It’s not the same thing.
In any case, writing with a virtual orchestra as a hobby is fine when understanding that the orchestra will never be replicated.
Impressive renderings, but most of all impressively beautiful pieces. As a former exchange student at Dalhousie, I’m personally grateful for this precious hommage to the Maritimes!
Paolo
Thank-you so much. I can quite honestly say that I really enjoyed finding ways of incorporating the four musical “anthems” from the Maritime provinces into that scherzo. This is the only movement of the 5-movement symphony that incorporates themes from elsewhere. Otherwise, the whole symphony shares two themes in various guises: “nature” and “man”.
That’s quite interesting! And a great job on putting the video together (whoever did it).
Ah! thank-you.
I made the video with my very VERY limited talents as a video editor and with the minimal access to images for the subject matter. (4 out of 5 movements have videos with accompanying images, while 1 movement has excerpts from the score, since I could never find appropriate images to go with it.)
there is a world of difference between the best real orchestras and the average amateur band. Instruments can sound totally different and there tends to be far less in the way of organic transitions. Don’t forget that sample libraries record actual musicians playing – it is not artificial at all. What makes it often seem to is the lack of musicality in the flow because of poor quality programming by the engineers but things are far better than they were a decade or so ago. I can say without the slightest hesitation that a good mockup of the top libraries sounds more like a “real orchestra” than a mediocre amateur orchestra badly conducted. Or perhaps more importantly, it certainly gives more pleasure.
Well, if the only purpose is to recreate classical scores, of course you will fail 90% of the time, a live symphony orchestra is the most difficult thing to create on a PC. But if the aim is to compose hybrid music, where it’s “allowed” to leave the classical orchestral form, then the result changes, as the aim is not to compose a realistic thing at all costs, the aim is simply to compose. I believe that in the future, thanks also to innovative software such as Note Performer, it will be easier and easier to get an “almost” real orchestra in your PC. For someone like me who composes from a PC and gives great importance to playback, this is excellent news.I hope that version 4.0 of Dorico also has a number of important updates regarding expression maps, integration with vst and playback control.
Giving pleasure to the composer who has no other means to realize his/her work, sure, most definitely. There are zero issues there.
Of course, the instruments recorded are live, but once the music leaves people and becomes a sample, something dies. I have recorded with a full 80 piece orchestra, 40 strings, and solo instruments. I also have top-level sample libraries.
Beyond the fact that due to how harmonics work, as such, no virtual orchestra can ever be 100% tuned correctly at one point, there is a sonic depth of field, something that I cannot actually explain, as to why the real thing sounds so, so much better, it’s fuller, has depth, and most importantly, it’s a unique print. It’s real, it’s live. People play it, not a processor,
Bad orchestras are a real problem and I completely agree with you.
Is anyone gonna grab the new Abbey Road Two - Iconic Strings? I’d be curious how it compares for string quartet/quintet stuff.
Not going to to that kind of comparison but I’m glad to read yours.
I recently asked my daughter to realize a version of the Müller Gloria (fka Mozart from 12th Mass) for my online choir. She used Garritan and it was ok but I was in for a fair amount of work (especially the tympani) so, to save time, reassigned to NP.
First, the tympani sounded exactly as I liked. Onto the strings. Many of the complaints about bowing articulations I’ve read are exactly what I like about NP when it came to 1/16 notes — sweet! Yes, I can spend hours tweaking out something a bit more refined but when I have weekly deadlines…why would I?
She was quite pleased with the final result and Dad knows that a $129 license for NotePerformer will be in her stocking this year.
So what? We’re working here.
So, NotePerformer doesn’t sound exactly like the real thing. Who cares? It doesn’t need to. NP’s value as a time saver cannot be overstated.
Besides, some (many? most?) of us are still working with lockdown ensembles and don’t get to just sit in front of a group of players in a live setting.
To me, these notes lack attack. The notes come in with a waaah. They should have a crisp start, and start at full volume. Not always crescendo to final volume for every note.
Are those renders ‘out of the box’? The staging is very good.
that’s exactly out of the box. I’ve done nothing to it to tweak in any way whatsoever.
as I said, this is a Finale file playing back the score as it will be sent to the publisher.
Agreed,
I was referring to a Jpeg with “detailed” instructions of a person with “amazingly finetuned ears” as of how to EQ and create orchestral strings to “perfection” which to me is pretty much impossible and akin to chasing unicorns.
I have zero problems with imperfect samples because that’s what they all ultimately are just that, imperfect and fake. But, hey, as long as they serve a purpose, then why not?
My reply was to a person who posted an animatronic AI robot and compared it to the evolution of the replication of the orchestra. That’s why I answered as per, “it’s not the same”, and I also could not get the correlation of a creepy robot imitating a human, and the replication of the orchestra…
Thanks for your feedback. Although the attacks of the Note Performer version are a bit too hard for my taste, I agree that for this piece somewhat harder attacks are better.
The problem you described was partly caused by a bug in Dorico resulting in a wrong playback of measured tremolos: it uses the length of the note representing the entire measured tremolo to choose the appropriate playing technique when an Expression Map uses note length conditions - and e.g. picks a playing technique corresponding to half notes instead of 16th, which is surely completely off. In the new version I have notated the measured tremolos explicitly to circumvent this and I hope this bug will be fixed in the next Dorico version. Moreover, so far the playback is a compromise we had to take since Dorico only offers five different versions for the playback of “natural” notes. The underlying Articulate Presets allow us to vary the attack continuously, and here is for example a different setting of the Expression Map with a bit harder attacks (and they could be chosen even harder)- still no programming:
Is this closer to what you were expecting?
As soon as Dorico allows us to use more of the many different attack version (hopefully soon in v4), the playback will definitely become more realistic in general.
I’m late to the party but I absolutely agree with @dan_kreider here.
I plan to purchase Abbey Road Two eventually. Unfortunately, I’m not sure when that will be…
hi. To be brutally honest, there’s still no real attack on the notes. No high frequency components in the attack portion. The dynamic is better at the start, so less wah, but it’s missing the crisp sound you get from bow changes. This makes it sound very synthey to my ear. Compare it to an actual recording and you will see what I mean.
even this one: Mozart: Eine kleine Nachtmusik - Concertgebouw Kamerorkest - Live Concert - HD - YouTube
p.s. measured trem was not what I was talking about either - I was thinking about the main initial statement in the 1st violins (and other strings), and then pretty much every note thereafter that wasn’t slurred into.
I was going to leave this interesting thread in peace but I must say, I am also somewhat disappointed by the @symphonic-riot rendering here. With no corrections at all (other than the tempo which is clearly too fast) I decided to do the beginning with the entry-level Synchronized Chamber Strings SE and my Expression Map. With five minutes work, I would correct the lack of attack on some sf’s and obviously write out the two stroke tremolos in full notation to start with. Whether this is an improvement, I’ll leave others to judge.