Sidebands; Intermodulation                                                  Back to Kjaerhus MPL-1 Pro Second Edition Vs. Others        Back to fader8's audio blog Index

Measurements at 96kHz, 24 bit, 0.7Hz resolution, Blackman-Harris windowing.

Limiters are tricky to test. While it's common practice to set them up so they operate only for extremely short durations to control transients, they still offer their "impression" during those moments. But where they can have a real sustained impact on the sound is with regard to bass, which can have very high sustained energy in a mix.  With the exception of maybe some Hip Hop, bass is typically a more complex combination of tones. Unfortunately, this doesn't sit well with limiters. When these signals get clamped down, the system becomes non-linear and the signals intermodulate to create non-harmonic distortion products, i.e. not the "good" kind. 

So this little page is all about bass, (tones, not fish). Note that the artifacts presented here will only occur if the limiter is directly effecting the bass signal. The processors presented here are all tested in their Audio Unit format.

Test Signal: Low distortion sinewaves at 40Hz and 185Hz. The file's output at unity peaks at -1dBFS, as shown on the PPM meter. 

It's a steady-state signal.

Here are the settings for the tests. I tried to stay as close to 1dB gain reduction for all of them. That's usually what I keep things down to anyway. Setting the release times all to 1 second keeps that common to all, except those that have no release parameter available. I noticed little difference in results when toggling between a 1 second manual release and any program dependent/auto release modes these may have.

Kjaerhus MPL-1 Pro Second Edition:

Universal Audio UAD-1  Precision Mastering Series Limiter

The idea with these settings is to get 1dB of gain reduction and then recover it to -1dBFS at the output. The PPM meter is inserted after the limiters. Both of these were a breeze to dial in these settings and get good level matches.

The UAD and the Kjaerhus are pretty competitive here. But the Kjaerhus has the edge. The PPM meter is showing the MPL on the left and the UAD on the right.

On to the Voxengo Elephant. I have to say it wasn't quite as easy with this one to dial in the settings to match the previous two. A little persnickety. Metering could be much better. 300mS is the longest release available. In the mode editor, setting the transient time and shape appears to allow compromising between low and high frequency intermod products. A useful ability, in this case.

Since the Kjaerhus faired a little better in the previous comparison, I'll use it as a benchmark here for the Elephant, shown here in blue.

Again, the PPM shows the Kjaerhus on the left.

And the venerable Sonnox Oxford Limiter. I tested the limiter section of the Oxford Dynamics too.

Note that when using the Enhance control, it will significantly increase the level of bass intermod products.  

Here's the Oxford Limiter in blue. It's fascinating how different these all are. I'm not including another plot for the Oxford Dynamics as they perform identically if you're not using any of the additional features of the Limiter.

Here's the Elemental Audio Finis, now an RNDigital product with a weird name.

Here's the Finis in blue. I used this one for a long time.

On to the Flux Pure II Limiter.

Here's the Pure II in blue. Not too far off from the Finis.

Mustn't leave out the Waves L2.

Here's the Waves L2 in blue. Not too shabby. Don't touch that ARC button though, or all hell breaks loose. At least on this signal.

Oh yeah, the PSP Vintage Warmer 2.

Vintage Warmer 2 in blue.

And in the "just for fun" department, here's the SSL Duende X-Comp trying to be a limiter . . .

X-Comp in blue.

In summary, this test only regards the intermodulation of bass frequencies and makes no claims as to the value of these limiting processors. While it's clear that the Kjaerhus shows a strong advantage here, others aren't too far off and offer different features and may well perform better under different circumstances. This is only looking at one very specific measurement dimension, under very specific and "non-dynamic" conditions.

   Back to Kjaerhus MPL-1 Pro Second Edition Vs. Others                                      Back to fader8's audio blog Index