Michael Elman, the creative director of audio group Wave Generation, explores the issue
In the game audio world, we constantly use Hollywood as a benchmark for quality. Be it in mixing, sound design, or music scoring, producers and audio directors are frequently referencing movies as a way of inspiring game audio artists.
I do not disagree with this practice; Movies have been around a lot longer, and the film industry – particularly Hollywood – has mastered a lot of the processes required for delivering entertainment to its audience. The music and sound departments are most definitely part of that mastery.
There are of course, fundamental differences between the two products, and I think it is very important to understand where we can borrow from Hollywood’s processes and ideas, and where we must take a different path. In this post, I will focus on the subject of using live musicians and orchestras in game scoring.
To that end, I make the following observation: While using live players and orchestras in film scoring is widely considered to be an important step, in game scoring it is not crucial to take the same approach.
For example, when our lead composer at Wave Generation, Nicolas Marquis, started work on the soon-to-be released MySims SkyHeroes for EA (developed in Montreal by A2M), I asked him what live instruments we would need to record, based on the music style. Working closely with A2M audio director Jean-Frederic Vachon, we had determined that the score would have an action/adventure feel, featuring rock, orchestral, and electronic music. Nicolas' answer was quick and didn't require too much thought: “Only the guitars,” he said.
The deadlines were reasonable, and Nicolas could have chosen to use more live musicians if he wanted to. In the end, his priority was to deliver a powerful score that supported the gameplay, and that could be delivered in a variety of splits should the developers need different mixes for dynamic changes in the music.
He knew he could successfully carry the score with a string section and support it rhythmically with electronic beats by using predominantly digital instruments. At the same time, he could give the sound an edge with a real electric guitar.
So in the case of this particular project, it just wasn't important to record the score live, and this has been true for many of our productions. Let me be clear – it sometimes sounds better to record a game's score live with an orchestra, band and soloists than to rely on digital sampling. However, this is only true in the following cases:
• The score is orchestral and has tons of linear cinematics or music cues.
• The score is for a music game that requires very realistic sounding performances, as in Guitar Hero.
• The composer is using instruments in his score that he (or perhaps anyone) cannot properly recreate with sampling and sequencing. For example, it is difficult to digitally replicate acoustic and electric guitars, so it’s better to use the real instruments in this situation.
On the other hand, digital orchestral scores sometimes sound better than live recordings. I once had a client call and say they preferred the composer's digitally sequenced demo tracks over the live orchestral recordings they had done elsewhere with a major US symphonic orchestra.
This was probably due to a variety of factors, such as the recordings may not have been properly mixed and mastered or the performance (that was a one-day session) was not executed as well as the composer had originally "performed" it while sequencing. In any case, a recording of a one-day live orchestra session may not sound as good as a one-day digital session.
Aside from sound quality, there are a host of other benefits associated with digital sequencing:
• It's cost efficient. This means that there is more budget to deliver a greater quantity of music or spend more time on the project in the pre-production phase.
• It's much easier to work with. Live recordings can be done in segments and parts, but it still does not give the flexibility that MIDI sequences have. In an "all sequenced" environment, you can make last minute changes to tempo, instrumentation, and of course, the composition itself. You can also export any combination you need of split tracks for interactive and dynamic music systems.
Sometimes we will get a request for recording a live orchestra, and for the reasons I mentioned earlier, it is a great idea and we are happy to get it done. But sometimes it feels like the producers who are requesting a live orchestra are really trying to say: "go the extra mile for us.”
We certainly appreciate their passion for what the music brings to their game. When we get this type of request, I like to brainstorm with them on what else we can or should do to hit it out of the park, and that usually leads us down a different path.
Sorry Michael, but I think your article misses the point.
Stylistically, an orchestral score is not appropriate for many (if not most) game scores. This is also the case for a large percentage of films. Today, you will be hard pressed to find a purely orchestral score accompanying a film (a la John Williams, or John Debney..). Using a large Hollywood Orchestra is very expensive, and out of reach for most film and game budgets.
The notion that any trained orchestral composer would prefer a sampled orchestra over a real orchestra is completely absurd. Some composers may prefer using samples because it better fits their expertise, but in the hands of an experienced orchestrator, a good Hollywood orchestra will always sound better than a sampled one.
Hi Mark,
Thanks for the comments.
Which point am I missing? I did not say that any trained orchestral composer would prefer a sampled orchestra over a live one.
I did say, that there are arguments to be made to game producers over where to spend their money, and a live orchestra is not always the top priority.
I think what you've written is quite valid with regards to the flexibility a midi sampled score gives you, but I agree with Mark that you'll never get the gravity of a real orchestra with a fake one. I agreed with the point about recording electric guitars...but acoustic?? I'd say that most orchestral instruments are harder to simulate!
I've been asked many times over the years to quote for a real orchestra...and it's always budget that limits it to a sampled score with a handful of real players (a fake cake with real icing!) at the end of the day.
Anyway, must fly...I've got a John Williams style fake orchestral score to complete! :-)
Hi Doug,
I think composers should decide on which instruments they record live and which ones they sequence based on their personal expertise. If you find it easier to sample Acoustic guitars, then go for it. The main point I would like to make is that game producers should not feel like they need to record a live orchestra for their orchestral or semi-orchestral game scores. While it's a great thing to do, it often doesn;t make sene for financial and production reasons. It is in that regard that I am comforting them on the advancements in sampled productions.
I agree with everything you say, but... I feel you miss the fundamental reason for using an orchestra apart from the fact that a live recording should sound better. To me using an orchestra should produce a more emotionally resonating and significantly better performance. An orchestra should consist of highly qualified expert players capable of with the aid of a great conductor making an interpretation of the music that resonates at an emotional level that is difficult, near impossible for a composer with samples to achieve. That said there are many occasions when a composer can achieve that emotional resonance without an Orchestra, and it is the job of the composer and musical director to make the decision that an orchestra is required or not to achieve the desired result. The problem to my mind is that the industry lacks the audio/musical directors that are able to make the right choices and the decisions to use orchestras are taken, like you suggest, as a political decision, "this will raise the perceived production value of our game" or as an emotional decision not based on the musical emotive facts, by the composer," my score will sound better and I will gain respect because my score is being performed/recorded by an orchestra". Both of which may be valid reasons but are not what should really be the deciding factor- which is to my mind. "Using an orchestra or live musicians will enhance significantly the players experience/ emotional engagement with the game."
It is my opinion that music plays a very significant role in the overall enjoyment of a game, and is worth investing in getting it right and failure to do so poses a more significant risk than is generally appreciated, I would like to have statistics on this to back up my assertion. With these I feel it would be possible to make informed choices on how much priority and consequently budget and attention music should receive, which is a big deciding factor on use of orchestra/live musicians. How could we acquire this information?
While we'll probably be arguing Michael's point on the use of live musicians for a long time, my favourite part of his argument is that movies and games are two superficially similar mediums that deep down are quite opposed.
In the early days of games, music had an identity that was unique to it. It may have been brought on by the technical constraints, but it was still a unique genre, one to which people are still emotionally attached. When those limitations were dropped (or pushed back), we became overcome with what I call "cinema envy", trying to turn games into films, including the scoring. While some stunning pieces of music were created in that genre (and I'm guilty of using that style frequently for our games), I think the next step of evolution for our business will be to redevelop a music aesthetic that is unique to games.
Thanks everyone for keeping this discussion going!
Simon - I think we all agree that more priority and budget would be welcome to all of the games we work on. As far as proof of music's value in games, it's of course impossible to quantify. However, as producer's gain experience working on games with great music, and without it, I hope the importance becomes more obvious.
Jean-Frederic, I agree with you. I am all for being inspired by film scores, but we need to strive for scores that in and of themselves inspire future game composers, not only for their melodic and production quality, but their overall technical approach as well, from production to integration.
It's incorrect to say that guitar is harder to emulate than orchestra. Maybe if you're listening to orchestras that are imitating mock-ups, like Hans Zimmer and Steve Jablonsky crap, but not if you listen to actual orchestral music. The person who said the mockup sounded better was probably talking about something written FOR samples, not for orchestra. An orchestra doesn't act like a bunch of pads and hits, like samples do. And few composers who use samples know how to notate what they hear. The level of quality that's resulted from sampled orchestras is astoundingly low. If it doesn't matter that it's real or not, it's because standards have fallen, and "orch hit" General MIDI with some prefabricated loops might as well do just fine for game developers who are accustomed to spending 1% of the budget on music as opposed to 5%. Most console games that want orchestral sound could hire an orchestra for 5% or less of the budget, on par with film portions, but they don't because game companies have lower standards. The same goes for most cheap ass straight to video B movies starring horrible CGI alligators, which is about what most games aspire to be in the music department. But if you take the top tier of game composers and compare them to the top tier of film composers, no comparison, the film composers win by a mile. Not for long, if film music standards continue to lower, though, but games will never have a John Williams until game producers develop some taste.
Furthermore, there is no such thing as a game music style, or a film music style. Early game music is just electronic music, which already existed. Film music can be anything, and invented or owns nothing. Musical styles do not arise out of visual entertainment. Visual entertainment uses existing musical styles. Game or film, doesn't matter. Most game music is recorded music, and should be held to the standards of recorded music, not the standards of the SNES, which by the way had better standards than most sampled scores in that it maximized the potential quality. To try to say that game music stands on its own as a style that can't be compared to film music is narrow minded. It's all just recorded music, and most game music is just on the level of straight to video B movies.
