Welcome back, third and final episode of "How to Write a Soundtrack". Today we have a really content-packed episode, so get comfortable, because we'll be talking about many, many topics. Where do we start, Fabri? Let's start, of course, with the recording of the soundtrack which we promised right from the first episode; it was supposed to be the culmination of our journey.

We haven't pressed "Rec" yet. Before pressing "Rec," we need to put something that captures the sound and perhaps even some chairs for the musicians to sit on... Darn it, I hadn't thought of that. So let's start with a fantastic timelapse...

We're inside the auditorium at Largo Mahler in Milan, the sacred venue of the Milan Symphony Orchestra, which has performed many pieces of the soundtrack. Fabrizio set up a camera just to film the orchestra to make this timelapse. Here, you can see all the technicians busy at work... We can also see someone who looks a bit like a composer...

And someone who looks a bit like Vito Lo Re... All these people are here to set up the microphones because the amount of microphones and equipment on a project like this is huge. And we can also see the Decca Tree Goffredo Gibellini and Digital Records, with Gianluca Porelli and Giuseppe Corradino - who masterfully handled all the sound engineering of the soundtrack - loaded a nice van from Rome full of wonders, full of microphones, full of stands... headphones...

mobile control room, converters, and also these beautiful Neumann TLM170Rs that we find on the Decca Tree. In this case, Goffredo set them to omnidirectional mode. Why? Because in the auditorium the sound is truly beautiful and placing the microphone in omnidirectional mode allows you to capture the complete sound, including the nice reflections.

However, if you wanted to exclude - in a hall not particularly effective in terms of acoustics - this microphone also allows you to switch to cardioid mode and thus remove the bad reflections from the input. In this case, though, they were very nice. Later, we will also see how to properly process the sound from these microphones, because the sound from these microphones is like a RAW image from a beautiful camera. Yes, because even though they are wonderful instruments, like everything, you need to know how to use them and make the most of them.

A good microphone alone does not automatically give you a good recording. You need to extract all the wonderful information that the microphone has captured. We also made a backstage video as well. You also saw the rotational movement of the hand he made?

It was related to the bending of the horns and indicates the "descent" of the sound. Obviously, all the musicians have their own headphones through which they listen to the click and everything they’re not playing. But it's quite interesting that each stand can adjust the volume of the click or the base track. This is important: when you are making such a complex recording, every minute counts.

Unfortunately, a lot of time is often wasted: e. g. the trombones ask the conductor, perhaps, to lower the click in their headphones. Then the horns say, "For us, it's too loud instead..." In short, time is wasted, it may seem like a trivial thing but even putting numbers on each bar, or having a monitor that precisely shows where we are.

It may seem like you’re saving a little, but in the economy of a session, it's a lot! Let’s take an example of the session changeover time. We close the session of a track that we’ve finished recording and open a new one for a new track. How long does it take?

A minute? Let’s assume we record 20 tracks, how many minutes are we wasting? 20 minutes, which is huge because in 20 minutes you can record two tracks, at least. What are these boxes?..

That wonderful box contains the converters. They convert into 01 - so into a digital string - the electrical signals that come from the microphones. This is the microphone setup, It almost looks like a game of Battleship! On the right, we see the sections of the orchestra with the indications.

The numbers represent the channels. We had 32 channels. For the strings, we used KM84. The choice of the orchestra’s disposal is another element that guides the type of performance.

It can also be functional to the material that we’re recording. And in terms of the material we’re recording, you can decide on a compromise between recording different sections separately or having them all play together. This is a choice that you need to make based on what you are recording. Yes, obviously it's clear that the composer and especially the one who will mix the score must know well what the critical points are, and what needs to be highlighted.

Recording everything together has the advantage of greater uniformity, but many times it happens that maybe one individual instrument ruins everything. So it's better to record it separately to avoid problems. Or sometimes it's simply a matter of balance. It's not necessarily because they played their part wrong.

I would say that we can identify two main factors. The first factor is post-production. If, for example, I decide to do some treatment in post production on the brass, obviously, I must have recorded them separately. The second factor is that compromise, as you mentioned, related to the acceptability of a take.

The more people, more sections are playing the piece together, the higher the chance someone will make a mistake. So, an easy piece exposes the individual sections less to error. So, there’s a higher probability that we can use the interplay between all the musicians and the beauty of playing all together because there will be little chance of mistakes. With material that’s perhaps not so usual, written in a more difficult or particular way, we can run into the danger that in every take someone makes a mistake.

And then going forward a greater number of times and thus losing more time; it's better to quickly close a section and work with more separate takes on the one that's causing problems. Speaking of wasting time - or rather NOT wasting time - it is fundamental in a project with such a large ensemble, to make a production plan that is followed absolutely to the letter. Not all musicians play at the same time, and it's not necessary for all of them to stay there until the end. There might be a musician who plays only in 2 pieces out of the 20 we have.

So, in the production plan, it will be indicated when those two pieces are played in the rehearsal schedule, so that musician will be present exclusively at that point. For a musician to stay maybe one, two, or three hours doing nothing, is not useful. Generally, we start with the largest ensemble possible, also because starting at the beginning, the orchestra is fresher, and gradually the orchestra thins out until we get to the minimum necessary requested by the composer. Later, we'll see that we went from a full orchestra of 14-12-10-8-6 to a string quintet.

It depends on the more or less intimate character of a piece. In some cases, the thinning sound is more useful to describe a certain type of sensation or scene. Let's see at the treatment of the brass that we mentioned earlier; the track is called Lich Pulsar. Let's hear this "mean" tuba and trombones that go through a very intense post-production: in addition to the Tape Recorder that simulates the Studer, we also add the "Gorilla Drive," a compressor and a drive that greatly pushes the aggressiveness.

Obviously, treating only the brass forces you to record them separately. Here is the pitch bending. Here is the delay we used, a model delay with a somewhat vintage sound. A delay like this couldn't be used on the entire orchestra, so we apply it only to the instruments we’re interested in.

That's why I was talking about the importance of the composer knowing BEFORE what to record separately. Here again, we use the Gorilla Drive, which saturates and adds a bit of dirt to the sound; sound that also passes through the Teletronix emulator. Let's now see the difference between the sound from the microphone and the sound passed through the Avalon 737s. This is the sound recorded by the Decca Tree.

This, instead, is the sound equalized by the Avalon. If you have the chance, listen to this passage with headphones. What are these Avalons? This is not an emulation, these are the two real Avalon 737s that you have here in the studio.

All the material passes through here, from the panoramic mics and also the section spots. I don't have a Neve console like the one Goffredo has in Rome, which can process more channels simultaneously. I process through these analog outboard units the signals I want, group them, and then bring them into the mix, enhanced and refined. Another very important topic is the tuning of the orchestra.

Generally, it's done at 440, but sometimes someone prefers a brighter tuning at 441 or even 442, right? For example, when I go to Prague, I often find the orchestra tuned to 442. In this case, to match the tuning of the orchestra with all the digital systems precisely set at 440, or even with instruments previously recorded and tuned at 440, How do you do it in these cases? You use oversampling, that is, you record at a higher sampling rate proportional to the 2 extra Hertz.

That ratio gives us the sampling frequency with which we will record, and then, bringing it back to the correct for example, the usual 48,000, everything will sound "right". Speaking of "sounding right," let's not forget a small detail: the orchestra always sight-reads. The orchestra doesn’t have time to study the parts at home, so everything you hear is sight-read; often the first take is already good, or the second is already good, and the third must be perfect because, of course, you can’t rehearse a piece 20 times and record 20 takes, because there wouldn’t be time. In the Decca Tree, we captured all the natural reverb of the auditorium.

The transition from the sample, from pre-production, to the final mix often changes the approach to the overall sound and the mix because the reverb generated by the hall where we are playing and recording can offer a soundscape very different from what we imagined. For example, if we apply this Decca Tree, as is, in the mixing session where we were working with samples, most likely the reverb we had before on the samples will no longer work here. Another reverb will be needed because we have the natural one from the hall, and this difference in reverb also reflects on the character of the music itself because let's remember that reverb also gives a psychological connotation. We saw this during the episode about "American Beauty." The psychological characteristic may require, within the same soundtrack, some open moments, o maybe some epic moments, or closed moments, or some intimate moments, such as in this scene from the documentary where we hear the tercets of Dante about Paolo and Francesca.

In this case, it was the last piece of the session because here we have a string quintet, just 5 strings and a piano recorded here in the studio. Let's listen to the scene. The story of Paolo and Francesca is a "closed" story, between two lovers; it has a very tight, very intimate sphere, so do we tell it with 100 musicians and 100 brass instruments?... No, we tell it with something more intimate: a quintet.

So: we did our homework well, we recorded, the orchestra performed well, the sound engineer did an amazing job, we mixed... What are we missing? A small detail that’s not such a small detail though: the mastering. I don’t recall a project we worked on together where you didn't emphasize the importance of mastering.

Let’s be clear: there are different levels; sometimes a master is done at the end of the mixing process. In this case, we do a small limiting to ensure the levels, we compress a little to bring out the right sound and to work on equalizations that remove resonances. My mastering engineer – Andrea De Bernardi at Eleven Mastering Studio – shows us in this piece some of the moves he made to find the resonances and work selectively on specific frequency bands; compressing some bands more and freeing others... this work pulls the most we can out of our mixed soundtrack; it's a fine-tuning job that helps build that completeness of harmonics, sound, and solidity that you want to achieve, that makes you say "wow!" Of course, between the composer, the person who did the mix, and the mastering engineer, there must be unity of vision, otherwise you risk having something flattened that you actually wanted...

because he's an artist, he has his own vision; maybe that vision – not knowing your mix goals – might not align with your vision. Like everything, you communicate; you have a Ferrari at your disposal and you have to decide where to go. Andrea works entirely "in the box," meaning entirely via software. He has all the plugins and software he needs on his computer.

This allows him to reopen master sessions for songs that have been changed along the way or cut, or have had last-minute changes. With hardware, we would lose all the settings we used. Having everything "in the box," meaning entirely in the digital environment, allows you to operate later with the same identical result that you had previously sought and perhaps found. Before we say goodbye, why not watch the song over the film's images?

Well, we’ve been talking about it for 3 episodes... shall we watch something? This time we’ll say goodbye watching the images from "Mirabile Visione - Inferno" by Matteo Gagliardi with the final piece recorded, mastered, and synchronized. See you next time!