Mixing By Numbers (Part 2)

31 03 2011

Yesterday I ranted about the dumbing down of mixing. Today I hope to provide a more balanced view on the new One Knob series from Waves.

Whilst writing my post yesterday I had the idea to do a One Knob mix to see how versatile these plug ins can be.  The One Knob mix took about 35 minutes to do.  Here’s a quick outline of what was used where and roughly how much the dial was turned to in brackets.

Overheads – Pressure (3)
Kick D112 – Phatter (7), Pressure (5)
Kick NS10 – Pressure (6)
Snare Top – Phatter (6), Pressure (6), Wetter (3)
Snare Under – Pressure (3)
2x Toms – Phatter (6), Pressure (6)
Drum Bus – Pressure (2), Wetter (2)
Bass – Phatter (3)
2x Acoustics – Brighter (7), Wetter (1)
2x Crunch Guitars – Wetter (1)
L Dist. Guitar – Brighter (8), Wetter (1)
R Dist. Guitar – Wetter (1)
Solo Guitar – Driver (4), Wetter (2.5)
Lead Vocal – Driver (1.5), Brighter (7), Pressure (4), Wetter (3)
Backing Vocal – Driver (1.5), Filter (8), Pressure (3) Wetter (3)
Master Bus – Louder (4)

Please don’t go judging my general working practices on this…

So as you can see I used all the plug ins and tried to do what I would normally do in a mix.  The first weakness I’d like to address is that Wetter is designed to work as an Insert, not as a send, meaning I had to drop loads in on the session.  By dialling up Wetter you’re not only increasing the mix of the reverb but the size and length of it too.  So in my mix, where I have several different reverb settings it has become a mess.

Pressure is much better used as a bus compressor, it really didn’t work on the close microphones on the drums. At lower settings the compression is quite subtle but it soon gets out of hand and when I dialed the master bus Pressure up to 10 something very strange happened… I actually felt physically sick! It was the same kind of weird sickness I get when I hear serious phase problems.  I had to dial it down pronto.

Phatter makes things bassier… not much to say there. I think it might do some low end compression too. Not sure so I won’t guarantee that.

Brighter is a really nice clean and transparent high boost. But I recommend getting a good parametric EQ and using that.

Filter is really for use on Dance tracks. Like Brighter and Phatter it does sound nice, but it does nothing that many other quality filters and EQs do already.  I’d also like to point out that Filter has a resonance selection button! I wanted One Knob, not One Knob And A Button.

Driver. This is the one I liked the most. The very low settings worked nicely as a lo-fi effect.  Higher gains were pretty good a boost to guitars but it’s not good enough to be the whole guitar tone.

Louder is the ultimate nemesis of Dynamic Range Day, Turn It Up.org and friends.  It’s a limiter that sounded like it was limiting very quickly and didn’t actually stop my master output from clipping anyway… so… why?

So in conclusion; I don’t get it. I don’t think you can deny the quality behind each one of these little things, individually they sound good. But with multiple instances of these plug ins they seem to clash with themselves, especially Wetter. So as a one off effect for one channel in your mix, a couple of these plug ins would work well. Namely Driver and Filter.  The EQ two Phatter and Brighter are really nice, but if you own any good parametric EQ plug ins already they’re pointless. I would seriously stay away from the compression ones and I can’t see the use in a one-trick reverb.

By Graeme Rawson

Mixing By Numbers

30 03 2011

As a semi-professional sound engineer I hope that people will see my work as something worth paying for.  I need the income I make from my recording business and I love doing it.  Over the past few years more and more people are cracking Cubase and hundreds of plug ins and joining the industry as hobbiests or “bedroom producers”.  I don’t want to sound like I’m moaning about the development of the subject, as a teacher of Music Technology I’m glad there is increasing interest. But I can’t help but think there is a certain amount of dumbing down going on. New plug-ins like the Waves One Knob series are removing skill and therefore value out of the industry.

On the base level of the industry, where I currently ply my trade, artists are usually funding recordings out of their own  pocket and are seemingly more reluctant than ever to pay for quality.  The studio I work in prices a day’s recording competitively with other similar level studios in the area, this figure is a good chunk under £200 a day.  The same price (I’m reliable informed by the owner) that it was around 20 years ago! Showing no signs of inflation and not representing the technical improvements of that last two decades.  This reluctance to pay, I believe, is coming from the view that a good record can be done on the cheap using software, however it was obtained, in a bedroom. I’ve heard some passable demos come out of this environment, but never a marketable professional recording. What’s missing is not the quality of software but quality hardware and crucially the touch of an experienced and skilled ear.

This ear is easily purchased but for some reason it’s not seen as a skill worth paying for. You wouldn’t find a plumber willing to work 10+ hour days for barely minimum wage, but in our industry if someone want to make in into the higher echelons of professional recording they must work almost endlessly for nothing at all.  Is this because our trade is viewed as a luxury to do? I love being a sound engineer and therefore am willing to work for perhaps less. But I could not afford the time working unpaid to make a move into the truly professional sector, so I made the choice to work, paid such as it is, at the entry level.

So I’m hoping to make a living using my skills as a mix engineer, but is this advantage being taken away slowly two? Can you now make good mixes without any understanding of the science behind it?

I’ve not tried, but it’d be interesting to see how good I can get a mix by just using preset settings on plug-ins. This would be easy enough and I’m fairly sure it’d sound okay too, but if everyone did this, wouldn’t we get a lot of records sounding the same? We’re already hearing Apple loops in charts, I cite Usher – Love In This Club and Rhianna – Umbrella.  So now, thanks to Waves we can all sound the same in new, uncharted ways. If we all just used One Knob plugs we’d be  able to do everything a mix engineer could possibly want… A dial for more reverb please, just turn up the compression to “on more”.  Tweak the bass up to a phatness level of 8.  Make it louder despite clipping every bus in sight with the One Knob Louder plug.  And do you know what’s most annoying? They’ll probably sound great.

I’ve downloaded the One Knob demos and will soon provide you with a glimpse of the future… the One Knob mix!  I promise to do my best with it, as much as I might want it to sound bad. Stay tuned…

By Graeme Rawson

AI: Autotune Intelligence

17 03 2011

Autotune has been around for donkey’s years. Every now and then it stirs up into mainstream media and everyone seems disapointed for 5 minutes that it’s been done to their favourite pop star or Glee actor. Then it’s all forgotten again the next day.  I would say, although I have no proof really, that every single pop/rock track released in recent times has some sort of pitch correction on vocals and the majority use Antares Autotune.  It’s not really a bad thing when used right, I have no objections to Autotune being used to help good singers get it perfect. Often it’s the tone, feel and emotion that makes a singer unique or marketable, not his/her ability to pitch 100% perfectly.

My issue is that slowly, over time, we are conditioning the listening public to not notice the Autotune and therefore we are getting braver and making it more obvious and it goes around in cycles.  Hell, we’re even conditioning ourselves to ignore it.  In casual conversation with various people (as I’m inevitably asked about it every time it’s in the news) I have found that many people now believe the artifacts caused by obvious Autotune as just part of some peoples singing voice!  “Doesn’t he just sound like that?” when referring to T Pain is one that springs to mind.

There are two main artifacts to look out for when using Autotune. The first and most obvious one is the clicking or blipping sound made famous by Cher and often now used as an effect. Heard here:

The second one is subtler and in someways is more guilty when it comes to duping the public. It’s a synth like tone that becomes noticable on longer notes as it seems to take over the natural voice. Heard here on the “yeah”:

I am going to discuss a couple of feature of Antares Autotune and how you can use it invisibly to improve your vocalists performance without robotizing it!

Retune Speed

This dial is the culprit of the blipping artifact when turned up too fast. The fear that some engineers may have is that if it’s too slow the Autotune won’t have time to tune the shorter notes. Although this is true to some extent it is often enough to just have a note moving towards the right pitch and never quite getting there rather than having it switch quickly and causing the jump blip.  My advice is to er on the side of slower.

Know your Key and Input Type.

I’ve seen these two really useful features overlooked to bad effect.  Autotune defaults the key to a chromatic scale, meaning it will retune the voice to the nearest semi-tone. This is fine if the singer is always within 25 cents (quarter tone) of the correct pitch. If they slip over 25 cents towards the next semitone then the odds are Autotune will retune the vocal to a note that is not in the correct key and it’ll sound awful. By selecting the correct key you’ll reduce Autotunes options when it’s selecting the retune note. If your singer is still over halfway towards the next note (likely 50 cents) in the scale then he’s not very good… but you’ll have to get in there and edit in in the graphical mode.  At least you know that every note Autotune thinks he’s trying to sing will be in key and sound superficially fine.

Input type is even easier to overlook (it’s in the top left corner, by the way). By selecting the right input you change the algorithm that Autotune uses to analyse the input and will greatly improve it’s chance of retuning naturally and accurately. Incidentally; notice the bass instrument option? Autotune on bass is a great idea for tightening up that rhythm section.

Here I’d also like to mention the “Targeting Ignores Vibrato” button, I nearly always click this on.  As it allows the singer some tolerance in performance. The natural wobble of notes can matter a lot to the feel of a song, particularly an emotive song.

Create Vibrato

This is designed to remove the droning tone I spoke about earlier on longer notes by creating a vibrato after a certain length of time.  Having a delay before the vibrato starts helps it feel more natural as it won’t vibrato quick notes and will come in after the longer notes have settled in.  But there is a massive problem: I once heard a Rhianna track (I tried to find it for an example but she’s released thousands over the last 3 years!) that had this very feature enabled. It wasn’t an obviously Autotuned song, but the vibrato came in at exactly the same point note after note after note, gave the game away… which leads me onto my next point.


This is where you can really get clever to make your Autotune invisible and your singer great! By automating the onset delay, shape and amount of vibrato the engineer can make it feel much more natural as it’s constantly changing every note. It doesn’t take long, and you don’t even have to be accurate, just change the settings around as it plays through.  This isn’t where automation stops helping Autotune though, try changing the retune speed to suit the lengths of the notes currently being sung.  A faster speed can be used throughout a sequence of shorter faster notes and then slow the speed down for the big long ones in the chorus.  The posiblities are endless, but there’s one more thing you should think about automating.  The bypass.  Sometimes I have found sections of vocals where no matter what else I do the Autotune can be heard, so I’ve simply excepted the natural sound on that sequence of notes and automated it back on later.