Bruce Richardson and everybody

[Joanne Babunovic]

Bruce,

Listening on your website - what are some of the key steps to creating that level of space and air in your music? If most of those comps contain a large amount of live instruments, that might explain it to some degree.

How much of the air and space you achieve is just simple panning? Do you always try to pan to realisim? ie. where instruments are typically placed on the stage?

What is your usual effects template for sample instrument compositions? How do you usually apply reverb/EQ - and what reverb/eq do you use? Maybe the less effects the better in most cases?

For someone just learning how to apply effects - what approach would you recommend. ie. just start using a straight reverb and then progress to convolution etc?
Thanks,

Joanne

[Bruce A. Richardson]

Hi Joanne,

Short of writing a book in response, I can only offer up a few basics which can hopefully get you going. People spend a lifetime becoming really great engineers (I am not particularly a great engineer, just functional and a quick study). So, there are thousands of tricks that people use to make mixes sound great.

In general, when you are creating a recording, you are creating a space from scratch.

So, the first thing you want to decide is the size of the space, and where you want the "listener" to be within it.

Then, you need to make your source tracks conform to that idea.

Using samples is tricky, because you don't have control over the microphone placement. A stereo-recorded sample recorded at close range will have a significant amount of "spread" between the speakers--which will make it sound intimate and close to you. If you want to make such a sound seem farther away, you have to imagine how that intervening air would affect the source of the sound, and emulate that.

Say it's a clarinet. If a real clarinet player were in the room, two feet from you, the sound would be wide in your perception, warm, with all frequencies hitting your ears. If you told your clarinet player to stand ten feet from you, the sound would be thinner, and it would simultaneously be more focused to a smaller spot in the room, while the sound of the room's response would be more prominent and spread around you. Imagine it to be a pebble dropped in a swimming pool...the ripples would come from a tiny point, but they'd be bouncing over the entire pool and back over each other.

So, this is where your panning, EQ, and reverb come in.

Say this clarinet sample were recorded in the close-stereo position...but you want it to sound as if it were 15 feet from you and to the left, sitting in a woodwind section.

First, just throw away one of the stereo channels. Why? Because you will never be able to avoid the impression of a 30-foot wide clarinet when you start adding the reverb. You'll want it to be a mono source.

Now, pan it to the position you want it to occupy. It's coming from the right location between the speakers now, but is still too "fat" to sound plausibly far away.

Here's where you start simulating the effect of air. Audio quite literally burns up traveling through air--the farther away from a particular instrument you are, the more it is going to lose its weaker overtones. Only the strongest and most fundamental survive the distance. So, your clarinet needs to "lose" some of the content that is not at the "core" of its tessitura. I'd set a lowpass filter to 3k and a highpass filter to 300 hz, and start reducing those ranges. Try 6 db at this point. Your imaginary player should start sounding a little thinner and farther away.

The last step is to simulate the behavior of the room. This is where reverb is used. Early reflections give us cues as to where the walls are. Remember our pebble in the pool? The early reflections are the first set of ripples that have bounced off the edge of the pool, and returned to the "source."

But the ripples don't stop when they come back to the place where the pebble fell in. They keep going, and bouncing and re-bouncing, until all the surface is covered in a complex interlocking pattern of hundreds, then thousands of ripples. The "tail" section of reverb represents this...the amount of time, after the first reflections, that sound continues to bounce around in a room before friction converts it to heat and it dissipates.

So, setting up the balance of the reflective sound to the tail sound in the reverb also becomes a part of your soundstage design.

Convolution achieves much the same as traditional reverb. In fact, most convolvers now have traditional reverb controls which alter an envelope over the impulse waveform to "sculpt" differences in the early reflections and tail components of the recorded impulse.

See what I mean about this being complicated?

I think that I've covered the essential basics. You just need to imagine what kind of space you want to create, compare the tracks you have with how those instruments would behave in real air, and then make the panning, eq, and reverb decisions which will accomplish that goal.

Engineering wise, the single thing to remember is that EQ and panning are filters. You are filtering components of sound which are in excess of what you need to make your concept work. So, in using them, always imagine "taking away" what is keeping the sound from working, rather than adding something to make it work. Subtle distinction, but a very important one.

[SteveHanlon]

I'm happy your here, Bruce! You explain it so well.

Quite frankly, I think a camcorder, a mic and you speaking would be wonderful tutorials...on music, of course.

You can leave the 'how to make a bloody mary' for the Bloody Mary forum. ;)

[Joanne Babunovic]

Bruce,

Perfect level of explanation and thank you. Listening to your music, you obviously have an exceptional capability with building spaces and direction and placement, and to offer up such clear and easy to follow techniques like this is appreciated and generous.

I was struggling with the exact points you talked to. My samples are all stereo and therefore impossible to adjust and fine tune the placement to build sense of space. I thought maybe the solution was to double the tracks i.e. have the Violin123 stereo sample panned way right on one track and then take that same violin123 stereo sample on another track panned way left - but all that does is create mud. I thought about somehow converting two stereo tracks to mono, but it appears it's alot easier - just drop one of the stereo tracks.

I got a hint at the main application of eq when you talked about working general voice audio for your film. EQ makes things closer or farther. This is likely intuitive and obvious to most, but was not to me. Another point - I did always think of eq and thining or taking away, but never thought of "panning" as a taking away effect - as you say, an important concept.

I download the 15 day free version of WAVE IR1, which I think is the convolution reverb that also traditional reverb that you mentioned. Needless to say, very much looking forward to applying these techniques and again, thank you.

Joanne

[dpc]

I agree that Bruce's lesson was a model of clarity and extremely helpful. He saved about ten pages there with one of the best explanations of sound I've ever read.

In Giga 3 you can collapse the stereo image with the width parameter found on each instrument channel. Then you can place the instrument in the room with Gigapulse. Problem is I still don't really get Gigapulse very well (I'm a little slow on that end of things.)

Bruce,

Any helpful hints on the basic approach to Gigapulse? We should be buying you sixpacks or something.

Thanks,

Dave Connor

[Drew B]

Good question Joanne.
The sense of space and air in Bruce's near-recent demo of VSL Glass & Stones was the first thing that struck me. I think I commented it on it at the time.

Bruce, do you ever narrow the stereo field of your samples ?

[Joanne Babunovic]

Thanks David - I unfortunately don't have GS3 and typically use GSv2.5 for strings and gold via kompakt for the rest. I am in the process of trying to figure out how to cut out a stereo channel. My guess is that it needs to be done at the sampler source (SI on GS and Gold via Kompakt are my usual libs) rather than in my Cubase SX sequencer.

Steve and Drew - although you are likely more advanced in mixing than I am -I guess hearing the basics with clear and creative explanations never gets old. After I heard Bruce's stuff, things I was relatively satisfied with now sound like a muddy jumbled mix of misplaced instruments - but I guess identifying a bad mix is a first step forward.

[S Johnson]

BUMP!

A very good read! Thanks, Bruce.:D

[Joanne Babunovic]

Thanks David - I unfortunately don't have GS3 and typically use GSv2.5 for strings and gold via kompakt for the rest. I am in the process of trying to figure out how to cut out a stereo channel.

Yes, a great writing from Bruce. Do you believe I've been hanging in thin air here since last summer, waiting to hear how to get a mono instrument?

[Jim Clark]

Hi Joanne,

I'm not a cubase user (I use Nuendo), but can't you just bounce it (export audio) as either a mono file, which will sum left and right, and then import the mono file. Or you could export as a split stereo, then discard the side you don't want to use and import as mono? I think either would work, with a slightly different result.

[Bruce A. Richardson]

Yes, a great writing from Bruce. Do you believe I've been hanging in thin air here since last summer, waiting to hear how to get a mono instrument?

Oops. In GigaStudio, assign it to a discrete channel in the DSP station, then go into the DSP station mixer, turn down the channel you don't want, and pan the other side where you'd like to put it.

[Joanne Babunovic]

That works Bruce - thanks. Seriously, when are you writing a book?

[Joanne Babunovic]

Hi Joanne,

I'm not a cubase user (I use Nuendo), but can't you just bounce it (export audio) as either a mono file, which will sum left and right, and then import the mono file. Or you could export as a split stereo, then discard the side you don't want to use and import as mono? I think either would work, with a slightly different result.

Thanks Jim

[Nigel W]

Good question Joanne.
The sense of space and air in Bruce's near-recent demo of VSL Glass & Stones was the first thing that struck me.

I'd like to hear this-where do I find it please?

Bruce, very much appreciate the fine discourse on mixing and space:-)

Nigel

[Houston Haynes]

Thanks David - I unfortunately don't have GS3 and typically use GSv2.5 for strings and gold via kompakt for the rest. I am in the process of trying to figure out how to cut out a stereo channel. My guess is that it needs to be done at the sampler source (SI on GS and Gold via Kompakt are my usual libs) rather than in my Cubase SX sequencer.

Joanne, there's an intermediate step you can take with Cubase SX that can help you "narrow" the placement of instruments after you've crafted the part and are ready to start mixing. You can render the parts out of GS and then import the stereo files into Cubase and then "place" them from there. Cubase SX has three panning modes - the Stereo Combined panner being your best friend in this circumstance. Of course, if you are working with a sample library that doesn't have mono compatibility in mind, you might run into problems. But if it holds up, then you can use Stereo Combined panner to narrow the "field" of the sound without throwing away have of the sound that makes it sound like - well - the sound you started with.

There are lots of other tricks for improving placement (such as having different impluses from different locations in the same hall - and sending instruments that map that location through them) but as Bruce and others have pointed out, starting with a good stereo spread is the first step. Check out the Cubase Unleashed tutorial demo at www.auxbuss.com for more tricks on how to get the most out of Cubase. The free PDF of Cubase SX Complete is also a worthy read - I've had it in trade paperback and still refer to it on occasion.

[Joanne Babunovic]

Thanks Houston. For the GS stuff, I was using Cubase SX panning to place things and that seems to help. I'll check out the Cubase tutorial.

One question - you suggested applying different impulses to different instruments. I am suppose to apply the same reverb to everything, at least that is what I've read. I can adjust the amount of that reverb per instrument, but the flavor stays the same for everything. Are impulses different? Do you routinely apply different impulses to various instruments?

Thanks,
Joanne

[Houston Haynes]

Joanne - there are "positional" impulses that are generated in the same space with the impulse driver placed in different locations in order to closely calibrate the initial reflections. That's the thing that really "nails down" the location of a sound, although it can also invite other problems if not mixed and managed carefully. It's something that I'm still experimenting with (using Voxengo's impulse modeler) in anticipation of Garritan's "Real Spaces" impulse library that's on the horizon.

In large part, I use a single stereo impulse for general tracking/mixing, and then, if the mix is "close" to the instruments I mute/disable the stereo impulse on the main buss and begin to dabble with positional impulses on frozen tracks. Depending on the density of the music and the mix, it can make a big difference. However, I've also found that the net effect of going to all of the trouble of laying out renders of instruments and/or sections with positional impulses yields a negligible net effect on the final mix (compared to careful stereo panning of the dry signals and routing them all through one "generic" stereo impulse. You really need to spend some time with it on a monitoring system you can trust in order to get to a point of knowing when/where/how it pays to go to a more detailed impulse layout.

[Joanne Babunovic]

Thanks for the detailed explanation Houston. I'm only working with basic reverb, so as you say, I'll need to get something like Voxengo experiment and begin a working understanding.

Much appreciated,
Joanne

[geronimo001]

Yea! thanks Houston, you've just answered a couple of question right there.

Also, thanks for the Cubase tutorial link, i really needed that one.

Geronimo.

[Houston Haynes]

Yea! thanks Houston, you've just answered a couple of question right there.

Also, thanks for the Cubase tutorial link, i really needed that one.

Geronimo.

Hey - no problem. Heck - I'm going through all of the Cubase Unleashed tutorials front-to-back to retune my own workflow and processes. I've got so many key commands and macros for music-making versus sound design/mixing that I'm going to go back to square one (after a major DAW hardware upfit) to subdivide into two seperate setup configs, including preferences, menu visibility, timebase defaults (musical versus linear), and macro definitions that are at the top level of the controllers that I use. I'm finding that the workflows are diverging enough for me that it's really necessary to have "two hats" - and Cubase SX/Nuendo allows quite a bit of one-button re-config of the system. Of course, Nuendo goes farther than Cubase in that regard, but the fundamental operations that exist in both apps are all there in the Cubase Unleashed tutorials. The PDF that's available on their site is also a go-to resource. Great stuff.

[alanb]

How do the "Cubase Unleashed" tutorials compare to the other commercially available Cubase tutorials ("ASK Video," etc.)?

I watched about 5 minutes of the "Cubase Unleashed" demo on the Auxbuss website, and thought that it was good, but I wonder whether anyone can offer any useful comparison between that tutorial and any of the others that are out there?

A lot of folks rave about the "Cubase Unleashed" series over on the Cubase forum, but the Auxbuss site points out that "[t]he tutorials have been produced by well known Cubase SX3 expert, Bas (until recently, moderator of Steinberg's Cubase SX forum) . . . ." so I want to make sure that the raves are justified, and not merely products of forum-loyalty.

Thanks for any insight!!

[Houston Haynes]

No way to know from my side - I've never used any of the tutorials you mention. I'm not sure a comparitive analysis has ever been done. Keyfax has a series of Cubase tutorials, too - have never seem them either, but I bet they're pretty good.

[geronimo001]

Hey - no problem. Heck - I'm going through all of the Cubase Unleashed tutorials front-to-back to retune my own workflow and processes. I've got so many key commands and macros for music-making versus sound design/mixing that I'm going to go back to square one (after a major DAW hardware upfit) to subdivide into two seperate setup configs, including preferences, menu visibility, timebase defaults (musical versus linear), and macro definitions that are at the top level of the controllers that I use. I'm finding that the workflows are diverging enough for me that it's really necessary to have "two hats" - and Cubase SX/Nuendo allows quite a bit of one-button re-config of the system. Of course, Nuendo goes farther than Cubase in that regard, but the fundamental operations that exist in both apps are all there in the Cubase Unleashed tutorials. The PDF that's available on their site is also a go-to resource. Great stuff.

That sounds good, I'm in the process of building templates to improve my ''workflow'' and I'm a bit stock really. I'll start a thread regarding templates, please feel free to jump in!;)

alanb
How do the "Cubase Unleashed" tutorials compare to the other commercially available Cubase tutorials ("ASK Video," etc.)?


I have no idea but for a beginner like me, it looks pretty deep.

[alanb]

The PDF that's available on their site is also a go-to resource. Great stuff.
That free .pdf is called "Cubase SX Complete," as opposed to the $50 tutorial, which is called "Cubase SX3 Unleashed." The latency chapter of the free .pdf book has hyperlinks at the bottom of the page which read "Upgrade to Cubase SX2 Complete."

Soon as I saw that, I got nervous about relying too heavily on the book, since I'm running SX3, and don't know what info is outdated now...

[Houston Haynes]

OK - fair 'nuff. I've been a Cubase SX user since version 1, so I'm not as sensitive to the distinctions. I think that it's a better resource than the manuals, and offers a lot more *pre*scriptive elements than the PDFs from Steinberg, which tend to be more *de*scriptive.