Digital Orchestra...

[perfect_pitch]

I'm currently learning Rachmaninoff's 3rd Piano Concerto, and aim to record it with a second piano in October. While I would LOVE to do it with an orchestra, Perth is a cultural cess-pit where music takes a BIG backseat due to sport.

Anyway, I have a radical though. Is it possible to record the piano solo part, and have a digital orchestra play the instrumental parts??? Obviously the piano part would be a WAVE or MP3 file, but how would one try and wrap a virtual orchestra around a piano solo???

What software would one need? Is it do-able? Is it worth the effort?

[quantum]

You are talking about using sampled instruments as a sound source then programming the parts into software, correct?

I think the main challenge would be flexibility of the flow of music.  The programmed part can't respond in real time to the piano solo part.  This approach of virtual orchestra + live performer works acceptably in pop genres where tempo is mostly defined by a tick track, and such rigid conformance to beat forms a characteristic of the music.  This is obviously not the case it your situation. 

You would need to program that flexibility of flow into the virtual orchestra, then also match that flow when performing the piano part.  It would be like following a script for expression and tempo. 

I would recommend reading up on film score and multi track recording techniques.  You would be borrowing concepts from these disciplines. 

As for software, the two main options for note entry would be: notation software, or digital audio workstation (DAW).  Enter or MIDI record the individual parts into the software. 

If you entered your parts as notation, enter your desired expression, tempo changes, etc.  Unless you like the AI expression that software gives you, you will need to program virtual musicality into the virtual part. 

If you recorded your parts as a MIDI performance, you have the ability to express the music yourself and have that captured in the recording.

The final stages of assembling the parts, and polishing up the sound will likely be done in a DAW.  You will need a DAW that does both MIDI and audio, some only do one or the other.

Do small scale experiments first on easy music.  Short pieces, say 2-3 mins, but with similar expressive / interpretive demands as the concerto.  Work through the project to its completion, meaning, produce a final recording with piano + virtual orchestra.  See if it is worth the time and effort needed.

[perfect_pitch]

Okay... some food for thought. I want to mull this over while I'm still practicing. I might play about with MIDI in Movement 2 as there are many small snippets of Orchestra parts that recede to Piano solo sections.

Out of curiosity - what the heck is a DAW???

[quantum]

DAW = Digital Audio Workstation

Examples of commercial DAW:
Cakewalk, Pro Tools, Logic Pro, Reaper

Examples of Open Source:
Ardour, LMMS, Rosegarden, Traverso DAW


If you don't know where to start, here are my recommendations:

Free:
Audacity - mostly just audio, MIDI playback only.
LMMS - MIDI focused, workflow is a bit different.
Cakewalk - commercial, full featured, and free.  Windows only.
GarageBand - The scaled down version of Logic.  Mac only.

Freemium:
Ardour - Linux, Mac, and Windows.  The question in using Ardour is: how much is support worth to you.  Try it free with a Live USB of either Ubuntu Studio or AV Linux.

Commercial:
Reaper - low in price, high in features.  There are Pro Tools users that talk about switching to Reaper, that's how good it is. 

[perfect_pitch]

Okay - a lot to look at, and get my head around. I'll look into this.

[j_tour]

I agree that's a lot to look at, and thanks to quantum for giving an up-to-date précis.

The last time I had set up a DAW, it was under Cubase SX, and that was more for editing band performances, and as one of many hosts for softsynths which, at the time, were barely adequate for even Hammond Organ+Leslie speaker, or Fender Rhodes, or other basic keyboard sounds.  At about that time, it seemed composers and arrangers I knew seemed to favor Sibelius's various products for reproducing orchestral sounds from a score or via MIDI.  I think the world has moved on a bit.

I am curious if the OP wants a backing track, or, as in the initial plan (as it seemed to me), to have a second pianist provide the accompaniment in real time on an acoustic piano.

I don't think it's unrealistic to hire a second pianist, or, really, a keyboardist in this case, to reproduce the orchestral score in real time.

You might have to pay them extra to get good at using the various MIDI functions (on, no doubt, a machine capable of dividing the keyboard into different zones, and different sets of discrete groups of "instruments," which should be well within the capabilities of even a modest controller keyboard).

It could be done, though, and certainly has been, in real time.  At least with an adept accompanist.

I wouldn't think a backing track would be all that effective, or give the best display of your interpretive efforts.

ETA Yes, it's possible, and AFAIK, it's always been possible to indicate to the backing track little changes in tempo and dynamics.  It could be done via (yet another) pedal, or a breath controller, but I don't think that's reasonable.   For example, I know a few jazz guitarists of good reputation who use backing tracks, recorded by themselves, but they have at least half-dozen pedals set up (not all the backing tracks:  you know guitar players, they like their effects!), but they seem to manage.

Like you, the people I know who play live with backting tracks have their set lists figured out, generally, use various pedals (or even one single pedal with some lateral motion, like in those old "home organs," to control at most a few aspects of the backing accompaniment.

I don't think that's what you want.

Then again they only got six strings, and unless they have 1958 Les Paul sunburst, they don't have sustain, either. 

If it's not a live performance, I'd just play to a click (more or less) and add the overdubs of orchestra in post. 

If live, any pianist could pick up the rudiments of doing the orchestra, pretty sure.  Given some practice. 

But you'd have to pay them more and there's a learning curve that's steep, playing different zones on the KB, and swapping the right patches at the right time.  Those kind of people aren't so much piano players, at a given moment, however much they can play, but they're good at doing that kind of stuff.

[perfect_pitch]

I'm happy to go with the backing track, and as a piano player try and record the piano solo first, before replicating the violin parts, horns etc...

Being in Perth, looks like the odds of me getting to play this with an orchestra are almost zero - most orchestras don't want to take on something as monumental as the Rach 3, and some won't work with soloists (instead doing film scores to movies, playing with Dance companies etc...)

How good are VST's??? Are they good enough to be used in movies? The piano one I have is rather decent (not perfect, but decent) that one could ALMOST be tricked into thinking it's a piano.

[klavieronin]

How good are VST's???

The best VST's come with and equivalent price and they can take a LOT of learning. And even then, for something like the Rach 3, you'd be looking at hundred's of hours to get it sounding realistic.

Have you considered something like this?

https://tonic-chord.com/rachmaninoff-piano-concerto-no-3-in-d-minor-op-30-accompaniment/

[j_tour]

The best VST's come with and equivalent price and they can take a LOT of learning. And even then, for something like the Rach 3, you'd be looking at hundred's of hours to get it sounding realistic.

Have you considered something like this?

https://tonic-chord.com/rachmaninoff-piano-concerto-no-3-in-d-minor-op-30-accompaniment/

Oh, with respect, I don't think it's that bad.

Maybe it won't fool an audiophile.

But, compare today's softsnyths to those even ten years ago.

Cats play live jobs using some little tablets pretty much every day. 

"Gigabyte Samples"? 

It is to laugh:  modeling even in the past ten years has become not a fad, but a reality. 

[j_tour]

I'm happy to go with the backing track, and as a piano player try and record the piano solo first, before replicating the violin parts, horns etc...

Being in Perth, looks like the odds of me getting to play this with an orchestra are almost zero - most orchestras don't want to take on something as monumental as the Rach 3, and some won't work with soloists (instead doing film scores to movies, playing with Dance companies etc...)

How good are VST's??? Are they good enough to be used in movies? The piano one I have is rather decent (not perfect, but decent) that one could ALMOST be tricked into thinking it's a piano.

I follow now:  you don't need live performance, in real time.

Yeah, I'd go with what others have said above. 

The best will cost the most (sampling libraries, and a platform aka DAW to host them), but, I think you'd be surprised at how good older tech can sound. 

Just watch any movie from 1985-2015 or so:  compared to now, it was maybe composed and "performed" on an Amiga or something. 

What I would do (and, no, I don't have the chops for Rach3Concert) is play and record the solo (i.e., piano) part, and then, yeah, just add in in post the orchestral parts. 

[klavieronin]

Oh, with respect, I don't think it's that bad.

Maybe it won't fool an audiophile.

But, compare today's softsnyths to those even ten years ago.

Cats play live jobs using some little tablets pretty much every day. 

"Gigabyte Samples"? 

It is to laugh:  modeling even in the past ten years has become not a fad, but a reality.

Don't get me wrong, VST orchestras are quite good. I use Vienna Symphonic Library myself and am very happy with the results, but it can take a long time, especially if you don't have much experience with that sort of thing.

I don't know of any physically modelled orchestral libraries though.

[j_tour]

Don't get me wrong, VST orchestras are quite good. I use Vienna Symphonic Library myself and am very happy with the results, but it can take a long time, especially if you don't have much experience with that sort of thing.

I don't know of any physically modelled orchestral libraries though.

I absolutely defer to you:  at best I'm talking about fifteen or twenty years old tech with which I had only a tiny bit of experience on, if that.

No, I know next to nothing about today's tech, except from reading a few magazines while...resting...a few minutes every morning.

My only point was, if I had to guess, pretty much anything would sound just fine.  It wouldn't probably sound cheesy, is all. 

[j_tour]

Here's a bit of a question I was just thinking about.

Assuming p_p lays down the piano audio, just how much time is it going to take the average pianist to add the orchestral parts, assuming a decent library of orchestral instruments and some basic familiarity with how to control playback using a keyboard?

I think that's the easiest way, playing into a keyboard via MIDI into the DAW. 

Otherwise, you know, it'd be entering things in the so-called "piano roll" format, or at best step-entering things, also using the keyboard.

I'm thinking that would take absolutely forever.

I know people do it, you know, people who do film scores or composers and arrangers of various stripes.

I don't want to sound discouraging, but from people who have experience doing this, is this really even possible for a novice at using softsynths and DAWs in a scale of reasonable time, like, say, under six months, say?

This workflow is seeming pretty daunting.

Maybe it's better to lay the orchestral parts down first, rather than have to make many, many adjustments to the orchestra in post.

Granted, that's going to confine the OP to whatever choices he makes about tempo for the orchestra, but it seems to me that would take far less time.

And even perhaps more importantly, the project would get finished, instead of being in a state of dire morass of endlessly fiddling with the orchestral voices.

I have no doubt that a competent pianist, which I'm sure p_p is, will find it easier to "follow" a prerecorded track, rather than trying to essentially program the orchestra after the fact.

So, that's the best way I can figure it out, to actually do the project from soup to nuts without spending ages and ages fiddling with the equipment.

And then!  We can talk about in-ear monitors and how to avoid any artifacts from bleed-through of the microphones!

ETA Oh.  And then the recording has to be mixed, so you're going to want stems of, say, at least violins, cellos, etc.  Mastering shouldn't be that hard once it's mixed, just even things out a bit.

Oh, man, this is starting to sound like a bigger project than I'd have thought.

Well, you're a brave guy, p_p.  I think you can do it, but you certainly need better advice than I'm able to give.

EETA Perth isn't a complete backwater.  Even just recruiting some university students to do, say, strings alone, live, and rent some studio time with a tracking engineer, that might give a solid foundation to which you can add other orchestral elements.  I don't know.  It's a tough problem.

You could do a "gofundme" project to rent studio time. 

That's a tough situation.

[klavieronin]

I honestly don't think it would be worth doing it yourself. For starters, a truely decent digital orchestra would set you back over $1000 or more, a semi decent one will be at least a few hundred dollars (but of course there's no guarantee it would have all the articulations you would need). Then there's the time it would take to learn the program, play in the parts, choose all the right techniques and articulations, set the dynamics and timing, fix any mistakes made during your playing of each part…

If it were me, I would just buy a pre-recorded orchestral part, played by a real orchestra, like the one I linked to earlier, and use the time I would have spent creating the orchestral part for practicing the solo part. Cheaper, easier, quicker, better results, what more could you ask for?

[j_tour]

I honestly don't think it would be worth doing it yourself. For starters, a truely decent digital orchestra would set you back over $1000 or more, a semi decent one will be at least a few hundred dollars (but of course there's no guarantee it would have all the articulations you would need). Then there's the time it would take to learn the program, play in the parts, choose all the right techniques and articulations, set the dynamics and timing, fix any mistakes made during your playing of each part…

If it were me, I would just buy a pre-recorded orchestral part, played by a real orchestra, like the one I linked to earlier, and use the time I would have spent creating the orchestral part for practicing the solo part. Cheaper, easier, quicker, better results, what more could you ask for?

Thanks.  I didn't actually clock that your link was to a ready-made orchestral part. 

Certainly, that would work, and gets rid of having to fiddle endlessly, to to speak, with the orchestral backing.

How do you think the best way to proceed for recording the piano should go? 

The only way I can think of is in-ear monitors (or, really,  just bog-standard earbuds, or whatever headphones one has handy) and just do the acoustic piano tracks using whatever microphones one would normally use on the acoustic piano.

It'll still take a little experimenting to get the right mix into the headphones, and then some work to make the end result sound mixed properly.

But, yeah, instead of six months of pulling out one's hair trying to recreate the wheel, that could be a handful of weekends pulling some late nights.

Much easier.

[quantum]

How do you think the best way to proceed for recording the piano should go? 

The  process used in multitrack recordings would be my pick.  Using in-ear monitors or closed back headphones.  Setup the DAW to have the orchestral tracks playback while recording the piano part to it's own track. 

I would mix the orchestra parts first, to make sure they all balance with each other and make musical sense.  The orchestra / piano mix does not have to be exact at the time of the recording.   The piano mix would be fine tuned once the complete piano part  has been recorded. 

Actually, I'm really interested in how this project develops.  Using a backing track of an actual orchestra is obviously less work, but it might leave a disconnect in regard to creative ideas.  The idea of being involved in moulding the orchestral parts, has the potential to create a very cohesive artistic idea.  The ability to shape not only the piano solo part, but to form the interpretation as a whole from the perspective of all instruments involved. 


Another small scale experiment, would be to play the orchestral part on the piano, then add the piano solo part as another track in the DAW.  This removes the complexity of dealing with virtual instruments.  One could focus in working with two tracks in the DAW.  Following that, one could expand to say separating the orchestral part into a few independently recorded tracks, for example: strings, winds, brass.  One would be learning the actual parts, and this work could be carried forward to the full scale recording where you would move over to playing a MIDI keyboard and virtual instruments. 

[j_tour]

The  process used in multitrack recordings would be my pick.  Using in-ear monitors or closed back headphones.  Setup the DAW to have the orchestral tracks playback while recording the piano part to it's own track.

Exactly, but I suspect it's even itself a bit of an unusual experience playing with cans on (or in) the ear for an acoustic player used to hearing things live.

I guess it would have to be a learning experience to work within a studio.  I'm not claiming to be a master studio player, nor even a good one, but it did take me a while to get used to using the various monitors and even talkback, but it's good experience.

Actually, I'm really interested in how this project develops.

Me too.  At first it seemed relatively simple:  map out some zones on the controller, add some expressive details after the fact, but the more I think about, the more I think perfect_pitch has got quite a little project on his hands!

I'm also really interested to know how how it turns out:  at first and now umpteenth glance it really seems easier and cheaper to buy a plane ticket to a smaller central European country and hire an orchestra and do it live, but I'm really curious how it works out.