Dynamics Trouble: Peak Limiting
Peak limiting is just one of a number of processes that can be applied to an audio track, but it may be the most common and most damaging of the amplitude modifying effects. The basic idea of any dynamic processor is to alter the natural amplitude variations as produced by a musician. As any performer will tell you, it’s virtually impossible to play any two notes or events exactly the same. Machines do this very well but humans don’t. Of course, great performer produce great performances specifically because they are able to subtly alter the loudness, pitch, and timing of their music.
A peak limiter is an automatic process that detects the amplitude of an instantaneous musical element or sound and has the ability to attenuate the signal level. There are a bunch of parameters associated with a peak limiter that allows the engineers to fine tune the processing with regards to the attack and release of the processing as well as the amount of reduction that will be applied. There’s also a knob that can be used to “make up” the gain that is lost due to limiting.
During the pre-digital days of record engineering, we would use dynamic processing sparingly on vocals, guitars, drums, and percussion instruments as a preventative measure against overloads and distortion. A singer has the ability to quickly change dynamics. One moment they’re singing softly and the next they belt out a note that sends the engineer scrambling for the input fader or record level knob. A peak limiter provides a measure of safety in those instances. No one can react fast enough to an instantaneous peak nor would you want to instantly pull the fader down 20 dB. Instead, the limiter detects the increasing level and over a small amount of time (usually 15-25 ms) it reduces the amplitude.
Peak limiters can also be used to shape a sound creatively. If the engineer wants to “tighten” up the bass part and smooth out the dynamics, a peak limited with long attack and release times and an aggressive gain reduction ratio can do the trick. Does this remove some of the player’s intentions? Of course, but if a producer or engineer feels that the music would benefit from that type of signal processing then it’s there choice to do whatever they want.
In a well-equipped analog studio, there might be a dozen dynamic processors. The outboard rack would house some peak limiters, some gates, expanders, program limiters, and hybrid devices. But these machines were used for specific reasons…and there were treated as special. These days every input or track in Pro Tools can have a slew of dynamic processors activated. The input microphone can be compressed, the bus output can have a limiter in line, and during the mixdown yet another dynamics processor might be applied…and all of this happens prior to the mastering session.
It’s almost as if recording engineers and mastering guys are in a competition to see who can create the loudest records. They’re winning and we’re losing. If you think you’re better off going to a live event to avoid this sort of dynamic intrusion…think again. All of the same processes are used during live concerts, too.
There’s a place for dynamics tools in crafting a hit record but consumers have no say in what they prefer. And unfortunately younger listeners have never heard a track with real world dynamics.
Dr.
Again, thanks for taking time to keep this blog alive, it keeps me thinking. Bear with me, I will get to a question. As you might remember I was a motion picture guy, 30 years behind the camera. When I left the industry digital was in its infancy. Today film is a very rare luxury. I just read an article by a very imaginative Director I got to work with. Doug Trumbull, (blade runner). It was about frame rates, and how there really is not a need for them or soon will not be. So to the question, where in the digital capture of music is there a concern for saturation?
I’m not sure saturation per se relates to digital audio coding, recording or playback. Frame rates certainly can be aligned with sample rate and perhaps bit depth could be tied to color depth. But saturation is a value used in reference to intensity of color, right? The dynamic range is close.
This is perhaps a significant reason that Apple, and most of their consumers, don’t care about lossless files, let alone higher bit depths and sample rates: the majority of tracks they’re listening to are such that the benefits are negligible, or possibly even too revealing for this kind of production.
Very true.
Accurate Dynamics have always been one of the most important aspects of music reproduction, and one that has been the least respected in high end audio circles. We threw away our high efficiency speakers of old for acoustic suspension designs in search of the ultimate flat frequency response graph, forsaking dynamics, harmonic distortion, etc.
We need Bob Carver to bring back a modern version of his Phase Linear 1000 Auto Correlator Noise Reduction/Dynamic Range Recovery unit. Worked fairly well to restore the necessarily restricted dynamics of vinyl and reduce it’s surface noise. Something really effective could be designed today with digital tech.
Starting about 25 years ago I had PH1000 in a system comprised of Klipsch La Scala’s driven by a VTL 100 tube monoblocks together with a pair of 7 foot tall Hsu subwoofers driven by a pair of NAD 2400THX amps running in the bridged mode and delivering tons of power to the subs.
That’s still one of the most dynamic and “REAL” sounding systems I ever heard. Broke my heart when I had to sell the system to retire and move to a much smaller residence in Fl.
Now THATS Rock and Roll.
LOL
Sad but true. An entire element of the vocabulary of music- dynamics- is being simply tossed out the window. No wonder so much of today’s music sucks.
All the live gigs I go to are small, and as such I get as near to the stage as possible – next to it if I can – just to avoid the PA. Even if the mixing engineer’s not messing with the sound too much I prefer to get as close to the musicicans as possible to hear them, or their own amplification, and not the PA.
Mark, I was watching you on Home Theater Geeks and I just want to ask you directly: Don’t you believe that the Loudness War has a much more profound effect on the degradation of fidelity than the bit or sampling rate?
Absolutely. The demands of the commercial music business does more to limit great sounding records than the formats, sampling rates, and encoding flavors.
“All of the same processes are used during live concerts, too.”
That is a good point – and a very important too.
Some live events are so miserable, that you hardly can stay there (unless you wear hearing protection, and already have had a couple of beers).
I used too think, that it is because I am getting ‘old’.
But lately I have come to more or less the same conclusion as you described above.
The guy(s) at the mixing console(s) are going totally crazy – I wonder, why the musicians don’t notice that ;-(