The nature of production (part 2)

Ok here's the second post in a series inspired by this post on DOA.

Now judgement is a hard subject to talk about as it is, by it's very nature, a subjective topic.  I guess to be able to judge effectively about your mix you need to know what sound you are going for - if you don't then this will make certain mix decisions very tricky.

Once you know what results you are after then you need to ask yourself "does this sound good?", "is this right yet?" and "if it's not right yet, what's wrong with it and how can I get it to the point that I want it to be at?"

One of my music tech tutors in college once said to me: a tune is never done, but after a certain point you need to draw a line under it.

Sometimes you need to be able to recognize when a certain synth patch or sample just isn't good enough and no amount of processing will ever get it to where you want it to be; hell, sometimes you need to recognize when a tune is always just going to be a 8/16 bar loop.

Judgement cannot be taught and you can't really write advice about judgement.  It comes with time and experience (the better I get at production the more I am convinced that your favorite producer x with the fantastic mixdowns and sound design has no production secrets; just a really good knowledge of their tools and a highly developed sense of judgement).

I don't think there's any more to say on this.

The nature of production (part 1)

I stumbled upon this post on Dogs on Acid which I found interesting and I hope you lot do as well.

Basically this post argues that production is a 3 stage process:

1. Listening

2. Judgement

3. Action

Most threads on any production forum I've seen are only about stage number 3.

So I'd like to address my thoughts to listening and in another post on judgement

Listening

One of my lecturers at uni say that the most important aspect of production is being willing to listen to the detail in audio and I 100% agree.  However certain factors can prevent or distract us from listening to the fine detail in audio.  Here are a few to be wary of:

- your monitors may be misleading you as to what you mix actually sounds like.  I produce on a pair of Edirol MA 7As (seriously, DON'T buy a pair) which only meaningfully go down to 100-150 hz, the mid range reproduction isn't all that great and the same for the highs.  

How to combat this problem: listen to professionally produced and mastered tunes at the same volume as you produce so that you can learn what a "good" mix sounds like on your monitors.  Over the 3 years I've had them I've learned their sound so my mixdowns now translate reasonably well to other monitors (even the hench genelec monitors at the uni studios).  If I was to mix them on flat monitors then I'd do a better job but at the moment I can get most of the way there with my rubbish budget monitors.  So learn your monitors, their strengths and weaknesses.

- the acoustics of your room can horribly mislead you as to what you're listening to actually sounds like.  I don't know a whole lot about acoustics, but I do know that sound reflects around your room and these reflections can cause phase cancellation on what you're listening to.  Basically without proper acoustic treatment of the room you mix in, you're not hearing an accurate representation of your mix.

How to combat this problem: get proper acoustic treatment or reference your mix on a reliable pair of headphones.  Have a look at this website, I haven't looked at it in depth, but what I've read looks good for educating yourself about the science of acoustics.

- Your ears lie to you on a daily basis.  Here's one I'm sure a lot of people who've been producing for a while are aware of: the fletcher - munson curves (also known as the equal loudness curves).  You can see the graphs here.  Basically the volume you listen to sounds at radically changes the perceived volume of bass frequencies and high frequencies (to a lesser extent), but midrange frequencies are pretty consistent in perceived volume.   This is why I hate plug ins which increase the volume of a signal (like distortion) but don't a have a master output fader so I cannot bring the volume back down to be equal with the original signal.

How to combat this: when you're a/bing the effect of some processing make sure that both the wet and dry signal are the same volume (another thing to bear in mind about a/bing is that your short term audio memory is only a few seconds long so switch between processed and unprocessed quickly to really hear what the difference is).

Also it's good to give your ears a rest during long production sessions because your ears get fatigued (especially when listening at loud volumes), you also get used to the sound that you're listening to and after a while you're not able to make objective judgements about the mix.  Apart from playing COD for a bit or actually going out and talking to people you can listen to pink noise for a bit to inject a bit of objectivity into your listening.

So there we have it, expect part 2 on judgement when I'm bored.  Also what would you guys like me to write about next: processing breaks or a rant about originality?

I'll leave you with a sick tune

Progress on uni project

For one of my modules at uni I'm attempting to make an algorithmic composition that can be controlled by your motions (using data from a web cam).

I'm in the prototype/learning how the hell to use the Gem objects in Pure Data at the moment and so far I've come up with this:

Which does this to the input of my webcam:

I still haven't got the pix_data object to work (arrrrrrggggg!), but basically what the patch does at the moment is it only shows movement (thanks to the pix_movement object) and then the green and blue channels are removed (using colourRGB) to reduce noise and make the video input easier to analyze.  Then the pix_blob object tracks the 'center of gravity' of the video and gives me data for the distance away and X and Y coordinates.

I may post updates on this project and eventually audio, maybe even the patch when I'm done.

Back to the metaphorical banging of my head against a wall that is pure data...

It's not about what you've got, it about how you use it.

This is an interesting article and I'm going to come back to it later on in this post.  This might get a bit rambling and may shoot of onto some tenuously related subject so please bear with me.


Ok I'm going to tell you how I used to make music back in the day (a good 4-5 years ago).  Back then I played guitar and had got one of these, which I used to record and create music with.  The PS-04 was pretty limited compared to say a modern day DAW and yet I found it an intuitive and easy way to make the music I wanted to within the whole 4 track limit (ok plus a really bad step sequenced drum machine and even worse bass machine).  It had some basic effects for recording and 2 effects sends plus some basic editing capabilities (you couldn't see the waveform or anything for that matter).  I loved it.


This was back in the day when I had no idea what a compressor was or proper gain structure and was dimly aware that if it said clip that might be a bad thing.  The mixes I got out of it were terrible but creatively speaking within the confines of that little multitrack I felt like there were no rules and I was free to do what I wanted.


Ok fast forward to now, you may have noticed from my earlier posts that I'm rocking Logic which is so much more powerful - unlimited audio and midi tracks (well until the cpu tells me to piss off), flexible routing of buses and all that modern day DAW blah blah blah.  You know, sometimes I find myself falling into the same pattern of doing things in different projects, almost as if I have to do things a certain way.


So first real point: your interface for making music - whether it's a guitar, violin, pair of turntables or a DAW - has influence on the music you make.  Certain workflows make certain musical results or styles of composition easier than others (I know as a guitarist how easy when writing riffs or solos or whatever how easy it is to fall into the trap of thinking in the box patterns of scales).  Arguably it's easy to fall into the pattern of "copy and paste loop then delete different parts to outline different sections" in a DAW.  My Dad once told me of a guitarist (can't remember who, I'll ask him next time I see him) who wrote all of his riffs on the saxophone so that they didn't sound like the sort of riffs that the very nature of the guitar encourages.


I think being aware of this is the very first step to combating this potential problem.


Ok now back to the link at the beginning of this post.  I find it interesting that you can exactly recreate the effect of those overpriced waves eqs (plus other analogue modeled ones) by using your own DAW's bog standard eq + x.  I remember the first time I used the waves renaissance eq thinking that "I can't really hear why this is so expensive" and I'm glad to know that I wasn't tripping.  One of my mates waxes lyrical about how fantastic the waves eqs are and how much better than the logic ones they are and over the last year and a half I could never hear it.


Compressors are a different issue and have too many variables to be recreated by a flexible one, but i guess what I'm trying to say is that it doesn't really matter what gear you use to make music or do mixdowns, it's all about how you use it.  It doesn't matter if you've got X plugin that's hyped on gearsltuz it doesn't automatically mean that your mixdowns will become any better.  Same with DAWs, if you're making real tight mixdowns on Fl studio, getting pro tools will not improve them.  Going back to what I was saying about how I used to make music, now I know what I'm doing I could make a mix that I'd be proud of using that PS-04 despite it's limitations.


A rookie guitarist with a proper Les Paul who can't properly do bar chords is going to sound worse than a really good guitarist with a squire strat.


Incoherent rant over.

DIY Dub Delays

Here's a way that you can make your own dub delay.

1. Send a track to a new bus

2. Add a simple delay to the bus, choose your delay times to taste and make sure that no filters are doing anything to the audio.  For this to work you want to keep the feedback between 0% and 10% - we'll be adding our own feedback loop later on.

3. Now add a distortion effect after the delay and a filter after that (I like either high passing or low passing but you could even go for band passing if you want).

4. Now to create the feedback loop, send your dub delay bus to itself (so if I'm using bus 1 for the dub delay, I send bus 1 to bus 1).  I'd recommend putting a limiter on either the master or at the end of the dub delay channel to avoid damaging your speakers and your ears (my mate blew his by turning up the feedback on Logic's tape delay a little bit too much.

That's the basics of it, but you can get creative with this technique and add other kinds of effects to the delay chain.

Here's a couple of Audio Examples of me messing about with this technique.

A different way to do rising build ups.

Ok here's a different method of doing rising builds ups that I nicked from a mate of mine.

First things first you select either the first beat or the first two beats of the section you want to build up into and bounce down a copy of this (you just need one or two beats - the picture below I just went for one).

Ok, import the bit of audio into a blank audio track and add a nice long reverb (try starting with 100% wet, sometimes this trick works better with a little bit of dry tucked in there are well), now bounce a copy of this snippet of the drop with the long reverb out.  In this picture I used a nice long reverb from Logic's Space Designer, but any nice nice sounding reverb will do the trick.

Now open up this new copy with reverb up in your sample editor of choice and reverse it (I'm using the one built into Logic), so it looks something like this.

Right now we're ready to position it so that it builds up into the drop (I had to chop a little bit off the front of my reverse reverb build up because it tailed off a bit) leaving you with something like this

I like this technique because it doesn't sound quite as rinsed as the band passed white noise sweep.  You can always experiment with further processing the build up (maybe a bit of subtle flanging or some other stereo modulation effect).

Edit: Here is an Audio Example of the reverse reverb build up

Hello World!

I'm making this blog to talk about my favorite music, production techniques, coding for Pure Data, maybe share my musing on music as a whole (I'll try not to talk too much rubbish when I do this) and also to share about my music and what I'm up to.

I will leave you all now with a fantastic tune, "My Angel Rocks Back and Forth" - by Fourtet