Colin O’Malley is an Emmy nominated composer for film and television, as well as games, concert halls, Disney theme parks, and the recording artist Yanni. While he is passionate about writing for orchestra, he is widely admired for his ability to turn samples into musical sounding works in demos for various developers. It is this challenge that I asked him to discuss with me.
Hi Colin thanks for doing this. Let me say flat out that I am a fan. Apart from your impressive body of work for film, TV, and games, you have earned a reputation for doing great demos for various sample libraries. It seems that no matter which library it has been for over the years, when I hear one of your demos it is always well-crafted, musical, and shows the library in the best possible light. So lets get to it. Tell me a little about your training, main instrument, etc.
I played piano growing up. In my high school years I got really into synths. I was double stacking an Ensoniq ESQ-1 and Korg M1 and rocking the talent show. I wanted to be Yanni, literally. I studied music in college, but when I got out of school I still didn’t understand the orchestra very well. I would say most of my real training came from landing projects with budgets for orchestra and composing lots of bad music. I wrote tons of over worked, busy, block chord material. Occasionally there were moments that clicked. I’ve spent a lot of time figuring out why. Voice leading is part of it, but in many cases it’s simple stuff. More unisons. Using basic counterpoint instead of forcing harmony to make myself feel smart. Stuff like that.
Career wise, I started off doing really low level stuff for Disney. Arranging Lion King songs for Disney conventions. Really groundbreaking stuff. I’ve continued doing a lot for them through the years. I’m now the Music Director for the Disney Christmas Parade on ABC. I also hold the Guinness World Record for completing the most arrangements of When You Wish Upon a Star
In 2004, I started working for Yanni as an arranger and orchestrator. I would prepare charts for his tours, but mostly my job was to arrange and mockup orchestral tracks. It was great because I was getting paid to see to see how far we could push the envelope with orchestral samples. We also recorded lots of custom experiments. He taught me a lot. I have a masters Degree from the University of Yanni.
How does this generally come about for you? Do developers approach you to do demos or do you find a library you like and offer to do a demo?
The demos I’ve done have been for good friends who happen to be developers. It really is a small community of people who are obsessed with sampling, as we are. In the interest of transparency I should tell you that I spent an irresponsible amount of time on the few demos I’ve done. Aside from basic pride and vanity, I felt very responsible to my friends. Composers are a tough crowd and first impressions of a library are very important.
Without being specific here, has there ever been an instance where you tried to do a demo for a product and just could not make it sound good enough to a point you were comfortable sharing it?
No, I haven’t been in that situation. I’ve only done 5 or 6 commercial demos at this point, and they’ve always been for libraries that I love.
Do you generally write things out on score paper first or compose directly into your DAW?
I only get into written notation when I’m prepping a score that is going to be recorded live or doing orchestrations for another composer. I’ve dabbled with sketching themes on paper or humming in my iPhone, but I basically compose everything in my DAW. I think there is something to be said for writing away from the DAW though. The idea of being uninhibited by sounds or playing skills is definitely interesting to me. I’m sure this happens to all composers, but I frequently have dreams where I’m writing or performing music that is waaaay better than anything I’ve actually written (at least it seems that way in the dream). I can also fly!! My point is I think we get in our own way sometimes as composers. Anything that might help us get around that is a good thing: writing notation, humming, sniffing glue….whatever it takes.
By the way, what DAW do you use? Do you also use Mac or PC slaves?
I’m running Logic 9 on an 8 core Mac Pro. I have 4 PC slaves streaming into my Mac over light pipe. I love the idea of working on one monster Mac, but I don’t think we’re quite there yet, at least not for large orchestral templates. I’ve stuck with hardware instead of some of the newer audio over LAN solutions simply because I need a lot of inputs and really low latency is important to me.
It has always seemed to me to be a losing proposition to try to bend a library to do what I hear in my head for a real instrument or section if it is not sounding good. When that is happening, I reach for another library and if it still is not working, well, then I simply compose something else.
When you are doing demos, how much do you write to the samples and how much do you try to emulate the real deal?
I guess I do work around the strengths of the sounds I’m using. However, with some of the newer libraries that is becoming less of an issue. The articulation depth and sound quality is reaching the point where it really becomes about writing. For me this is both exciting and terrifying. We have nowhere to hide. In that vein, I think 90% of what makes a mockup work are the writing and arrangement. As proof I have hard drives filled with well mocked up slop that I’ll never play for anyone. At this point most of my frustration comes from problems with my writing, not the libraries I’m using.
Regarding pursing realism in my music, it depends. Looking at the handful of commercial library demos I’ve done, yes, I was very concerned with realism. Increasingly I’m drawn towards composers that are mangling, processing and turning the orchestra and everything else upside down. In that case, realism goes out the window to some extent. I love that.
I have made it clear on a number of occasions that I make a distinction in my own mind between a library sounding “real” and sounding “good”, with an emphasis on the latter. We have a bunch of libraries, all recorded with good players playing good instruments in decent rooms, well recorded by good engineers and they all sound very different and each has their devotees among knowledgeable users. Which is how it should be but I do not get hung up on whether one or the other sounds more ”real” to me if I like the sound, as none of them sound like a real orchestra when you put the recordings next to each other.
I certainly have preferences in tone between different libraries, but most of the recent releases sound both real and good to me. Again, I think writing and arrangement have a lot to with realism and making things sound good. Having said all that, I think you’re right. If it sounds good who cares? Realism in a traditional orchestral sense can be limiting in a way. It depends on what you’re trying to do. I love Williams and Zimmer equally. There is not a single right answer for how things are supposed to sound. Classical snobbery is really boring to me.
Did you ever spend much time trying to mockup Classical pieces? Obviously in that scenario, you have to try to make the library do what the piece dictates. Some guys do a remarkable job at this but I would rather have bamboo put in my nails :)
No, but I’ve definitely set out to write pieces inspired by classical works or film music I love. I also mix my compositions against my favorite recordings. It helps me keep my perspective on balance. It’s easy to lose that with samples. Many years ago one of my first attempts in front of a full orchestra was a Debussy inspired piece. It was awe-inspiring for about 15 seconds.
In your own scores when there is not a budget for an orchestra or only a small group of real players, how many core libraries do you typically use? Do you do a lot of mixing different ones together or go for a more homogenous approach?
I used to Frankenstein tons of libraries together, layering, EQing and endlessly tweaking reverb. I don’t do that anymore. With the exception of percussion, I pretty much use 3 orchestral libraries now. I definitely don’t do a ton of layering. In my opinion there is very limited real estate in a good arrangement or mix. Every sound has to have a place and make room for the others. This is true with a live orchestra, but even more so with samples. Layering can destroy the air of the sound and take up all the space in a mix if you’re not careful.
Here’s a typical pitfall I’ve run into. I’ve created this awesome cello sound layering 3 libraries. When I solo it up it’s PHAT. The problem is it takes up 80% of my mix. So now I’m going to raise the level of my violins 200% so I can hear them, and destroy all hope of natural balance in the process. Did I mention I haven’t added percussion yet? It’s a nightmare where good ideas die. I have a hard drive full of them. I call it “sample cranking.” There are guys that layer and mix monster sounds brilliantly, but I’m not that good at it. The mixes I’ve done that I’m happy with work because I’m really careful about what makes it in, and I don’t layer a ton of things.
How much time do you generally spend learning a new library? Do you go into a library and kind of exhaustively live with it for a little while? Or is it a pretty much a need to know basis?
I usually just dive in and try to write. I don’t read the manuals as much as I should. Most of the recent releases are pretty intuitive in my experience, and it keeps getting better. Most developers seem to know we’re all short on time. I do think my opinions of libraries change after I write finished pieces with them, often for the better. I don’t always trust my first impressions noodling on the keyboard anymore. Out of tune noodling might be magic in context within an actual piece.
I have a kind of standard workflow that I prefer, I want to play a part in real time, keyswitch or program change my way through articulations while controlling vibrato with the modwheel and volume with an expression pedal so that it gets the part 80% the way I want it without a ton of tweaking. Personally, I tend to avoid libraries that won’t let me do that well but other guys whose work I respect will have voluminous track counts with individual tracks for separate articulations, which I would find overwhelming. What is your workflow?
When I’m writing for traditional orchestra, which is probably 75% of the time, I have a large template with about 300 tracks. I don’t see this as a badge of honor or anything like that. It’s just the way I’m used to working. I also tend to obsess over individual attacks of notes. Key switching feels limiting to me in some cases for this reason. I probably don’t need as much individual articulation control as I have, but I’m used to it and it works. My orchestral template is like spaghetti. Every piece I write, I’m tweaking balance, EQs and reverb. Collectively it seems to improve over time. I’d be completely disoriented if I started from scratch. I do think my track count will come down in the coming years as new brass and wind libraries are released.
Do you gravitate as I do towards libraries that play nicely with the other children so to speak, that blend well together?
Yes. I definitely prefer libraries that are well recorded in a natural studio or hall. I find blending libs that have natural early reflections a lot easier. It takes some significant EQ cuts and reverb magic to make a bone dry library sit as I want it to.
Any orchestral instruments that you still are not happy with the sample library choices? For me it is trombones. They never sound ballsy so I am excited that we have a bunch of new ones that will be coming out soon.
I don’t have too many blaring holes at the moment, but I’m definitely very jazzed about some of the upcoming brass and woodwind releases. Things keep getting better and better.
I would assume that on your decently budgeted projects you have an engineer but for demos you are the engineer. Am I correct?
To this point I haven’t hired an engineer to mix any of my sample based scores or demos. I’d love to, but the time and budget have never been there. I’ve mixed some of my live recordings as well, but always work with a mix engineer when I can. In my experience mixing a well recorded orchestra is way easier than pulling off a sample based mix. In both cases, I would still prefer to have someone else handle the final mix. It’s very hard to be objective about my own work. I think a lot of us learn to mix out of necessity, but can’t compete with the knowledge of somebody that does it professionally.
So when you are the engineer, how big a role do reverb choices play for you? Hardware, or software, or both?
Reverb choices are important, but I think there are a lot of great options out there. Currently I use a combination of a hardware Bricasti m7, Lexicon PCM Native and Quantum Leap Spaces. They all sound excellent to me in their own way. One thing I should mention: I try to write using as little reverb as possible. I hold back until I get the arrangement most of the way there. It’s not always very much fun, but it makes me work for things more. It’s kind of like mixing on Yamaha NS-10’s perhaps? In the end I usually don’t need as much reverb as I thought I would initially.
Do you go in for the separate instances for each section with separate instance for early reflections and tails approach or something simpler??
I used to go nuts, 8 instances of Altiverb, separate early reflections and tails for each section. I was very scientific about pre-delay. In the last few years, I’ve abandoned that approach entirely. Part of that change has to do with exclusively using libraries that already have the early reflections I want built in.
How much time do you have to spend with EQ? Some of my engineer friends say that if you have to do a lot the library was not well recorded but then I see guys with an EQ on almost every instrument.
I do some very focused high frequency boosts and limited cuts to clear up mids and low end mud. It’s a lot easier when I only use libraries that I consider to be really well recorded. Also by limiting my reverb while I’m writing and making sure the sections are balanced I find myself EQing less. In some cases I think I’ve used EQ as a band-aid for problems in my orchestral balance an arrangement. If something is fighting it might be my writing.
Personally I also look for libraries that when I just play them I feel inspired. I want my workflow to be as transparent as possible with a minimum of fussing until I get to a final mix. I guess I am old school in that I still think in terms of compose, then orchestrate, then mix, whole others do serious mixing as they go. I have tried this but for me as a composer, it is a flow killer. Your thoughts?
I definitely get sidetracked mixing along the way. I can’t always stop myself, but I don’t think it’s productive in my case. A creative idea has a limited energy and momentum and you have to get it all out as fast as you can. When I start judging and over thinking I seem to shut down the creative side of my brain and start writing logically. It’s a big bummer. The best composers I know are by far the most uninhibited and prolific. It’s definitely an area I need to work on. Having a template in order is half the battle, but it also may be a personality thing. To answer your question, I think you’ve got it right. Focus on writing first and then fuss with the mix.
Anything else you would like to share?
I have a friend who is a really successful photographer. He told me the key to his success is that his wife is a great photo editor. She weeds through the thousands of pictures he takes and finds the select few that are really saying something. I think we should do the same with our music: write a lot, pull out the small gems and let go of the rest.
Good advice! Thanks Colin for doing this, I look forward to following your career and hearing more good music.