To what degree can technically demanding scores such as Tom and Jerry's (Scott Bradley's music) be performed on synthesizers? Which specific synthesizers are capable of the best quality performance?

I'm a layman, so if my question is wrong or embarrassing. I apologize in advance.

Thank you.

  • Start here and listen to the samples for yourself. Synthesizer is actually no generic term, but a specific waveform combination device, while you seem to use it very broadly. Same sounds is also too generic to be useful, but I would say electronics today comes pretty close. – guidot Jan 22 '19 at 8:04
  • 1
    Synths will go way beyond the sounds any orchestral instrument can produce, including what they do produce. Depends considerably on the player! Heck, if a steel band can sound like an orchestra... – Tim Jan 22 '19 at 8:19
  • These two answers contradict each other. So can they or can't they? – Elliot Jan 22 '19 at 8:34
  • Actually I think the two answers so far (mine and ghellquist's) agree very well! Both of us are saying that although synthesizers can do a lot, it's really hard to get that last few percent of realism with a synthesizer. Whenever you ask if something's theoretically possible (as per your comment on my answer), often the answer's yes, even when that thing is very hard and usually impractical. – topo Reinstate Monica Jan 22 '19 at 8:51
  • No, I mean these two comments. Not the actual answers. – Elliot Jan 22 '19 at 8:57

Computerized reproduction of orchestral music gets better all the time, and whether it's good enough for some purpose depends on the purpose. I would say that the percentage of needs where a software-based automatic interpretation is "good enough", increases all the time, at the same time as the general public (i.e. audience) gets more and more used to all sorts of artificial sound.

Here's an example of Note Performer 3 by Wallander Instruments playing a Star Wars score notated in the Sibelius notation application. Good enough for ... something?

The main idea with Note Performer is that in combination with a score-writing program it interprets traditional musical notation into a performance, rather than relying on a "programmer" musician doing the translation and feeding it low-level geek-oriented MIDI data. Because Note Performer is driven by higher-level notation symbols like hairpins and slurs from a notation program like Sibelius, Finale or Dorico, it knows (or tries to reverse-engineer) a bigger picture. I'm not sure (after looking at the PDF manual) at which software layer the interpretation from symbols to data happens in each case, but anyway, Note Performer constructs a fluid continuous flow of expression, controlled by musical symbols. (And it needs a one-second delay for doing this)

With traditional sequencers and sample libraries, everything needs to be "programmed" in terms of, say, note velocity numbers, key switches, control changes and program changes, which end up triggering samples and switching or cross-fading between sampled notes. Wallander's synthesis engine is not based on entire sampled notes, but a "component-sampling + acoustic behaviour modeling" approach where each instrument's sound has been sampled as small components that are used to construct the whole sound in an additive fashion, guided by complicated behavioral instrument models which decide how much of which components are needed in order to produce the requested notes and pitches with the requested expression parameters. I'd say it's more of a synthesizer, definitely not a sampler. And it's not based on physical component modeling either.

| improve this answer | |
  • That's an interesting piece to use as an example. Good pick √. Initially [first 5 seconds] it's extremely convincing - the samples at least give a very good first impression... but after about 10 seconds it has become 'mechanical' & I'm afraid I had to switch it off after about 20s. – Tetsujin Jan 22 '19 at 20:23
  • @Tetsujin It's not samples, it's closer to synthesis really. Part of the mechanical feeling might come from Sibelius. A well-known tune is a hard test, because of all the expectations. Anyway, I think many composers would rather work in an environment where the computer plays your score and you only "conduct" at most, instead of having to play the computer using geek language. – piiperi Reinstate Monica Jan 22 '19 at 21:07
  • I thought this was quite impressive in terms of the timbres of the instruments. The main things that felt musically deficient to me just had to do with the fact that there was no human expression being supplied by a conductor and section leaders. The tempo doesn't vary. Dynamics are fairly constant and don't sound like the ones human players would supply based on emotion and expression. They just sound like they come from the score markings, which is of course where they came from. But if this really did come straight from a score without any tweaking, then it's a technical tour de force. – Ben Crowell Apr 24 at 16:44

There are two types of synthesizer that you'd look to if you wanted to recreate a realistic orchestral performance - sample-based synthesizers, and physical modelling synthesizers.

Sample-based synthesizers (and sample libraries) actually 'cheat', in that they actually contain recordings of real players playing the different instruments - with a high quality library containing a large number of techniques played on that instrument. So how close you can get to an orchestral sound depends on how extensive your sample library is, and your knowledge of the library and how to use it. Even with a very extensive library, it might be that some instrumental lines - especially in prominent solos - are, in practical terms, easier and cheaper to record using a real soloist if a realistic sound is what you want.

Physical modelling synthesizers are a mathematical model of the physical behaviour of a real instrument. Some of them are starting to get very good. However, it is not necessarily easy to get them to respond in the way you want - a lot of detailed control information is probably needed, such that creating a realistic line may, again, be harder in some cases than getting a real player to play the line for you.

So in summary - you can probably get close to a real orchestral performance with synthesizers, but getting really close would require so much effort that it might not save you any time or money over using a real orchestra. Synth manufacturers have, for the most part, been happy to focus on what synths do well rather than really make the effort to put orchestral musicians on the street...

| improve this answer | |
  • So, theoretically, with enough effort, money, time and an extensive library, a composer can perform it all on a synthesizer? – Elliot Jan 22 '19 at 8:42
  • @Madara yes, if only because sample-based synthesizers already contain recordings, so they can (theoretically) contain a recording of any sound that an orchestral instrument can make. – topo Reinstate Monica Jan 22 '19 at 8:48
  • This track is completely played on a Korg M1 and Korg M01/W. This is not difficult to play? – Elliot Jan 22 '19 at 9:28
  • @Madara it's not particularly difficult to get results like that - but do you think that track sounds realistic? I think parts of it sound quite unrealistic. – topo Reinstate Monica Jan 22 '19 at 12:18
  • 1
    The advantage of Wallander's approach is in its balance of what is explicitly sampled vs implicitly model-generated. With physical modeling, all sound is created implicitly by the model. You cannot just command the model to start sounding like a beautiful sample, or sample the entire behavior of all the physical components of a real-life example. It may be doable for a guitar amp or compressor that takes an electrical audio signal as input and has relatively simple interactions between its (few) components, but how do you sample all relevant interactions of a good violin's physical components? – piiperi Reinstate Monica Jan 23 '19 at 21:38

My answer is no. At least for now.

Let's compare with a professional symphony orchestra. All players have played their instruments full time at least 5 years, often up to 45 years. The requirements are brutal, you could compare with the top teams in NHL ice hockey or perhaps Harlem Globetrotters.Before these at least five years, most players started in "minor leagues" some 10 to 15 years, often starting playing from five years old.

The orchestral instruments has often been developed over a long time. I play the bassoon and the instrument went through a long development period, starting somewhere in the 17th century, accelerating in the 19th century and still ongoing. (To be very clear, I am an amateur player and the pros are far, far, far ahead of me).

Synths can come close in sound and expression, but there are few players that can use the full potential of expression. I believe that over time, a few players will reach high levels, and with time smarter programs (AI) can move the field forward. But, alas, I really can see that in the near future (but that is me).

But in playing other types of sounds or music, the classical instruments fall short. The palette of sounds is much larger for synths.

My five cents....

| improve this answer | |
  • AI could be a great to generate all that control data needed for a realistic physically-modelled performance... interesting times! – topo Reinstate Monica Jan 22 '19 at 8:33
  • It all depends on how you interpret "complexity," and more important "quality" . – Carl Witthoft Jan 22 '19 at 14:01
  • And us amateurs produce a sound that the OP probably doesn't want to achieve :-) – Carl Witthoft Jan 22 '19 at 14:04

There are some high end sound modules and synths that have multiple pristine samples for every note every orchestral instrument can play. The problem to me would be intonation, how to play a keyboard so it sounds like I am blowing a french horn. But I think a skilled player with an aftertouch enabled keyboard might do a pretty good job with some diligence.

| improve this answer | |
  • True, but since every "real instrument" player in the orchestra has a slightly different sound quality, those sound samples will require as a minimum some very creative alterations (pitch, vibrato, overtone, and more) applied and overlaid to sound like, e.g., the 10-violin First Section. – Carl Witthoft Jan 22 '19 at 14:03
  • Good point. I should look on youtube. Possibly there are some examples there. – user48490 Jan 22 '19 at 19:53

It can get pretty darn close! And, although a convincing rendition requires a lot of skill, it's probably nowhere near the amount of skill each member of a live orchestra has to bring to the job.

Depressing, isn't it?

| improve this answer | |
  • It sure is. Any recommended synthesizers that comes close to that level of quality? – Elliot Jan 22 '19 at 13:58
  • A computer-based sequencer plus extensive (and expensive) sample libraries. – Laurence Payne Jan 22 '19 at 14:00
  • How much time and effort would it take? And with no loss in quality? – Elliot Jan 22 '19 at 16:33
  • "How much time & effort"... about 20 years, 10 if you're very good. To reproduce even a single instrument convincingly, you need to know in intimate detail how it sounds when played. You don't need to be able to play it yourself, but you need to understand how it works when someone good plays it. – Tetsujin Jan 22 '19 at 20:18
  • No, I mean if you're really good and know your stuff, how long will it take to play that. Not learning to play, to actually play it. – Elliot Jan 22 '19 at 20:39

Vienna Symphonic Library


Unless I misunderstand what those audio samples are, or the company is faking it, they offer a sample of famous orchestral works, performed on their virtual symphony.

Listen to The Rites of Spring, it's absolutely impossible to tell that it wasn't a "real" orchestra. I put "real" in quotes to point out these are extremely detailed, digital sample libraries of real instruments.

Get ready to shell out big money, these products are not cheap! The full library is $14,000 US!!!

Also, I'm sure the learning curve to using these virtual orchestra - to sound very realistic - is steep.

| improve this answer | |
  • Or listen to Star Wars Main Title as a direct comparison to piiperi's example... but then go listen to the original. I think the problem isn't that the sample set isn't very good [it sounds very coherent] I think the issue is the programmers weren't very good. They didn't quite understand how to turn that magnificent set of notes into a performance. – Tetsujin Jan 22 '19 at 20:34
  • If I understand what Piiperi's Star Wars audio is - a computer reading and performing the score - I think you way overstate that it sounds mechanical. – Michael Curtis Jan 22 '19 at 20:55
  • The Note Performer is of course mechanical; unlistenably so. However the VSL stuff is supposedly 'performed' by humans... they need more-talented humans. – Tetsujin Jan 22 '19 at 20:58
  • Claiming that Note Performer "reads" music is a bold statement, but that's the general idea. On the other hand, human musicians are able to perform written notation in a very mechanical and ugly way! ;) Maybe the example should be listened to like, "could an actual orchestra play like this". If they have a slightly immature or mechanical approach, and would need better conducting? Maybe, because the virtual orchestra plays in tune and in time and with consistent tone, you expect top-level artistic interpretation as well? ;) – piiperi Reinstate Monica Jan 22 '19 at 21:31
  • @piiperi, I don't know who your comment is directed to. But, I did listen to Note Performer with that very thought 'could an actual orchestra play like this?' If the audio file is a straight reading of the score - no human tweaking after - I think it's very impressive. Not 'unlistenable' like Tetsujin said. – Michael Curtis Jan 23 '19 at 13:51

Not the answer you're looking for? Browse other questions tagged or ask your own question.