Can You Tell The Difference Between A Real Orchestra And A Synthesized One?

orchestra

Synthesis has changed a lot since the days of Switched On Bach – and it’s getting to the point that recordings of virtual instruments, and even virtual orchestras, are nearly indistinguishable from the “real” thing.

Of course, this begs the question – what is the real thing?

How close are virtual orchestras coming to traditional orchestras?

You be the judge.

I’ve embedded an example below that contains short sections from three symphonies, by Beethoven & Schumann. See if you can identify the one that is virtual – and if you can articulate why you think it sounds “virtual”.

Image: jordanfischer

29 thoughts on “Can You Tell The Difference Between A Real Orchestra And A Synthesized One?

  1. I have to say, it is really hard to tell. I agree that the second one sounds "virtual", but only in comparison to the other two. The second one doesn't have the slow reverberation and deep bassy lushness that I would expect of a full orchestra. But, of course, that can be done electronically as easily as synthesis. When will we find out the answer? I'm expecting you to come back and say that all three were synthesized to just BLOW OUR MINDS.

  2. The second one is virtual – it's good, but there's a slight "machine gun" effect on the strings as The ShaggyFreak mentioned, also the timing is a bit metronomical – as with all these things the virtual stuff sounds good , until you hear it next to the real thing. But then is the point of virtual orchestras really to replace the real thing or are they a compromise to be used when (as is often the case) the budget is too small or to mockup an arrangement in order for producers to hear what you're going to do? Surely no-one really believes a virtual orchestra will really replace real ones if someone has the option (and the budget) to use the real thing?
    As a side issue did Carlos really intend the sounds in the original S.O.B. to be imitations of the real thing? Judging by some of the interviews that I've read with her (and given how much she says she hates S+S based instruments for example) I'm not altogether sure.

  3. The second one is virtual – it's good, but there's a slight "machine gun" effect on the strings as The ShaggyFreak mentioned, also the timing is a bit metronomical – as with all these things the virtual stuff sounds good , until you hear it next to the real thing. But then is the point of virtual orchestras really to replace the real thing or are they a compromise to be used when (as is often the case) the budget is too small or to mockup an arrangement in order for producers to hear what you're going to do? Surely no-one really believes a virtual orchestra will really replace real ones if someone has the option (and the budget) to use the real thing?

  4. I have to go with the majority so far and say the second one. And I did make sure to listen and decide before reading posts so as to not color my opinion. Without getting some decent volume to give you more specifics (my office is in the school library – loud music is verboten) it simply feels much more sterile than the other two examples.

  5. Section 2 sounds the least pleasing to me.
    The dynamics seemed very 'hyped' on first listen and it didn't feel like 'one big instrument' in the way that the other two do.
    After reading the comments and re-listening, the harsh attack and machine-gunning was also evident to me.

  6. Second one: strings sound more synthetic and the dynamics are unnatural (too fast cresc/decr). The instruments aren't as 'together' as they are in the other two, but maybe that could be solved by mixing and configuring the reverb.

  7. The second one sounds like many take of individual instruments mixed (copy-pasted?) togoether. The first sounds like a big mess of too far miked big room analog take. A little too boomy for a classical experience, but I have the worst headphones on earth. The third one sounds somewhere between the two. Perhaps too clean for me, but not as harsh as the second, wich is definitively virtual. How couldn't you hear the difference ?

  8. What exactly is meant by "synthesized" and "virtual" here? Are they based on samples or are they actually synthesized from oscillators?

  9. which one is virtual? so the question is either of them has been naturally played and the other one is played – or – written using recorded samples of real orchestral instruments?

    so, virtual is definitely the second one. It's cold and clums.
    If I'm wrong , I'm very impressed!

  10. The second one is virtual. Just look at the way its envelope leaps up and down! No real orchestra can play dynamics that well.

    I still like it, though. Way to put grumbly real musicians out of work!

  11. None of them are real. They're all coming out of speakers!

    Since I actually made the second example, I should say that the purpose was not to imitate a real orchestra and sneak by un-detected. Rather, the point was to play this piece how I wanted to play it … with the shaping and balance between the moving notes that I felt best fit with the other elements of the score.

    It's not so important to me that it sound exactly like a real orchestra — just like it was not important to Glenn Gould that his piano sound just like a clavichord from Bach's day. The important thing is the musical expression. Can I make it musically compelling?

    To answer that, it may be best to hear the full version of No. 2, here it is:
    http://www.fauxharmonic.com/music/beethoven_symph

    – Paul Henry Smith

  12. Paul – thanks for your comments.

    Your project is fascinating – because it starts to raise questions about what a "real" orchestra is.

    Is it a bunch of people using classical instrument technology, is it a sound, is it the process of making music out of old scores, or is it something else?

    For the record:

    The first example is from Classical Masterpieces | Deutsche Welle's podcast of Beethoven's Symphony No. 3, the "Eroica".

    The second is from The Fauxharmonic Orchestra's performance of Beethoven's Symphony No. 2, in D, III – Scherzo.

    The third example comes from Classical Masterpieces | Deutsche Welle's podcast of Robert Schumann’s Symphony No. 3.

  13. I listened to the track on some desktop usb speakers at work, and I honestly have to say that I don't hear anything that would make me think that it's fake.

    I guess people are always going to prefer the "real thing" but I also think that we are gonna get more and more used to the sound of the digital realm. especially with today's quality VI libraries.

    My point being that if a artist uses a VI over the real thing, the average consumer does not hear the difference. Only a trained ear on decent studio sound will really be able to tell the differences.
    Nevermind the fact that if it was not a fake, maybe it's just some person's unique editing & mixing style.
    I see my friends listening to one album and then the other of various bands and if the recording sounds good, then a person just accepts that as that band's "sound".

    SoI for one do not at all mind using this quality Virtual Instruments in a song.

  14. Of the three, I have to agree and say that the second one is probably synths. The third sounds like it could be synthesized as well, although it has better production than the second one.

  15. Clearly the second one is virtual. It is very nice though. The strings sound a bit thin and mechanical. Especially the quick notes sound too even as if they were played on a keyboard rather then on a real violin. Violins are notirously difficult to imitate because a real violins sound changes at every pitch level AND during the stroke whereas the imitation always uses some sort of a loop which gives the fake violin a repetitive quality. The parts with the entire orchestra sound more convincing although the attack sounds slightly unreal.

  16. I said, the second, but because the woodwinds sound artificial to me, a bit like the woodwinds used by the “Sibelius” program.

Leave a Reply

Your email address will not be published. Required fields are marked *