.ch/docs/r/r128s1v1_0.pdf

…

And as you can see in the image here, after a complete loudness measurement of my Crisis Core remix from start to finish, this remix is also well within the permissible parameter limits:




I don't know exactly why the Maximum Short-term Loudness value is already flashing red here with a measured value of -19 dBFS (according to the EBU R 128 guideline, up to -18 dBFS would be permissible) - but perhaps the values have been slightly adjusted over the years.

And since EBU R 128 only requires you to adhere to the maximum value of one of these two additional parameters anyway, I'm still well within the green range with the Maximum Momentary Loudness value, although my main focus is primarily on the target program loudness of -23 dBFS.

...

Another interesting aspect in connection with the implementation of EBU R 128 in broadcasting is the fact that, towards the end of 2023, an extended guideline for streaming platforms was also initiated:
https://tech.ebu.ch/docs/r/r128s2.pdf

It would be really awesome if this became standard practice in the near future, because then it would no longer be worthwhile to stubbornly master soundtracks for maximum loudness (since the soundtrack would be reduced to -23 dBFS anyway), but instead every musician would have the same fair opportunity to make their compositions as dynamic and lively as possible at a standardized loudness level.

…

Based on EBU R 128, the resulting expanded dynamics in music, and my “Life Force” mixing concept, which I have developed extensively in recent years, I want to demonstrate with my upcoming video game remixes and some larger original compositions that dynamic music can sound really good, organic, and lively in every conceivable music genre - from classical orchestral music to electronic music to rock and metal - and that it can be superior to the often dynamically limited “compressor music", especially in terms of sound quality.

Edited yesterday at 02:20 AM by Master Mi

[DarkeSword 🎮]

This topic title is ridiculously long, please edit it down to something around 10-12 words.