What's new

Recording a brass library: where to start?

bababooey

Active Member
Hello,

Im a 19 year old hobbyist composer and got a group of friends who are willing to record a brass library with me.

We got 3 trumpets, 3 tenor trombones, 2 bass trombones, 1 tuba, and 4 french horn players.

We already have a studio, we just need a plan.

None of us have the experience we actually need but we really want to do it, it should start small, with just the most important articulations, we can expand the library later if we want to.

I would like a few tips from someone who has some experience, at least in theory.

Since we are all musicians and nobody expects us to get paid, the budget question is out of the question, it's just time we need.

So my main question is the process for something like this, recording, editing, scripting, testing would be my guess, i would like tips on the individual steps.

We would need to buy the microphones for this, do you have any recommendations?

The price doesn't matter for now because I take the money we need from my share portfolio.


Thank you in advanced for your answers!
 
I like to record each note of each dynamic, so C3 pp, C#3 pp, D3 pp, etc. Then I do the next dynamic and once all dynamics are complete I move to the next articulation.

With some instruments though it can be better to do all dynamics or articulations for each note before moving to the next note. For example we do this when sampling dulcimers because you have to keep returning the strings, so once it's tuned you do everything before moving on.

Experiment and see what order you prefer.

Give your players plenty of breaks to rest their lips and get moisture out of their instruments.

Decide what you want to sample, ask yourself things like are you going to record ensemble samples, are you doing individual articulations or are you going to do phrases, do you need multiple dynamic layers, if so how are you going to deal with phasing/chorusing during crossfades, are you going to crossfade or just use velocity for dynamic control.

I'd avoid scripting unless you are already a programmer or what you want to do is very simple. Consider using SFZ, it's a very versatile format and easy to use.

For editing use either Reaper or Ardour. I recommend these two because both can be controlled by scripts to make batch processing 1000s of samples much easier, and both have scripts available designed specifically for editing samples.

Some good tips here - https://versilian-studios.com/2023/10/10/the-art-of-guerilla-sampling/

You might also want this book - https://www.amazon.com/Mic-Microphones-Microphone-Techniques-Impact/dp/0367470365
 
Hey, I wish you the best with your endeavor.

None of your concerns about equipment, editing, any of that matters as much as ROOM!

So, being students, do you have access to a sweet room? An appropriate band room, concert space, school theater, university, hall, architectural space engineered for sound?

That is first. Absolutely first. You can do it all with a ZOOM recorder if you have a legit space.

This is what you've got over adult-type sample developers. Access to a place that sounds good.
 
Hello,

Im a 19 year old hobbyist composer and got a group of friends who are willing to record a brass library with me.

We got 3 trumpets, 3 tenor trombones, 2 bass trombones, 1 tuba, and 4 french horn players.

We already have a studio, we just need a plan.

None of us have the experience we actually need but we really want to do it, it should start small, with just the most important articulations, we can expand the library later if we want to.

I would like a few tips from someone who has some experience, at least in theory.

Since we are all musicians and nobody expects us to get paid, the budget question is out of the question, it's just time we need.

So my main question is the process for something like this, recording, editing, scripting, testing would be my guess, i would like tips on the individual steps.

We would need to buy the microphones for this, do you have any recommendations?

The price doesn't matter for now because I take the money we need from my share portfolio.


Thank you in advanced for your answers!

Just a few quick tips which are more general ones, as the specifics of the recording will depend on what you personally want:

Wear comfy, quiet clothes (think joggers etc)

Make sure everyone eats before each session, especially in quiet sections you can ruin a recording by having a hungry person's belly making noises.

Allow about 20% extra time than it will take to record (that's on top of what you're allowing for corrections), have spare material to record incase you have the extra time unused, or it will end up being used for things you didn't consider (big instruments being too heavy for that long etc)

Tune very regularly, I'm sure that's obvious but it's insane how quickly things go out when there's no reference.

Try to have fun - but keep the reigns on. It'll be a very boring day for everyone except you, because you've got the end goal in sight - for everyone else it'll be the most boring session they've likely ever done, and it can be quite taxing to stay interested for that long, similar to actors when there's not much stimuli people start keeping themselves entertained, which can cost valuable time if its not kept in check.
 
What everyone else said is solid advice! Let me add:

If you can, do tests and build tests instruments as early as possible. You never know how something responds as a virtual instrument until it's mapped and programmed, no matter how good it sounds in the sessions.

Going note by note is less musical than planning phrases, but allows greater accuracy for tuning. If going note by note, I always play a reference note in the headphones before each note for tuning and finding each other. It's fool proof.

Click track is essential for short notes, but it's also useful for long notes, as it ensures accurate starts of each note AND makes the session much easier to navigate and edit, when everything is on the grid.

Click bleed is your enemy, so I like to use spoken click tracks. Avoid any words with "S" in them, as it tends to bleed through much more than other consonants. Say "go" instead of "start" for example, or rather count in with subdivisions but don't have any sound on the actual down beat. Then you don't have to worry about click bleed at all, except for note ends, which need to be just as together as note beginnings.

As @d.healey said you have to decide if you want to group your recording by dynamic layers or by pitch. You will get more consistency by grouping dynamics from note to note (record all soft dynamics first), but you will get less pitch and timbre consistency compared to recording all dynamics of the same pitch before moving to the next. Stereo image can also change because us musicians move and for close mics in particular even a few cm to one side or another can make a big difference - this will become apparent much later when crossfading between the dynamic layers. The bottom line is both methods have trade offs...Whatever you do, schedule anything that will be mapped and played together in the same day (like dynamic layers), because instruments and equipment can sound radically different from one day to the next.
 
We would need to buy the microphones for this
No - you can rent:


 
No - you can rent:


Actually i think it would be better to buy some vintage mics, these rent prices are sooo high and in the end i dont even own them.

If we just say we need 12 microphones (3 tree, 3 close, 3 amb, 3 out), 30 days of total recording i would be at f***** 30k€ for which i can buy 12 good microphones (i think).

Wouldnt it be better to just buy microphones used or new instead of renting for a fortune of money?
 
My first thoughts are:
I know what I like in brass recordings, but I can't make recommendations to you without knowing what YOU like.
If you're asking these questions, you're not ready. Hold off until you're confident in your choices, know what you want, and have a plan.
If the studio doesn't already have microphones suitable for recording brass, then the space itself is probably not good for recording brass. Unless you're using a university hall or something like that?
The Recordings are king. Put all of your thinking power and energy into capturing the best recordings and performances you can. The rest is the easy stuff.
If you want it to sound like a film score, engineer it like a film score. Don't engineer it like a sampling session.
Space/distance is key. Brass instruments need room to bloom. They don't sound epic by being close.
Don't aim the instruments at the mics!!!
A "close" mic does not actually need to be close. Capture the section more than the players.

Write up as much of your plan as you can, WITHOUT outside suggestions or thoughts.
What sort of sound do you want? What recordings or soundtracks are an example of the sound you're trying to achieve? Who engineered those sessions? What techniques or microphones do they use?
Most of the above information is available online in bits and pieces. There are standard mic choices for hollywood scoring stages, and there are different standard mic choices when recording a live concert. Not just the microphones, but the mic techniques and placements, too.
If your studio has a low ceiling height, then that changes which microphones to use. If your space is not wide enough to seat the sections in the typical seating arrangements, forcing you to record everything central in the room, then that affects microphone choice. The microphone stands you have at your disposal can even determine the best approaches.
If we just say we need 12 microphones (3 tree, 3 close, 3 amb, 3 out), 30 days of total recording i would be at f***** 30k€ for which i can buy 12 good microphones (i think).
Firstly, you've got the wrong idea about mic perspectives and how many mics you need.
3 for a tree, yes. 2 for outs, 2 for ambient (if you need that), and realistically only 2 for Close. You do not need to mic up every individual trumpet, if your intention is to capture a trumpet section.
Secondly, you do not need 30 days of recording. If you're only wanting the basics, you should get away with a single day for each section.
Although, it'd be worth considering mixed sessions where you alternate between sections for each articulation. This will make sure those sections are consistent with each other's performance, and also gives each section more time for resting, which is important for brass.
Plan your sampling session with notation so that you can do the calculations of how much time you'll need.

Are you using an experienced engineer? They might have some mics of their own that they like to use.
If you're not using an engineer, by golly do as much research as you can on microphone technique.

I would also highly encourage you to choose your articulations based off listening to music (a LOT of it!) and listening to what the brass players do. Do NOT simply look at existing sample libraries to decide on what to record. In many cases, sample library "articulations" are not necessarily representative of how musicians actually think about what they're playing. They are built on a foundation of old sampling methods imposed by the limitations of the tech.
If the project is meant to be small, then you'd be better off focusing on a specific style of performance, and capturing the subtleties that make that style unique. Getting a limited set of plain vanilla samples will result in plain vanilla music. Find the articulations that are used in the music you love, and try to capture those.
Have musical examples ready that you can play for them. That's the quickest way to plant the idea in their head of what it is you're aiming for.

On the topic of click tracks, you can just put a low-pass filter on it and cut out the highs. Think of it as a "thump track". That cuts down on headphone bleed.

Now, obviously all of my thoughts above are in the interests of getting the best possible result. If it's mostly a learning experience or for fun, then you can dive in head-first and just try out stuff, knowing that you might wanna do it all again in a few more years when you feel more confident.
 
My first thoughts are:
I know what I like in brass recordings, but I can't make recommendations to you without knowing what YOU like.
If you're asking these questions, you're not ready. Hold off until you're confident in your choices, know what you want, and have a plan.
If the studio doesn't already have microphones suitable for recording brass, then the space itself is probably not good for recording brass. Unless you're using a university hall or something like that?
The Recordings are king. Put all of your thinking power and energy into capturing the best recordings and performances you can. The rest is the easy stuff.
If you want it to sound like a film score, engineer it like a film score. Don't engineer it like a sampling session.
Space/distance is key. Brass instruments need room to bloom. They don't sound epic by being close.
Don't aim the instruments at the mics!!!
A "close" mic does not actually need to be close. Capture the section more than the players.

Write up as much of your plan as you can, WITHOUT outside suggestions or thoughts.
What sort of sound do you want? What recordings or soundtracks are an example of the sound you're trying to achieve? Who engineered those sessions? What techniques or microphones do they use?
Most of the above information is available online in bits and pieces. There are standard mic choices for hollywood scoring stages, and there are different standard mic choices when recording a live concert. Not just the microphones, but the mic techniques and placements, too.
If your studio has a low ceiling height, then that changes which microphones to use. If your space is not wide enough to seat the sections in the typical seating arrangements, forcing you to record everything central in the room, then that affects microphone choice. The microphone stands you have at your disposal can even determine the best approaches.

Firstly, you've got the wrong idea about mic perspectives and how many mics you need.
3 for a tree, yes. 2 for outs, 2 for ambient (if you need that), and realistically only 2 for Close. You do not need to mic up every individual trumpet, if your intention is to capture a trumpet section.
Secondly, you do not need 30 days of recording. If you're only wanting the basics, you should get away with a single day for each section.
Although, it'd be worth considering mixed sessions where you alternate between sections for each articulation. This will make sure those sections are consistent with each other's performance, and also gives each section more time for resting, which is important for brass.
Plan your sampling session with notation so that you can do the calculations of how much time you'll need.

Are you using an experienced engineer? They might have some mics of their own that they like to use.
If you're not using an engineer, by golly do as much research as you can on microphone technique.

I would also highly encourage you to choose your articulations based off listening to music (a LOT of it!) and listening to what the brass players do. Do NOT simply look at existing sample libraries to decide on what to record. In many cases, sample library "articulations" are not necessarily representative of how musicians actually think about what they're playing. They are built on a foundation of old sampling methods imposed by the limitations of the tech.
If the project is meant to be small, then you'd be better off focusing on a specific style of performance, and capturing the subtleties that make that style unique. Getting a limited set of plain vanilla samples will result in plain vanilla music. Find the articulations that are used in the music you love, and try to capture those.
Have musical examples ready that you can play for them. That's the quickest way to plant the idea in their head of what it is you're aiming for.

On the topic of click tracks, you can just put a low-pass filter on it and cut out the highs. Think of it as a "thump track". That cuts down on headphone bleed.

Now, obviously all of my thoughts above are in the interests of getting the best possible result. If it's mostly a learning experience or for fun, then you can dive in head-first and just try out stuff, knowing that you might wanna do it all again in a few more years when you feel more confident.
You are right, here is the sound we are aiming for :




So i want at least 3 types of shorts, so i can do fast short passages with alternating notes without the muddy and unprecise sound i get when i do this with f.e. CSB or AROOF.

I want this crisp clear LSO early 2000s era sound, our hall is about 450m2 and has a rectangular shape, high ceiling and acoustically insulated, should actually work quite well.
 
Hello,

Im a 19 year old hobbyist composer and got a group of friends who are willing to record a brass library with me.

We got 3 trumpets, 3 tenor trombones, 2 bass trombones, 1 tuba, and 4 french horn players.

We already have a studio, we just need a plan.

None of us have the experience we actually need but we really want to do it, it should start small, with just the most important articulations, we can expand the library later if we want to.

I would like a few tips from someone who has some experience, at least in theory.

Since we are all musicians and nobody expects us to get paid, the budget question is out of the question, it's just time we need.

So my main question is the process for something like this, recording, editing, scripting, testing would be my guess, i would like tips on the individual steps.

We would need to buy the microphones for this, do you have any recommendations?

The price doesn't matter for now because I take the money we need from my share portfolio.


Thank you in advanced for your answers!
Good luck on this, sampling can be so much fun and also a total slog. Usually both at the same time.

- write out sheet music for everything you're going to record. you don't need to write out the individual dynamics, you can use the same sheet music for piano, mezzo, forte... this way you can accurately estimate the time it will take to record. (number of bars times number of beats in bar {4 in 4/4} divided by the bpm). you have to account for the reverb tail of the room to make sure there's enough space between notes to capture the entire room.
- record phrases, not individual notes. if the studio is super washy, you'll have to put some gaps in around the target notes to account for the reverb, but it will still make the recordings closer to the way musicians actually play.
- for brass players, plan for 100% of playing time for resting. So, once you add up all the time the sheet music adds up to, double it and that's how long you need to actually record it. add in setup time and a couple of hours to get sounds on the first day and an extra 30 minutes each day for latecomers (if no one's getting paid, someone will be late). that's your total studio time. it'll be a lot more than you think it will be.
- you're buying mics?? the studio you're recording in should have plenty of good mics. ribbons is the ticket for brass.
 
I would also highly encourage you to choose your articulations based off listening to music (a LOT of it!) and listening to what the brass players do. Do NOT simply look at existing sample libraries to decide on what to record. In many cases, sample library "articulations" are not necessarily representative of how musicians actually think about what they're playing. They are built on a foundation of old sampling methods imposed by the limitations of the tech.
This. 1,000%
 
Oh a very important tip. Use an audio editor that displays a spectrogram, you will see sounds that you won't hear but that your users might hear once they start doing crazy things with your samples :)

If possible also have a live spectrogram during the recording - if the DAW you're using doesn't have one built in (ProTools) you can add one as a plugin or use a standalone one and route the audio to it, I like Friture (https://friture.org/index.html). You'll spot things in this that you might miss with your ears in the moment.

Just look at that click bleed!

1720519317192.png
 
Oh a very important tip. Use an audio editor that displays a spectrogram, you will see sounds that you won't hear but that your users might hear once they start doing crazy things with your samples :)

If possible also have a live spectrogram during the recording - if the DAW you're using doesn't have one built in (ProTools) you can add one as a plugin or use a standalone one and route the audio to it, I like Friture (https://friture.org/index.html). You'll spot things in this that you might miss with your ears in the moment.

Just look at that click bleed!

1720519317192.png
Yeah, definitely. I mean basically, use Reaper to edit samples. It's better in every conceivable way.
 
Hi bababooey
You are young. At your age, you still have the idea that you can change the world, that you can do everything better than it already is. This conviction, hope and great faith in oneself is part of young life. That's great!
So I think that you should definitely try to create a library with which you can recreate the Star Wars sound...
You'll learn a lot in the process - unfortunately including the fact that you can't do it any better than all the big library manufacturers, who have been producing such libraries for a long time with a lot of experience and usually with the best equipment. Here is a small sample for motivation.

As I said, it's still important to gain experience. You will learn a lot about microphones, acoustics, sample layers, programming and so on.

I wish you a lot of success and fun of coarse!
Beat
 
Last edited:
My number one tip: make sure you leave enough tail after the end of each note. The musicians will want to rush through, as it it will seem to them like a note is finished is soon as it is released. But for sampling you need a lot of ring-out of the room tone or the samples will feel cut off. I had to throw my first library in the trash for that reason. Using a click, as suggested, and sheet music that has that long tail built-in will help. And it would also give the brass players a moment of rest for their chops.
 
Top Bottom