One of my favourite musician jokes is the one about the travelling businessman who goes into the hotel lobby bar for a drink. After the piano player plays a few songs, he goes up, puts a crisp fresh $100 bill in the tip jar and says to the pianist [with a thick NY accent] “I’m from New York and with all this travelling, I’m feeling a bit homesick right now, do me a favour and play the authentic version of New York, New York”. The pianist had been able to spot the unusually generous tip in the jar and emphatically says “sure thing” and starts to play “dum dum dadadum, dum dum dadadum, Start spreading the news…” The businessman, with his hands waving for him to stop, blurts out “hey, I asked you to play the authentic version of New York, New York”.
The pianist glances to the bartender in a bit of a state of panic, pauses for a second and then carefully counts off and starts again with more intensity; “dum dum dadadum, dum dum dadadum, Start spreading the news!… ” Once again, the businessman waves his hands for him to stop and says “are you messing with me? I’m born and bred in New York and I nicely asked you to play the authentic version of New York, New York, …you’re not even playing in 5/4”. The pianist looks at him and says “I am sorry sir but that is the only way I play the song, ...in 4/4”. The businessman replies “. I played a little in the day, slide over and I’ll play a few bars for you”. The pianist hesitates as he does not usually allow patrons to play but glancing again at the tip jar, he motions to invite the businessman to sit down right next to him.
The businessman adjusts the microphone, counts to 5 and plays; “dum dum dum dadadum, dum dum dum dadadum, Start F$^&%ing spreading the news…”
So why did I pull that one out? Well, we see lots of articles on compression techniques and feature lists for both hardware and software refer to a blend mode or a control to mix the compressed signal with the dry signal as parallel or “New York” compression. The idea of parallel compression is that a blend of the sustain of the compressed signal and the attack of the dry signal combine in manner to give a result that is the best attributes of both signals. Prior to boxes with a dedicated parallel function, the technique to do this was to “mult” (split) the signals, and then mix them back together again on separate console strips. It is widely accepted that this technique was first used in one or more reknown studios in New York. The problem is that it isn’t authentic New York compression, there’s a missing detail that makes all the difference.
Parallel compression is fantastic, both on mixes or a subgroup. Now with Logic Pro X’s new compressor, one can quickly learn how it sounds in different applications. The plug-in’s 7 modes all have a mix control with even some presets that use this parameter by default so I’d invite you to experiment and learn how it sounds and interacts.
Authentic New York compression works a little differently while still being a parallel compression process. The key difference is that the compressor is detecting a sum of the left and right channels and most essentially, that the compressed signal output is 100% mono and centered. The compression applied tends to have a setting that is such that there are considerable artifacts such as pumping, way more than one would normally use. When re-combined with the dry signal, this allows a masking effect where the initial attack provides a fully stereo spatial image and the compression component focuses the signal’s energy straight down the middle without the listener hearing the compression artifacts. This is one of the magical audio sleight of hand techniques that can really make a mix come together. It’s actually a great way to use the so-called “all button” or “British mode” on an 1176 style processor. Our friends from the UK will be relieved to know that I’ll refrain from another joke at this time.
I really like to use this technique when mixing guitar parts that are not super distorted and grooving a bit… think classic rock in the Stones and Aerosmith vein. Often these types of layered parts aren’t perfectly tight and likely aren’t even exactly the same chord voicings with occasional inflections that are different from each other in places. Panning these parts wide makes each part distracting during the “rough spots” while panning them together makes what I call “brown gravy”… it tastes ok and glues everything together but in reality, just makes a mess of the mix as soon as it gets brought up to be prominent. Authentic NY compression will magically make the parts work together and take their rightful place in the mix. It can also be equally effective with drum busses, room mics, synth layers and back vocals so let’s see how this is done in Logic Pro X. I would recommend trying this out, I think you’ll like it and quickly recognize some albums from your collection where this technique was an essential element in the mix.
The easiest way to set this up in Logic Pro X is to create an aux for the compressor and bus sends (ideally pre fader) to that aux on the individual channels of each of the elements to be processed. Each element is panned as desired with a wide overall field and sent as well to the compressor bus. The aux channel strip hosting the compressor plug-in should be toggled to mono for NYC compression or stereo (default) for parallel compression. A suitable compressor is instantiated on that aux channel strip. Alternately, on the aux channel strip, one could open a Gain plug-in instance (Logic Pro X utility/Gain) followed again by the compressor of your choice. The Gain plug-in is set to unity but with the mono switch engaged. The compressor is set to a fairly high ratio and a threshold so that there is considerable gain reduction. The aux return fader is then used to blend in the compression effect back into the mix. As this is an additive process, please assure your gain staging isn’t overdriving your master bus.
It is very important that the sample accuracy is retained between the dry and wet signals so make sure plug-in compensation is engaged and that the plug in you are using is correctly reporting its latency to Logic Pro X. If you are not sure, use one of Logic Pro X’s native compressors, they are excellent and you’ll avoid a potential problem. For those who are brave, if you encounter an issue with a preferred third party compressor, a fine resolution sub sample delay plug-in such as the Eventide Precision Time Align will be a great asset. If your result isn’t massive or sounds funny, you have a phase issue to correct.
Here are some audio examples for you to listen do and hopefully gain the inspiration to try this technique yourself. My colleague here on Logic Pro Expert, Eli Krantzberg graciously supplied us with an unprocessed stereo file of a live drum kit. We’ll hear it first as is, the compressor buss in both mono and stereo, then with parallel compression, and finally with authentic NY compression. I then recorded some live guitars and we’ll hear them fully panned, then centered and finally with parallel and authentic NY compression applied on the full width tracks. In all cases, I went with sound settings that best demonstrate the core of this discussion. I might choose to process an actual mix with a bit more grace and a more delicate stroke…. or not.
I hope this neat trick helps make your next mixes stand out from the bridge and tunnel crowd, no offense!