A zoom on the recording process

[RainbowStorm]

Something that I think is pretty underestimated in recording is the way each recording process component is "consumed". By consumed I mean actively making one or many decisions within the process component. By process component I mean for instance tracking, optimizing, balancing, fine adjusting. When I open a mixing book it's pretty much focused on coloring and balancing, which in a very professional context might be the most important thing for a mixing engineer, but for recording engineers it kind of results in a pretty unbalanced focus overall, resulting in bad overall sound quality. So the weight of the consumption of each process component causes an unbalanced overall recording process. Sound quality is actually not very focused on the mixing process, the mixing process is much closer to the music quality in my opinion. So such an unbalance can easily make a record much worse sounding than it needs to be.

During the last two years I've focused a lot on the mixing part, just to find out there are other components as well that are as important as the mixing part. We all know that the tracking phase is extremely important, since it includes the choice of instrument, room, mics, mic position, pre-amp, effects and so on. But something that I have constantly consumed very inefficiently is the optimization process component, including early coloring. By that I mean making each individual track and individual element sound the best possible once it's tracked. In recording books they mention "pressing the solo button" and describe it a little, typically in a few sentences. But in my opinion this phase is much more important and complicated than that. First of all, if we don't pay attention to this process we will easily add effects in the wrong order, simply because the decisions are taken as we are mixing rather than before that. A typical example is noise reduction somewhere down the road after reverb, another example might be applying a stereo effect post fader after some delay effect and losing all the possibilities with being able to efficiently pan an already very wide track by applying the stereo effect pre fader. Actually I noticed the importance of this component as I happened to change the order of panning. I had always been panning all my tracks after balancing them with the volume faders, but panning them first (with the tracks temporarily normalized) made a quite big impact on the final result. I realised this can really be a late time optimization step, simply finding their right default places in the stereo field before mixing takes place.

Another thing I realised the optimization process component could include was why not make early decisions about how wide each instrument should be to sound the best, including applying widening plugins, time offsets on tracks, adding delay on some channel when necessary... We all know that mixing with phase can cause phase problems. I think one of the reasons is that it's pretty hard to widen a track once the song is being mixed. Another way to view it is that everything that is applied after the volume faders and pan knobs have been set will automatically damage the air between the tracks because it will ruin the underlying air structure. For instance adding strings really widely late time in the mixing process is pretty hard, because the sound field is already consumed in a certain way, you can't just add the strings on top of it all and widen them because either you get phase problems, panning problems or you can't set the volume balanced enough. But when each track and element has a pre-defined stereo width and position the final result will be better when the consumption of the balancing component is finished. Of course it's allowed to modify it all later in the mixing process, but the point is that it's easier to balance a song if as much as possible of the final result is included once the balancing takes place. One such thing is applying compression on drums. Typically you fail with that because the volume faders are set on "holy" state once you realise you need to add softness on the kick drum for instance when the song is already balanced. The instinctive procedure is to be very careful with the volume faders, when in fact the best is to dramatically lower the volume fader and dramatically increase the compression output on that track to not make the other inserts on that channel unbalanced and to add the amount of softness that is really needed to create the desired effect. All this is much better heard when the tracks are being soloed in the tracking phase, hence leading to better decisions through making them in the correct moment and therefore not having to do so dramatical things once the song has been balanced. The optimization step can also include a lot of other important decisions that might have to be made, for instance compressing, EQing, adding room simulation and so on.

I view mixing as an iterative process, but optimizing individual tracks and elements before the mixing takes place can dramatically reduce the amount of decisions that have to be made in the mixing phase and more important reduce the amount of decisions that were never taken. This is only one important block in recording that you easily bypass too fast, resulting in a worse sounding mix.

There are other views on this as well, some engineers think it all should be mixed together anyway so one should better get used to it as soon as possible and just start mixing with the tracks unmuted/unsoloed. I think that's an efficient way of not caring to do any decisions or not caring about the importance of making decisions as soon as you are ready or once you have a really good opportunity to do so. It's like starting to mix from a random point, which is extremely inefficient in a song where there is an underlying importance order of each element in the structure. (a decision should always be based on the priority order of the options)

One more thing. I think it's easier to succeed with a mix by taking away things rather than adding things, which is why it might really help focusing much on the optimization phase early. Decisions are better made when you can do small decisions towards a greater decision, which can be done by removing stuff once you've got the information you need to know what you have to decide. One thing I think can really help in this context is to have a clear picture of what the mix should sound like in the end, before you start mixing, rather than the other way around. Focusing on the optimization step is a good way of really process that in practise too. But once you get new ideas from the mix they are worth paying attention to. I think that's the optimum approach.



BTW, of course this is only my view on this. There's nothing worse than throwing away steps based on decisions that have proved being correct earlier...

[3rd world order]

just out of curiousity.. how many cups of coffee do you consume each day? Great nick, BTW.

[Matthew Murray]

You sure do think about stuff a lot, Tony.

[RainbowStorm]

Originally Posted by 3rd world order

just out of curiousity.. how many cups of coffee do you consume each day? Great nick, BTW.

Hehe, none actually, but today I drank one cup earlier...

[RainbowStorm]

Originally Posted by Matthew Murray

You sure do think about stuff a lot, Tony.

Yeah I like to do it that way instead of just doing and suffer, but of course it depends, sometimes taking risks is the right option. But I've learned my lessons just like you will or have and I will continue to learn my lessons just like you will... No, but seriously, there comes a point when you need to dig deeper in order to set the limit higher. This is what every architect constantly is facing. So for me writing an article like this is not very expensive in terms of the energy consumption, I'm used to digging this deeply since I'm a software engineer as the profession.

BTW, any comments on what I actually wrote in the article? I would be glad to receive new perspectives to learn more...

[stevep]

I think you think to much....... Just push the faders UP and let it Rip

[stevep]

Oh yea,,,, If the meters dont Slam all the way to the right at the first beat and stay that way till the end its not a good mix.......



If you cant stand looking at them pegged and all the overload lights on all the time ……..... just put tape over them

[guid0]

I think there's an inverse relation between the quality of the arrangement/performance and the time you have to spend "optimizing" the tracks for a mix.

If everything sits right during the actual recorded performance, it will sit right in the final mix.

If there's no performance to speak of as in the case of recording track by track, or recording a bunch of talentless fruitbats, then I agree you have to spend some time "optimizing" to create your mix. Wether this means soloing/tweaking every track or putting up a decent mono mix of the recorded program before going stereo all depends on one's personal production values.

In the end, you have to make it sound good one way or another. Put them faders up and use your ears!

[RainbowStorm]

Originally Posted by guid0

I think there's an inverse relation between the quality of the arrangement/performance and the time you have to spend "optimizing" the tracks for a mix.

If everything sits right during the actual recorded performance, it will sit right in the final mix.

If there's no performance to speak of as in the case of recording track by track, or recording a bunch of talentless fruitbats, then I agree you have to spend some time "optimizing" to create your mix. Wether this means soloing/tweaking every track or putting up a decent mono mix of the recorded program before going stereo all depends on one's personal production values.

In the end, you have to make it sound good one way or another. Put them faders up and use your ears!

New perspectives, thanks a lot! So you think the optimization process should be skipped and instead the tracking phase should make everything right from the start. I have read about this approach before, I guess this is a pretty common view on the process. It's a pretty well known fact that as much as possible should be as correct as possible as early as possible, so for instance instead of using EQs in the mixing process one should choose microphone, microphone position and preamp instead. This is because it results in a less processed signal path and a better kept signal-to-noise ratio and better mixing material. I agree to a certain degree about your opinion, in that of the two processes tracking is the more important process in my opinion, because there are so many more important decisions made in the tracking phase and the sum of them makes the whole process more important. Is tracking enough? Not necessarily I think. Some things should be kept controllable and simply can't be made right in the tracking phase. One such thing is when the tracking phase doesn't include optimization of elements (a group of tracks) and element groups (a group of elements, most often the whole arrangement). As we all know tracks within the same element easily cause frequency fighting and elements within a group can cause frequency fighting as well. Depending on what the tracking phase looks like (many instruments are recorded simoultaneously or one instrument is tracked at a time) it might be pretty hard in practise to get everything right in the tracking phase, especially if the song is mixed at a different studio than it is tracked at. Sometimes different things like for instance compression has to be kept moderately in order to be controllable for the mixing engineer, a situation when frequency fighting of tracks within elements and elements within a group might be impossible to prevent completely as early as in the tracking phase. In these kinds of situations a lot of those important decisions are transfered into the mixing process, which easily makes the decisions never made since so many other decisions have a much higher priority in that process or the decisions are made in the wrong order without knowing it.

[guid0]

Originally Posted by TonyCrazyMan

Depending on what the tracking phase looks like (many instruments are recorded simoultaneously or one instrument is tracked at a time) it might be pretty hard in practise to get everything right in the tracking phase

Ay! That's the rub.

Putting more effort into "getting things right" at the onset usually translates into a more satisfying result.