Review – Pro-L 2 from FabFilter
Fab Filter has a well-earned reputation as a maker of top-end virtual mixing and mastering tools. That reputation expands with an upgrade on their already excellent limiter.
by Vincenzo Bellanova, Jan. 2018
In this review we will take a look at the new version of the limiter plugin from the Amsterdam-based FabFilter, Pro-L 2. FabFilter has established its name during the last decade with top-shelf mixing/mastering tools and effects. The process of “renewal” began some years ago with the upgrade to the surgical Pro-Q 2 and, then, with Pro-C 2. It’s now time for some new features on their limiter. Let’s see what’s new and then test it on some audio material.
What’s New: Algorythms
Pro-L was already an excellent limiter with its neat interface, clear controls and transparent behavior, which did not mean it couldn’t push your signal to the limits. The new features and improvements touch different areas, from metering to new limiting algorithms, and more. First, Pro-L 2 added four new limiting algorithms, bringing the number to eight:
- Modern, a new go-to algorithm
- Aggressive, best suited for heavy music and EDM
- Bus, better on drum groups or individual channels
- Safe, which is a smoother one that guarantees you won’t hear any distortion.
CPU and Oversampling
Fabfilter promises Low CPU usage, and they have added some new linear-phase oversampling options, up to 32x, that will result in a more accurate result, but an intensive CPU usage. Just to give you an idea, we tested it on Ableton Live 9.7, 16GB of Ram DDR4 and a Intel a X-SERIES i7-7800X n.12, without any track in the sequencer, 32x oversampling took 23% of the CPU, while 16x about the 10%. Not a terrifying result, especially since we could always stem our project or bounce a final mixdown and operate in a separate project. Fine and flexible improvement, the higher oversampling options now can offer even more accurate results.
Pro-L 2 now supports surround audio, including Dolby Atmos 7.0.2 and 7.1.2 formats with a handy channel link, an optional DC offset filter for the removal of super low frequencies, adjustable Look-ahead, and Attack and Release to carefully tune the algorithms.
True Peak Limiting
Pro-L 2 now features True Peak Limiting. We are probably familiar with the Nyquist theorem, sampling rate and the digitization of audio. While an analog signal has infinite amplitude variations, the digital reconstruction of the analog signal occurs with some approximations. In a few words, when we record at a sample rate of, let’s say, 44100 KHz, this means that we are measuring the amplitude of the analog signal 44100 times in a second. Imagine having a “dot” for every point of amplitude measured. Connecting them, we will draw a curve similar to the original analog signal (if the maximum frequency of the recorded signal does not exceed the half of the sampling frequency, called the “Nyquist Frequency”).
In addition, every dot needs to be “vertically” quantized to the nearest value. In the digital domain, the number of these values is not infinite as it effectively is in the analog signal. The number of values (between -1 and +1, which are the normalized amplitude’s limits of the digital domain) depends on the bit depth. 16 bits, as we find on CDs, means 2 raised to the power of 16.
Now, True Peak Limiting does not only consider the amplitude value represented by position of the dot, but also the amplitude of the reconstruction of the waveform, which can easily go over those values, and, obviously prevent those Inter Sample Peaks. In Pro-L 2 enabling this function is a matter of a click, right below the Attack and Release knobs or on the TP icon, in the lower right corner.
Another important feature is the new Loudness Metering, meeting the EBU R128, ITU-R BS.1770-4, and ATSC A/85 standards, and new metering scales such as K-System. These names can be a bit obscure, but they are really important when it comes to audio broadcast or interactive music for video games. Let’s try to figure out in simple words, what they are, although it is a fairly huge topic the explanation of which is not our primary goal. EBU R128 is a collection of recommendations and rules regarding loudness and permitted maximum level of audio signals during broadcast (for example, it suggests a maximum true peak level of -1 dBFS) that were introduced to avoid loudness inconsistencies between programs and broadcast channels, a formal regulation in the Loudness Wars. A new absolute scale was introduced: LU, LUFS or LKFS (Loudness Units, Loudness Units Full Scale or Loudness K-Weighted Full Scale, where LKFS and LUFS are terms that describe the same thing). A jump of 1LU is 1dB.
Loudness cannot be easily measured, because perceived loudness depends on frequencies (you might be familiar with equal contour curves) and can be highly subjective. Peak Normalization and Hard Limiting led to make the audio louder, without having a firm reference. The sophisticated ITU-R BS.1770 algorithm, through psychoacoustic models, introduced a new standard for measurements: the aforementioned Loudness Units, which refers to the perceived loudness. The EBU R128 suggests a -23LU loudness target, although this target may vary on different countries and applications. While this application can sound a bit far from the musician’s target, it may be useful to know that some middleware like Audiokinetic’s Wwise uses this metering scale, so this is not so uncommon after all.
For further readings:
Wwise 201 Course, page 309:
Some new functions include Audition Limiting, which let us listen to the difference between the input signal and the process signal, advanced Dithering with three different noise shaping options, and Undo and Redo and A/B comparison
The real-time level display has new display modes, such as an Infinite mode that shows all the audio flow for an entire session, or a Slow mode that writes the incoming audio at the half of the speed of the Fast Mode (Pro-L 1 had this one).
Side chain Mastering routes the external side chain to feed the path of the limiting process instead of the normal input of the plug-in. This can be very useful for stem mastering, when you need to deliver individual stems that have been processed with the exact same limiting as the original master. Use Pro-L 2 on the stem, while feeding the original (unmastered, unlimited) master to its side chain input.
The accessible and simple UI makes very easy to move around while setting the parameters. After choosing the algorithm, the advanced tab shows a few knobs with which we can govern the sound of Pro-L 2. Improving quality with the DC filter and True Peak Limiting is as easy as a single click. In addition, the clear, large display, with the full screen option gives us a perfect vision of how we are changing the incoming audio. The inline help window can clarify some of the functions of the small buttons in the lowest row and in the lower right corner: TP, DC Filter, Display Mode and metering scale.
The new algorithms sound very different from each other, and act exactly in the way their names suggest. We tried it on a drum loop, and the differences are noticeable. Bus has a strong but very clean and precise character, while we found the Modern algorithm being the more aggressive. The Safe one runs very smoothly and super-transparently on the incoming audio, perfectly suitable for raising volume and give consistency without affecting the original audio – no risks of distortion here, it works exactly as described.
The differences are noticeable also with an RMS meter inserted after Pro-L 2. We tried switching the algorithm, with the same settings, and the different character are, as expected, noticeable even in a visual way.
The same behavior is confirmed on more complex loops. Safe and Bus both work magnificently, respecting more the feel of the original audio without being aggressive (Bus especially, which is definitely my favorite algorithm), while Modern and Aggressive, again, are pushing way more than the previously two.
An already great tool, this version boasts some really useful updates that improve a variety of areas.
There is more flexibility with the new algorithms, so more choices and a wider range of tweaks, with the added look-ahead, attack and release settings, can really improve results.
True Peak Limiting and the new Loudness Meter enlarge the range of applications of the limiter. Despite that you can find some LU meters available for free (e.g., DP Meter II from Professional Audio Tools), I personally find it very useful to have everything in the same window.
Side chain mastering, new display modes and the DC filter are great new features that make Pro-L 2 even more useable, more precise and reliable.
Pro-L 2 is compatible with all major DAWs on both Windows and Mac platform, 32 and 64 bits, and, as all the FabFilter plugins, is available in VST, VST3, Audio Units, AAX Native, AudioSuite and RTAS (32-bit only) formats. List price is €169 EUR, and Pro-L 2 is also available in several FabFilter bundles.
New Features at a glance:
- Great transparent sound combined with maximum loudness.
- Eight sophisticated, carefully-tuned limiting algorithms, all with their own character and purpose.
- True peak limiting.
- Up to 32x linear-phase oversampling to minimize aliasing and inter-sample peaks.
- Low CPU usage.
- Highly accurate output and gain reduction metering, including true peak metering.
- Extensive loudness metering with support for the EBU R128, ITU-R BS.1770-4, and ATSC A/85 standards.
- Various meter scale settings, including K-System support.
- Surround support, including Dolby Atmos 7.0.2 and 7.1.2 formats with flexible surround channel linking.
- Unique, highly-informative real-time level display, showing input, output, gain change, and loudness levels over time with key gain reduction readings and four different display modes.
- Optional DC offset filter.
- External sidechain triggering for stem mastering.
- Unity Gain option to easily listen to the effect of the current limiting at unity gain level.
- Audition Limiting option to listen to the difference between the input and output signal.
- Adjustable Look-ahead, Attack, and Release settings to fine-tune the sound of the algorithms.
- Separate, freely-adjustable channel linking for both the transient and release stages.
- Professional dithering with three different noise shaping algorithms.
- Easy setup for common limiting situations with the included factory presets.
- Automatic adaption to stereo, mono, and surround channel layouts.
- Full Screen mode for precise adjustments and insightful metering.
- Multiple interface sizes: Small, Medium, and Large, as well as a separate Compact mode.
- Retina support on OS X and High DPI support on Windows.
- GPU-powered graphics acceleration.
- Supports common Pro Tools hardware control surfaces.