Distinguishing between mockups and live performances

[Med8r]

...I think that we are entering this "turn in the road", and that if the person behind the MIDI keyboard really knows what he's doing, he can make the gap between live orchestra and virtual orchestra incredibly small. Thomas Bergersen is a good example. I can still remember being amazed that demos like "Love Suspended", "Allegro Agitato" and "Sampled, Not Stirred" actually weren't made using live recordings...that still blows me away sometimes. I had to listen hard to them several times before I could pick out things that suggested it was virtual. Alex Temple is another good example. "Storming the Gates" still sounds real to me, as do many of the other pieces he's done.

Granted, I'm not a world-renowned conductor or sound engineer. They may be able to easily tell that those songs were virtual. But with all due respect to them...they're not the ones who matter. The music they make is for common, everyday people who haven't studied music all their lives. And the average person can't tell the difference between live and virtual, if the virtual is done well.

So that tells me that the "turn in the road" is here, and as sound samples improve in quality, ease of use, and realism, the gap between live and virtual will probably get smaller still.

[gstitt]

For me, I always just listen for the instruments I'm most famiiliar with. I played trumpet seriously for a long time before getting into composing, and I have always hated virtual trumpet parts. The newer libraries are better but even then, the limited amount of attacks is what usually gives it away for me. For a given phrase, a good player instinctually uses subtly different attacks based on the context. Round-robin eliminates the machine-gun effect, but you still don't have much control over the initial attack. With a lot of fine-grained CC1 automatation, you can kind of do it, but it isn't perfect. I had scripts to help with this in Kontakt, so I am anxiously awaiting Play Pro.

[Bryan]

For me I think there are a number of virtual tracks that I have heard that sound very real, until I listen to a recording of a live orchestra. Then it becomes clear that the first track (the virtual one) is fake. I think anyone with really good ears and knowledge of what live players sound like can tell the difference. If you listen to something virtual that is done really well without listening to live players then it will sound more real than if you compared the two.

[TGV]

Let me look at it from another angle: is is that important to make it sound real? We have a certain sound ideal in our heads, and large part of it is formed by sound tracks, recordings, and live performances by real orchestras. It's natural we'd like to emulate that. Plus: what's more thrilling than having your music played by a large symphony orchestra in a great sounding hall?

However, we have to accept we're never going to sound like a real orchestra, not even if we have all the samples you can imagine for 30 different solo violins, 10 solo violas, 10 solo celli, 8 solo basses, 6 solo french horns, etc. We can't play the required 80 or so tracks as real ensemble players. That would require huge amounts of study to master each instrument, plus huge amounts of time to get each track right.

So what about sounding good enough? We should keep some standards, but we don't have to fool anyone into believing they're listening to real performers. We just have to make them like our music.

[lejmer]

Originally Posted by Jeff Hayat

Sorry, I don't buy that.

And you don't have to. That's the beauty of free will.

[Plasuma!!!]

Real sources I always use for comparison are a few tracks from Total Annihilation (large ensemble / mid and far micing) and Fable (smaller ensemble / close micing). They're not masterworks by any stretch of the imagination - they're very simple music - but it's pretty amazing how, even in the repetitiveness of the scores, there is so much variation. They're rather thin and the timbres are bare and plain to hear, which makes them helpful for picking out the natural characteristics of individual instruments.

I just don't think sample libraries can ever hope to achieve absolute realism in this area. I've noticed a few virtual composers compensating for a lack of the natural phrasing idiosyncrasies in their samples with lots of variation in arrangement and orchestration, which severely limits what styles they can write for orchestra. It tends to be either that or using synths to augment the timbres, but that's trendy now anyways.

[Doug Rogers]

I can tell you from years of experience, the average person (who is the one that really matters here) cannot tell the difference between a real orchestra and a sampled orchestra if the composer knows what they are doing.

Also, it's a fact that most composers that actually have a budget to use real orchestras also use sample collections such as ours to add power to the final mix.

So, while it's not possible to mimic exactly the organic nature and nuances of a real orchestra (or the fact that most players are slightly out of tune much of the time ), it doesn't matter for the intended audience.

This of course means that you can orchestrate a film/tv etc. with a modest budget, something that was unheard of not so long ago.

Cheers,

- DR

[Bryan]

Originally Posted by Doug Rogers

I can tell you from years of experience, the average person (who is the one that really matters here) cannot tell the difference between a real orchestra and a sampled orchestra if the composer knows what they are doing.

Nor do most of them care for that matter. All they are concerned about is if they like the music and it sounds good. We the musicians and composers are the ones who actually care about these things.

[Jason.M.Watts]

Very well-put, Doug! It is REALLY NICE to read your comment on this topic! Thanks for taking the time to share that!



Jason Watts

[epwpower]

I recall reading an interview where a composer claimed he used symphobia almost exclusively for one of those silly CSI-style shows (don't remember which one). And when I watch those detective shows, I can sometimes pick out specific samples that I recognize, or a "tone" that reminds me of a specific library. But to the average listener, it doesn't matter! I think it gets to a point where composers are just competing amongst themselves to prove who creates the better mock-up while no one else in the business even notices the difference.

However, as for composers not being able to tell the difference, I think Mr. Bergersen has been winning with that one since many years ago. Just listen to his song "Mojo Madness" which was made a few years back before the "better" sample libraries were around.