I've been watching the VOD of Deadmau5 producing his Imaginary Friends track over the last couple of days to get inspiration from it and see if I come across certain techniques that I didn't know or thought of just yet.

The biggest question I have thus far is why would he be using audio tracks for his bussing needs. Thus far he didn't create a single return track for bussing things together and has been using audio tracks set to "in" (he uses Ableton for this track) for these needs. I understand there are a thousand different ways all leading to the same place but I am wondering what the pros and cons could be from using these different methods.

Personally, when I want to bus certain things I just create return tracks and use my sends to achieve this. It also seems a lot more handy because I can adjust the send levels for the different audio and MIDI tracks being sent to the bus. Am I overlooking certain possibilities or limitations from using either of these methods? Are there any real differences between the two in terms of either sound or possibilities?

 

Bussing for what? For fx (reverb/delay) or for grouping (all drums, all synth)? Can you provide link for video?

 

Most of the FX are on the separate instrument channels, so I think he has just been bussing things like his synths using this method for the sidechain and the likes. I've been thinking a bit about it and I think he has been using these input audio channels as an alternative to grouping the different channels inside Ableton. Correct me if I'm wrong, but sending several instrument channels to the same audio input channel makes it so that all of their sound only comes out through that channel. That would make this particular method great for making sure things go through the same sidechain compressor or something similar but makes it unable to be used for parallel processing.

This is the link:

He creates one of these audio input bus channels around 08:17:30 in the video, sorry for the bad platform but I was unable to find a mirror and their VOD replays are a lot worse than their streams as far as lagging goes.

 

Well, he's doing it the traditional way.
Let's dig a little deeper to help your understanding.
A bus, in it's original meaning, is an electrical conductor that is used as a common connection for several electrical circuits.
A bus in audio engineering terminology is exactly the same thing: an empty track on your console where you sum several other tracks using your patchbay (larger consoles have hardwired buses you can assign to the tracks by pressing a button) to allow even signal processing over several tracks.
In Ableton Live, you could do that simply by grouping the tracks, but it appears Joel is an oldschool guy who does it using audio tracks.
It doesn't matter which way you do it, there is no difference to it in Live.
The reason he doesn't use sends/returns is because he uses dynamics processors as inserts (stuff inserted into the track) on the bus. The rule in the analog realm is to do dynamic processing as inserts, fx (delay, reverb, chorus etc) as sends.
This has several reasons: first of all, if you use a reverb for depth staging, you want several tracks to go through the same unit. But more importantly, said effect units would cause serious phasing issues if used as inserts.

Was this what you wanted to know or did I get you wrong?

 

Yeah, that was extremely helpful and I would like to thank you for the very elaborate response! Can I pick your brain a little bit more with a follow-up question to your answer? Why would things like reverb cause phasing issues whenever they would be used as inserts opposed to using a return track and putting them on there?

The way I understand the grouping of tracks (within Ableton), the technique Joel is using and regular return tracks, the only real difference is that you can send different levels for different instruments to the same return track while this isn't the way the other two methods work. And the way Ableton works the return track will still get mixed with the dry signal at the master bus while using the audio input tracks would be the only input since they just reroute the instruments through that before hitting the master bus.

I don't know if this explanation is good enough and I hope it makes enough sense. But when following this logic, I don't understand how one of the techniques could cause phasing issues within the reverb while the other one doesn't. If there is one thing, using reverbs as inserts would make me think they would be less likely to cause phasing issues because all of the audio will go through there and there would no longer be a copy of the dry signal. So to my understanding, there would not be anything to phase against while return tracks would still have the dry signal from the original track present to phase against.

 

Generally you can't record on buss tracks.
So some daws that allow internal recording, audio tracks are used so that stems can be printed in realtime.
This becomes a must when using summing boxes.

 

you are confused HETIS..

 

You could say that. I checked some of my own theories out in my DAW and I guess some of my rambling in my previous post doesn't make any sense so sorry about that. Nevertheless, I still don't understand how an effect could cause phasing issues when used as an insert opposed to used on a return track. The sound of the FX plug-in is still there and as long as the parameters are the same, the sound coming from it are the same whether they are there on an insert channel or returns track. Is this an issue that would be limited to the analog world because of how the physical routing inside the mixer worked back then?

 

Ok, you got me a hard nut to crack there. I remember our professors teaching us the thing with the phasing, but I cannot find any reference to it in the literature. That may be the case because it's so basic nobody bothers explaining it, or maybe because it's not really an issue.
The effects you pack into the aux are usually time-based effects, like delay, chorus etc. So even if you dial the dry/wet knob to 0% wet you will add some latency. This might screw up the phase. But the more I think about it, it really is more of a best practice than an underlaying physical principle:
The first thing that comes to mind (still talking of a hardware setup here) is the quality of the electric circuits: the quality of the circuits in the mixing console is usually superior to the circuitry of an effect unit. So if you need only a fraction of the main signal to go trough the effect, you wouldn't want the rest of the signal to go through the (usually) cheap pcb of the effect unit.
Then, off course, you would want to get into the effect unit with the highest level possible (gain staging) so you use as much of it's dynamic range as possible as you would have to raise the signal afterwards and what happens if you boost a low signal in the analog domain? You raise the noise floor, and that's every engineer's nightmare because the higher your noisefloor, the less dynamic range is left for the music. So, you put compressors etc in the inserts because it frankly doesn't matter that much if a reverb that's barely audible goes into the noise floor a little sooner.
Then: inserts are usually always pre-fader, auxes are post fader by default.

But in the daw? It doesn't matter that much anymore. I put reverbs and delays as aux, for depth staging.
So why don't use the returns for summing? Because you want to send 100% of the signal to the summing bus. Because it can be confusing to have 6 knobs for send effects and one (or more) for sending the signal to the bus. And of course, if you don't make the return path "pre-fader", you will just send an additional signal to the mix - and if you do set it to pre-fader, you'll have to turn the track fader all the way down, loosing the channel meter in the process (this is also valid for the analog mixing console).

So it's not really about sound or phasing or noise floor anymore, it really is a question of best practice.

edit:

I forgot one aspect: if you use an aux bus for your summing needs, you won't be able to use the send fx on this bus. That would be a major drawback for me. Plus, it's really not convenient. In Ableton Live, select the tracks you want to sum, press strg+g, done.

 

Last edited by a moderator: Nov 26, 2015

Thanks again for your very elaborate response! This helped me immensely in understanding more about the signal flow of audio through your DAW and the different plug-ins. Signal flow and how to go about it is definitely one of my weakest suits when it comes to producing. Also nice to hear some stuff about all the analog gear for somebody that never had a chance to actually use any of that.

I will definitely keep a closer eye whenever I'm bussing or sending tracks from here on out. I very much am of the stance that whenever it sounds good it is good. Still, keeping these things in mind and becoming more accurate with my signal flow should be a good thing towards making sure there are less possible issues within my track that would make it sound bad.

Just like you said, all about best practices and usually they are best practices for a reason. But sometimes I'm just curious what made it so in the first place, I always try to never stop thinking critically about what I'm being told.

 

You're very welcome

Signal flow is a pretty logical step-by-step and what-if topic. Image a flow of water that's guided through a canal or pipes or stuff like that. The water is the signal and on it's way it passes several stages. Just try to follow every step from the source (instrument) to the end (the monitors).
Maybe this can help you:
https://soundsessionstraining.wordpress.com/2010/10/07/understanding-signal-flow/

 

I mix in Studio One. This DAW has FX returns and busses. The busses have sends where you can send to another bus or FX or both. Best practice currently is to end up with everything on busses. The reason is simple you are going to have to deliver stems to the record company. So Drums, Bass, Guitar, Keyboards, Vocals, Backing Vocals, Percussion, FX like uplifters, and maybe hook sounds on another bus. This also helps you greatly in your mix where you can trim whole sections or solo a bus with the complete drum tack etc.. If you send to FX off your busses then you will not have to worry about trimming fx when you trim a bus. But you will still need to on occasions send to fx from your channels. This gets a little tricky but you can get your head around it. I will return all reverbs on FX returns and all other sorts of FX on separate busses.

I suggest you build yourself a template in your daw with everything in place. This template you will modify continually as new tricks are learned and new fx are purchased. When you walk into a major studio you don't have to build your console it is there ready to go and if your assistant knows you he will have a lot of what you want already set up. Treat your in the box mixing and recording the same way. Build your template and import your audio into it or start recording with your template and the mix will develop over the recording process. This has worked for me over many decades.

When you have to deliver stems it is a lot easier to just mute what you don't need.