With the many different versions ”cheap China condensers” out there, and more being created every day, there has probably never been a better time to to sell de-essers than now. XILS-Lab, perhapsss mossstly known for their exsssellent virtual analog ssynthesssiszersss, have desssided to put an end to excesssive esssess.
De-essing is a wide subject. Some never touch a de-esser and others swear by them. De-essing starts with the selected microphone, the voice in question, placement of the microphone and the vocal technique of the singer. Put the ”wrong” type of voice in front of the ”wrong” microphone and you will end up with unusable vocal takes, however much de-essing you apply in post. Most of the time you will probably get something usable but need some type of taming of the highs, enter the de-esser.
Myself I am somewhat in the middle, I have no problems with using a de-esser in my DAW but I’m quite fond of my hardware de-esser so I usually tame it a bit on the way in already.
Most people agree that the best way to get rid of excessive esses is to automate the volume of the track. True. That is the best way as it gives the most natural results. But it takes time, and skill. A de-esser is created to take care of this automation automatically. In it’s most basic form it’s a compressor with a narrow band reacting to the higher frequencies only. There are a lot of de-essers out there already, some good but a whole bunch not so good.
The biggest difference (in my opinion) is in how the de-esser reacts to the incoming signals. Some react only at a certain level, you set a threshold, and no de-essing takes place if the sounds are softer than the set threshold. In theory that works but in practise it just doesn’t work that well. It gives an unnatural sound.
Enter XILS DeeS
Installation was very straightforward, install the plugin, authorise it and go. When I first loaded DeeS I was a bit confused and actually had to read a few lines in the manual. And now comes my biggest gripe with this plug-in. It is not created after any type of hardware, it’s not a virtual model of anything that already exists in hardware. Why on earth would anyone make an interface that doesn’t follow the flow of how you set it up? Beats me. But then again, it could be just me…
According to the manual (and somewhat regular de-essing workflow), the steps are as follows:
- Adjust the detection of the sibilants
- Set the level of the sibilants
- Adjust with some EQ to taste
The DeeS features a lot of controls, quite a few more than your regular de-esser. More controls means more control of course but also a need of understanding of those controls. The manual gives a somewhat good description of what each knob does but it’s when you have used it for a few minutes that you get how it all functions.
Lets start with the detection. DeeS has a control which lets you filter out plosives from the detection unit. This is a first for me and I must say that I like it. Quite often the detection unit in de-essers gets fooled by plosives such as hard ”ph”, ”ff” type of sounds. If you de-ess those sounds it just sounds strange and sometimes real bad. DeeS let’s you sculpt away those hard sounds and let the softer ”real” problems through. The ”essssessss”.
Depending on your vocalist the offending frequencies are in a certain region. The detection frequency knob lets you home in on them in detail. What the detection unit really is trying to do is separating the sound into two different ”parts” if you so will, one with the main vocals and one with the sibilants. The reduction knob is a kind of mix knob between the voiced part and the sibilants.
Lastly there is a meter showing the sibilant reduction, in percent (don’t know why they call it a VU-meter) and a quality knob which in reality is how fast the transition between voice and sibilant really is. You set it to do the transition between the voiced part and the sibilant part slower or faster, in time not quality.
You also get two different EQ’s, one for each part of the divided sound. You therefore have the means of adjusting only the sibilants separately which is great. Also, you could boost the voiced part a bit where the sibilants are removed and get a somewhat more natural de-essing. As the final control in the sculpting part we have a level knob which controls the level of the sibilants.
Finally, at the bottom we have a big display of the vocals frequency spectrum and the different EQ’s are drawn in their respective color as well as the detection area which is a tad lighter grey in color. Here we also have an input and an output knob. Both are used to compensate the audio coming in or out of the plugin.
There are also three different solo buttons in various places in the interface and they let you listen to the detection, sibilants or the voiced part in solo.
What I like about the DeeS is the amount of control it gives me. Not just in the detection part but letting me have control over the voice and the sibilants separately with an EQ on each is really great. You can get excellent results with DeeS but of course you can go overboard and get that lifping found af well. What I don’t like is the order of the controls. I’d much rather have the detection unit first, followed by the sibilant and then the voiced part. But it’s just a matter of learning I guess and I’m not that old, not yet. But as it is now it feels counter-intuitive
I tried both some female vocals and some male vocals as well as some voice over and DeeS handled the task of controlling the sibilants very well. What I haven’t had the time to test for this review is how it handles sharp transients like cymbals and fret squeaks. Most of the time though you will use it on vocals. And there it does its job just fine.
In this day and age when ”everything” is available digitally and has demo-versions it is important to get the user to connect with the plugin as soon as they lay their eyes upon the GUI. I don’t think that there are that many that reads a manual of how to use a plugin these days, especially if you just are trying something out. And that is a shame because the DeeS is capable of some outstanding results, if you figure out how to use it. Yes, there are a lot of knobs but everyone is there because of a reason, they are just put in the wrong place, from a workflow point of view.
That said I must congratulate XILS for getting version 1.0 this good. Yes there is room for improvement for future releases but not when it comes to the actual de-essing. I think they nailed that part pretty sssolid!
Note: ”cheap China condenser” is a term used for the many ”no-name” large format condensers that you can get dirt-cheap by the dozens. They all share the phenomenon of exaggerating the higher frequencies and the need of de-essing is therefore greater with these type of mics.
Minimum system requirements: A PC or Mac with at least 1GHz processor and 1GB RAM.
Plug-in formats: AAX, AU, RTAS and VST.
More info: http://www.xils-lab.com