Chopin Prelude Orchestration (audio)

Beautiful!

Speaking of the audio, what software and hardware was used?

Finale 2011, audio produced directly in Finale, using various plugins, without exporting anything to Cubase or similar.

Libraries used: Berlin Woodwinds for woodwinds, Cinebrass for brass, East West for percussion and LASS 2.0 for strings.

Mastering plugins: Ozone 5 for EQ, Harmonic exciter, Compressor (just a touch...), Stereo image; Reverberate (rich hall) to create the ambience.

The difficult thing is that the various libraries used are recorded in different environments and LASS is dry, so on the strings I had to apply a Kontakt 5 reverb preset and then the general reverb.

The total weight of the software and various plugins is around 3 GB, so even with a 32-bit system it's doable, but I optimized the Kontakt buffer to the maximum so that the RAM used in loading the instruments would decrease; otherwise, you exceed 4 GB. This is possible by having an SSD (solid state drive), which is very fast at reading.

It goes without saying that I made countless attempts and comparisons with demos and tracks from real orchestras available online; I also found some very interesting virtual orchestration lessons and it seemed to me to understand that a lot depends on the ear and the experience gained over the years, while those libraries are—the best ones out there are those (EW, LASS, Vienna, Cinematic Strings, Cinestrings, Berlin, 8Dio)—there are no others; when we listen to a demo that sounds monstrous, it's because there is a sound engineer who knows how to bring out the best from every single library.

I own all these libraries except Vienna, so I can say that my choice for strings fell on LASS for timbral reasons and RAM usage; Cinematic Strings are too dark for me, 8Dio Adagio presents quite evident sampling errors, I don't have the Hollywood series from EW but the previous one, which in terms of quality is inferior to LASS, but Hollywood is extremely heavy.

The only problem with LASS is that they are dry, so you have to tinker to create an ambience before even starting to compose.

geppino12575 wrote:

i also found some very interesting actual virtual orchestration lessons

It would be interesting if you had some links, but the question is: is there compatibility between traditional and virtual orchestration? I mean, it's not that the piece sounds "divine" on the PC and then doesn't work with the orchestra?

Since you have experience in both different sectors, what do you say?

Unfortunately, I don't have much experience in this regard, because it has rarely happened to me to have my arrangements performed by a real orchestra, but live they sounded decidedly better, in an environment with beautiful natural reverb.

In general, everyone agrees that to make virtual libraries sound good, you must first know how to write for real instruments and be well-versed in the various techniques and articulations of all instruments, even if you play them yourself on a MIDI keyboard.

There is also the difficulty that the virtual composer (let's call them that) must also act as the sound engineer, studying the various libraries carefully to put them together so that it seems they are all in the same environment, managing their stereophonic positioning, etc.; in short, a series of issues that those who just write usually do not consider, although there are examples in history where great masters calibrated their works (especially for organ) according to the environment in which they were performed.

It is also true, to support what you are saying, that people often think a passage sounds bad because it is MIDI, but in reality, it sounds bad simply because it is poorly orchestrated; this is why many recommend, before investing in software, to invest in studying composition and orchestration, in listening—especially and above all live—and in collaborating with real instrumentalists.

So, just out of curiosity, would you be able to estimate the working hours for this audio project?

From when you put down the first notes in Finale, to when you finished the final master.

Thank you

A couple of months, with some breaks because of the holidays in between. Obviously I haven't worked on it every day, and in compatibility with work commitments and some health problems I've had during this period. For the audio, I am proceeding by trial and error, whereas for the writing I am quite sure of what I am doing; that is why it took me about ten days to write, but the timing is stretching out enormously for the audio.

I confirm what Geppino said.

The head moves very fast and the hand moves quite fast, but not as fast as the head.

Then the draft needs to be reviewed for errors, and finally, one starts with the audio.

But there is a lot of tinkering to do, and in this case, Geppino gave us a very useful tip to avoid exporting to Cubase or similar.

Writing the piece at the tempo level takes 30% of the time.

For the audio, it takes at least 60%, and the remaining 10% is for putting together a final score playable with the parts, which usually doesn't take much time if the score is set up well.

Of course, those who write directly on the computer save a bit of time, but personally, I can't do it. I need to have paper.

Great job, I feel it's just a little too reverberated, but it's only an impression given by my personal taste... I wouldn't know how to justify it ..

Compliments Geppino!

It actually seems a bit too reverberated to me as well, I will try to remove something.

Regarding the fact of working in Finale or exporting to Cubase and similar, I must say that no one recommends working in Finale or Sibelius and indeed editing is more awkward, even if substantially with a good sound card one can record at minimal latencies even on Finale as if it were a sequencer (the writing becomes quantized) while preserving the integrity of the human factor, let's call it that.

I am tied to Finale because I am tied to the score and because I only know how to use Cubase in theory; in practice, I feel more secure starting from the sheet music, but the vast majority do the opposite: they start from the sound and therefore from their own MIDI performance and then perhaps try to derive the parts if they were needed for eventual real musicians. Furthermore, virtual libraries are created mostly for this second approach, i.e., starting from the sound, not for notation programs and therefore from the notes.

Everyone chooses the approach that is most congenial to them; at the moment I prefer to start from the score and implement that whole series of measures to humanize the performance.

opinion of a non-expert listener on mastering & co

The level is certainly very high; if I had to find something to point out, but just out of perfectionism (!!), I would say that, listening through headphones, it gives me the impression that the orchestra is a bit far away or anyway slightly lacking in sonic impact.

Also, the very pronounced spatialization of the instruments, while beautiful on one hand because you can distinctly hear everything, risks giving the audio a character that is too "dry" (settle for "dry," I don't know how to explain it better eheh)

geppino12575 wrote:

The total weight of the software and various plugins is around 3 GB, so even with a 32-bit system it's doable, but I have optimized the Kontakt buffer to the maximum so that the RAM used in loading instruments is decreased, otherwise it exceeds 4 GB.

How do you do it?

in effect my PC has 4 GB but it struggles immensely to load more than one library at a time, so in the end I always have to give up before even starting....

geppino12575 wrote:

Finale 2011, audio done directly in Finale, using various plugins, without exporting anything to Cubase or similar.

Libraries used: Berlin Woodwinds for woodwinds, Cinebrass for brass, East West for percussion and LASS 2.0 for strings.

Mastering plugins: Ozone 5 for EQ, Harmonic exciter, Compressor (just a touch...), Stereo image; Reverberate (rich hall) to create the ambience.

The difficult thing is that the various libraries used are recorded in different environments and LASS is dry, so on the strings I had to apply a Kontakt 5 reverb preset and then the general reverb.

The total weight of the software and various plugins is around 3 GB, so even with a 32-bit system it's doable, but I optimized the Kontakt buffer to the maximum so that the RAM used in loading the instruments would decrease; otherwise, you exceed 4 GB. This is possible by having an SSD (solid state drive), which is very fast at reading.

It goes without saying that I made countless attempts and comparisons with demos and tracks from real orchestras found online; I also found some truly interesting virtual orchestration lessons and it seemed to me to understand that a lot depends on the ear and the experience gained over the years, while the libraries are what they are—the best ones around are those (EW, LASS, Vienna, Cinematic Strings, Cinestrings, Berlin, 8Dio), there aren't any others; when we listen to a demo that sounds monstrous, it's because there is a sound engineer who knows how to bring out the best from every single library.

I own all these libraries except Vienna, so I can say that my choice for strings fell on LASS for timbral reasons and RAM usage; Cinematic Strings are too dark for me, 8Dio Adagio presents quite evident sampling errors, I don't have the Hollywood series from EW but the previous one, which in terms of quality is inferior to LASS, but Hollywood is extremely heavy.

The only problem with LASS is that they are dry, so you have to tinker to create an ambience even before you start composing.

Thank you for the detailed explanation

Listened to in headphones, great, I like how it sounds.

I have serious problems with these libraries, I also own East West, LASS etc....

but I don't like the result I reach every time...

In practice, I always discard everything with much frustration....

I don't know if it depends on the type of orchestration I do, or something else.

in general, it is the sound that does not satisfy me, I find it synthetic and lacking expression..

And I am not even an amateur, since besides being a musician, I have always

been fascinated by the technical aspect related to sound, so for 25 years I have been doing both,

both sound technician and composer :-(

I went to a sound engineer, who told me that I did a good job, except that he heard a bit too much "noise" resulting from the fact that reverb was applied to every instance of Kontakt (so 5 reverbs) which all go into the master bus and therefore add up, creating too much ambience. He advised me to work with dry samples, so wherever it is possible to choose the microphone placement, always choose close mic and then apply two reverbs to the master bus, one long and one short, and give more or less of these two reverbs to each section, in order to position the instruments more forward or further back depending on their actual positions in the orchestra. For example, the strings are closer to the listener, so they will have 80% close reverb and 20% far (the % are indicative...), the woodwinds are intermediate, so a bit more far reverb and less close reverb compared to the strings, the brass even further away, and the percussion even further still.

Furthermore, he told me that the further away the real instruments are, the more you need to cut the high frequencies, because the sound has to travel a longer path to reach the listener and high frequencies are absorbed by the performers' heads; therefore, it is good practice to apply some EQ and cut some highs, just as to avoid unwanted noise, it is advisable to cut some of the frequencies not used by the instruments.

He told me that my workflow is maddening, that working in Cubase is much more convenient, but all in all, if these are the results, I can consider myself moderately satisfied.

I hope this information is useful to anyone else who finds themselves like me, working on the score rather than on the tracks.

I am so sorry, but unless magic happens in the hall where I am, what he told you

seems too singular to me.

A concert hall that changes its type of reverberation depending on the sections?

The reverb, meaning the decay of reflections, will always be the same; the hall cannot switch from long to short,

it is scientifically impossible.

There can only be one reverb; it is only through dosage, meaning the amount of reverb, that you move instruments further away or closer :-)

If you then want to venture into exotic mixing techniques, I could go along with that, but I don't think that is the case :-)

In the event that what can change is only the pre-delay, or what is defined as Early Reflection,

in other words, the first reflections; so to summarize, the reverb length remains always the same,

but having an instrument closer, the first reflections will reach my ear sooner, therefore the parameter to act upon will be the Pre Delay, usually expressed in milliseconds in normal plugins.

It has been estimated that the best halls in the world all have a Pre Delay of no less than 21 milliseconds :-)

Micrologus wrote:

I have serious problems with these libraries, I also own East West, LASS etc....

but I don't like the result I reach every time...

In practice, I always discard everything with great frustration....

I don't know if it depends on the type of orchestration I do, or something else.

in general, it is the sound that does not satisfy me; I find it synthetic and lacking expression..

 

If this is the flaw you are finding, it probably depends on the fact that you work little on the parameters that make samples "human," which depend on MIDI messages, regardless of whether they are entered into a notation program score or a sequencer. These parameters are: velocity (volume/dynamics), variety of articulations, note duration (which in real performances never corresponds to what is written), tempo variations (for expressive purposes), small vertical offsets, legato and staccato, crescendo, etc. In practice, it is about "dequantizing" every aspect of the sound, possibly note by note, and doing so with the same criteria that would guide a real performer.

If I can give you some advice, try taking a piece with simple writing, perhaps just for piano (for example, an adagio from a Mozart sonata) and try to meticulously copy the recorded interpretation provided by a pianist; only then will you realize how important it is to intervene and where, using such procedures (these are all MIDI messages, even if introduced via "tools" or various windows), and what the ways to do so are in relation to the program being used. Once you have succeeded, you will know how to extend the same procedure to other samples.

This is an aspect of editing that has nothing to do with composition, nor with mastering and audio quality, but it is fundamental to making samples sound credible, especially when they are of professional quality (this is because, even if it may seem strange, less "ambitious" samples require less attention to sound more or less expressive).

I invite you to notice that a flesh-and-blood performer can be recognized even when playing a keyboard with poor sounds; therefore, the fact that you care about the sound and have professional samples probably has nothing to do with the fact that you derive from them a performance that tastes "mechanical."

Bianca wrote:

How do you do it?

in effect my PC has 4 GB but it struggles immensely to load more than one library at a time, so in the end I always have to give up before even starting....

I had the same problem as you for a long time and I solved it with a very modest expense, upgrading the computer to 16 GB. If your PC allows it, I highly recommend it. Otherwise, you have to settle for libraries that are not top-tier, but still decent (Garritan, for example, runs well even on 4 GB). Then one would need to look at some other characteristics, such as the hard drive speed.

CromaDiBrera wrote:

I had the same problem as you for a long time and I solved it with a very modest expense, upgrading the computer to 16 GB. If your PC allows it, I highly recommend it. Otherwise, you have to settle for libraries that are not top-tier, but still decent (Garritan, for example, runs well even on 4 GB). Then one would need to look at other characteristics, such as the hard drive speed.

I also believe that 4 GB is too little for decent libraries. Moreover, I use a laptop that is also on its last legs and I am thinking about whether it's worth upgrading the memory or replacing the entire notebook directly. I tried to intervene in every way with the Kontakt buffer but without any appreciable effect. The fact remains that doing playbacks under these conditions is a torture and it's just a waste of time...