When you use 2 limiters, you split the processing between the two, in turn mildly reducing the effect of distortion while achieving a louder overall level. Furthermore, if the 2 limiters you use have a unique or complex timbre, combining these can have a beneficial effect.
Using 2 limiters gives you the freedom and opportunity to combine settings. For example, you may make your first limiter compress each channel independently, then have the second one compress the channels collectively.
Although this sounds a little counterintuitive, this would retain the detail of each channel with the first limiter, and then achieve an even loud sound with the second which still keeps everything relatively distinct.
The shorter the release time that you use for your compression, the louder the signal will be - the reason being, it takes less time for the signal’s amplitude to return to normal after being compressed. A super short release time will cause distortion, but this can have a beneficial effect.
This distortion will amplify the signal, causing it to sound even louder, but this may not be ideal. A good mid-ground for your release time is 50ms, since this is the shortest you can make it without causing distortion.
Although this is true in most cases, each compressor is a little different, so use your ears and see how short you can make your release time before you begin to notice distortion.
Using two limiters allows you to make your signal louder without relying too heavily on one form of processing - but if you add an expander between these limiters, you gain back some dynamics that you lost from the initial limiter. Although unorthodox, this is great for making loud masters.
The first limiter is going to reduce the majority of the track’s dynamics. By the time the signal hits the second limiter, the majority of the transients have already been attenuated.
By putting an expander after this first limiter, you build back your dynamics.
Personally, I like to use the FabFilter Multi-band for this, and expand some of the lows, mids and highs.
Upward maximization or low-level compression serve a very unique purpose during mastering - they bring quieter details of the mix upward, making them easier to perceive and reduce the effect of masking. If you want to create a loud and upfront master, this form of processing is a must.
Whereas traditional compressors and limiters attenuate the loudest parts of the signal, upward maximizers and low-level compressors detect aspects of the signal which are quiet, captures them, and amplifies them while keeping the peaks at the same level.
Combine this effect with your limiting to achieve a full, complex, and loud master without having to push your limiter too hard.
Whereas the lows of a master hold the majority of the track’s dynamics, the mids hold the majority of the instrumentation and detail. You can augment this section of your master’s frequency response by using a traditional stereo compressor and isolating the mids with an internal side-chain.
Then, use automatic makeup gain to ensure that the level of the mids stays the same. By compressing the mids and then keeping them at relatively the same level, you make them louder in comparison.
Although internal side-chaining doesn’t mean that only the mids will be compressed, it ensures that compression only happens when the mids are loud enough to trigger the compressor.
Some limiters include various algorithms or styles that make them better suited for specific forms of mastering - if you’re trying to create a loud master, find a limiter that suits loud mastering. A Dynamic Limiter typically expands the transients before limiting, meaning you can push the signal while retaining dynamics.
A good example of this is the FabFilter L2 - its dynamic option expands the signal’s dynamics right before the limiter kicks in - making for a louder, more dynamic sounding master.
You may also want to try the loud algorithm since this too is suited for creating a louder sound without causing too much distortion.
Although over-processing a master is definitely a problem, the fear of using too much processing can hinder you from making a commercially competitive one. If you’re working on rap or pop music, you may need to use more processing than you expect to achieve a loud sound.
If you’re mastering a song, and you know it needs to be loud enough to compete or sound similar to other masters, be sure to use a fair amount of both upward compression and limiting.
One processor that I’ve found which is really helpful for this is One by IK Multimedia. If you include it in your signal chain prior to limiting, it provides just enough processing to push the signal into a commercially loud and punchy area.