Version française


? | Home page | Tutorial | Recording

Homepage
Home page



RECORDING

Preparing the session
Drums
Guitars and bass
Vocals

Here is some advice to make recordings in good conditions.
Top of page

PREPARING THE SESSION

Before anything else, we need to prepare the recording session. We are not going to look for a melody, nor improvise or rehearse... we are going to properly record a song that we have already written. To avoid doing the same things over again each time I wish to record a song, I prepared a blank template which contains all the tracks and buses I need. I may naturally add or delete some elements if the template is not appropriate for the current project.

What does my template look like?

Track-wise:
- Rhythm guitar tracks (from 2 to 8 depending on the project)
- Solo guitar tracks (usually 2 tracks to make the sound thicker)
- Two bass tracks (one with the direct raw sound, and one with an amp simulation)
- Lead vocal tracks (usually one or two tracks, depends if I record it twice or not)
- Background vocals tracks (if the project demands it)
- Keyboard tracks (same thing, the number of virtual instruments will depend on the project, can be none, can be 5 or 6...)

Then we have drum tracks. There is one track per drum element. They are automatically created when I insert my virtual drum plugin:
- Kick drum
- Snare drum
- Low tom
- Medium tom
- High tom
- Hi-hat
- Crash cymbal
- Ride cymbal
- Splash cymbal
- Overhead microphone
- Room Ambiance microphone
- Piezzo microphone
- One MIDI track on which the drum score will be placed.

Bus-wise, I have one group for the guitars, one for the bass, one for the vocals, one for the drums, each of them is redirected towards the Master bus, which goes out throuh my studio monitors. It looks like this:

Track/bus structure
Track/bus structure

What is a track? what is a bus?

Tracks are where you record an audio or a MIDI instrument. For instance, track 1 will contain guitar parts, track 2 will contain guitar parts of a second guitar, track 3 will get bass parts, track 4 will have vocals, etc. Everyone organize their tracks as they feel more appropriate.


Tracks
Tracks
Tracks in Sonar's Console view

Bus
Bus: uh... yeah, but no...

Buses do not carry recordings. They are groups of tracks. For instance, you can add a "Guitars" bus towards which you will redirect your guitar tracks. Thus, you may allocate common effects without adding these effects to each and every guitar track. Just put the effect on the bus, and each track will produce the desired effect. This will also allow you to mute all the tracks connected to a bus, or to solo them in one click. Again, you may organize your buses the way you want, and you will find the use of buses very convenient.


Bus
Bus

Buses in Sonar's Console view

Both tracks and buses can be either mono or stereo. In today's D.A.W.s, the number of tracks and buses you can create is not limited. Actually, the only limit are your needs, and the power of your computer's processor.

Track view
Track view

Sonar's track view.
The size of the "Tracks" and "Buses" windows can be modified at will.


Finally, before we start, and if your gear and software allow you to do so, choose to record in 24 bits, the internal calculations and the sound quality will be more accurate. Once your session is prepared, the recording can start. Start with what, by the way? Well, I always start with the drums, let me tell you why...

Top of page

DRUMS

Recording an acoustic real drum kit is far from being easy, even for professional sound engineers. It a time-consuming process, it's frequent to spend several hours placing the microphones around the drum elements before you can actually record. But we are in a home studio, and we will have to deal with a software drum kit, based on midi files...
First of all, why start with the drums? The answer is simple, we will use drums as a metronome. The drums sound will guide us and help us follow the tempo. This will allow for an even recording and the song will not speed up or slow down unintentionally. Of course, variations can be interesting and bring some life to an otherwise mechanical tempo, but let's consider that a studio session seeks recording perfection, even though it's only a home studio.

I usually have no idea what my final drum track will sound like. Actually, I only adjust it when the rest of the song is finished. But I still need its metronome function to record all other instruments. Thus, I create a drum track which repeats itself over and over again, and I try to have this loop match what I am about to play (no punk rhythm to record a ballad). For instance, I'll use one of these patterns:




Each sample contains one or several "normal" loops and ends with a "fill" loop that will play at transition moments, like going from verse to chorus, or chorus to bridge, etc. This can help when recording. I put as many patterns as needed on my drum track, and I place a few additional bars (usually 2 are enough) at the very beginning in order to feel the tempo and start recording in tempo. These additional bars will be deleted at the end.
Recording starts at measure 3
Now that we have a metronome drum track, we can proceed with the recording of the other instruments. Guitars and bass follow...

Top of page

GUITARS AND BASS

Let's not forget we are recording in a home studio, in an appartment and it is simply impossible to play with a good old 100-watt tube amplifier, without the neigbors calling the police. So we are going to have to record the guitars and the bass directly through the audio interface. No real amplifier, no microphones involved. The latter solution would be preferable, but on of the benefits of direct recording, coupled with amplifier simulators, is that you can always edit the sound later, without having to re-record. Just change the settings and you're done.

So now... bass or guitars first? There isn't one clear answer. Bass and drums are the foundation, the rhythm base of a song and everything else should rely on them. But other factors could also be taken into consideration: for instance, the person recording may be more comfortable with a guitar than a bass, and will rather play guitar first. Or maybe the song has a very important bass riff that compels you to record it first. In any case, you are the one who can decide. If you are uncertain, then the drum / bass duo is a safe bet. If this is in place, then the rest can easily be added.

Guitar or bass, the recording process will be the same. Plug your guitar into the pre-amp, the pre-amp is connected to the audio interface (or plug your guitar directly into the audio interface if using the interface's pre-amp), and set the recording level. This is very important! Before recording anything, check that you are not going to go beyond the maximum level (0 dB, zero decibel). In a home studio, you won't have a sound engineer besides you to make adjustments on the fly, while you are playing. You are the one to take precautions. How can you do that? Simple: try and adjust, it doesn't take long and will prevent you from making a perfect take, then realize the levels were too low or too high, forcing you to do it all over again.

Have a try: for a rhythm guitar for example, play the loudest parts and set the preamp and audio interface volume levels in such way that when you play the loudest, the recording level doesn't go beyond -6 dB. The absolute maximum that you should not reach or go beyond is zero dB. If you play in your try the same way you play during the actual recording, then you can be certain the recording level will be correct. If your average level is between -9 db and -6 dB, then your level is sufficient and you have a margin of error before clipping.

Clipping is the term used to indicate that you reach or go beyond 0 dB. Clipping is your enemy :-)

No clipping    With clipping
You can clearly see here the difference between a take with a correct level, and the same take with too much gain. On the right, the sound is crushed, there are no distinction between the peaks and the troughs. Let's zoom in:

No clipping, zoom    Clipping, zoom
If you zoom in, you can see that the right curve is beheaded, anything that goes beyond 0 dB is crushed. Why is that wrong? Because the sound gets distorted, saturated. And you should avoid this, unless you voluntarily seek to give a headache to your listeners.

Example of clipping with vocals:
    

With a guitar:
    
No clipping (left) and with clipping (right)
Alright? Your levels are checked? Then go for it, record your guitar and bass tracks.

Set your guitar as you wish by using the bass, treble and microphone buttons. Try to avoid very low sounds (muffled) or very high sounds (shrill). The sound you record should be pleasing to the ear. Remember that, when mixing time comes, you won't be able to magically bring to life what was not recorded in the first place. A high-pitched sound will never produce a warm low sound, even by tweaking the knobs. So think in advance what you wish to obtain in order to choose your sounds in a coherent and logical way.

Now stay in tempo and if you make a mistake, start over. Once the song is complete and starts spreading around you, it will be too late to regret a bad take. So do things well now. Not everybody is as talented as Jimmy Page (Led Zeppelin), Eric Clapton, Slash or Matthew Bellamy (Muse) and everybody can make mistakes. There is no emergency, and if you have to record the same take 30 times, do it. Do and re-do until you are fully satisfied. Record and carefully listen to the take, especially if your think it's the right one. Regularly save your project. You never know if your computer will crash and make you lose hours of work.

If this helps, record with the guitar sound you chose (see amplifier simulators), but for the time being, what counts is the quality of the recording, not that of the amplifier sound. This is what's so great about the amp sims, you can edit the sound whenever you want. We'll see that later, when mixing.

OK, you have now recorded all the guitar tracks you needed. Maybe your song requires only one track, or maybe you needed 8 tracks, that's up to you to decide. It can be interesting to double the takes, in order to thicken the sound. But remember, the more instruments and tracks there are, the more difficult mixing becomes, because it's more complicated to have all instruments coexist. Looking for a "bigger sound" doesn't necessarily requires multiplying the tracks, it can also be achieved by using the appropriate effects.

Of course, the bass is also a guitar (a bass guitar...), so the recording constraints are identical to those of the guitar. Check your recording levels to avoid clipping, and above all, stay in tempo, or the song will sound shaky.

Top of page

VOCALS

I prefer to record them last but there are no rules. If you prefer to record them first, then do so.

To record vocals, make sure the place is quiet, shut the door, tell the people who live with you to be quiet, and do not record while your neighbor is drilling holes through his kitchen walls! Also, turn off your monitors and use a headset instead to avoid recording the playback with your microphone.

Condenser or dynamic microphones?

Dynamic microphones are solid, they don' need a power source, they can take heavy acoustic pressure (like a kick drum or a saxophone) and they are not too expensive. They are also less sensitive to surrounding noises than condenser microphones. The cons are they lack clarity in the high range, which renders takes less clear and defined than with condenser microphones. They can be used with Jack or XLR plugs.

Condenser microphones are much more responsive and accurate. Their high sensitivity is double-edged, because they will capture any noise when recording. The fans of your PC are noisy? Chances are this noise will be recorded. Sound comes out of your headset? It will be recorded by your condenser microphone. Children are loudly playing outside? You might get that too. However, some condenser microphones are called "cardioid", or "hyper cardioid", and they only record what comes from a specific direction, ignoring (more or less) other sound sources from other directions. On the contrary, omnidirectional microphones record what comes from anywhere. Not ideal for a home studio. Condenser microphones are also more fragile (don't knock them) and must be powered through a "phantom power", whose standard is 48 volts. This kind of power is either present on your audio interface and can be turned on and off with a button, or it will require the use of an external phantom power source that you will then connect to your audio interface. You have to use 3-pin XLR plugs that carry the phantom power current. Finally, condenser microphones are usually rather expensive, some of them cost several thousand euros (or dollars, or pounds), but only professional studios or rich amateurs can afford those. On the plus side, the sound you get with a condenser microphone will have the best quality.

Be cautious though, a good dynamic microphone is worth better than a bad condenser microphone. No big secret here, for microphones like for anything else, very low prices are rarely synonymous with good quality.

A few known and renowned microphone brands: AKG, Milab, Neumann, Rode, Sennheiser, Shure...

Jack plug    XLR plug

Jack plug (left) and XLR (right)
Some pieces of advice: buy a microphone stand and a pop filter (you can also make one yourself with wire and a piece of tights from your wife / girlfriend / mother / daughter / neighbor). The stand will prevent you from manually holding your microphone and thus produce handling noises. As for the pop filter, it prevents the air to hit the microphone and produce unwanted blowing sounds when you pronounce some letters such as "p" or "b".
Microphone stand    Pop filter

Microphone stand and pop filter
Furthermore, try to stay in front of the microphone when you sing, don't move from right to left or back and forth in order to avoid big volume variations. Dynamic microphones will not record you correctly if you stand too far from the microphone (8 inches, 20 centimeters would be a lot, meaning you can easily go too far). Don't let this piece of advice prevent you from "feeling" your song. If thinking about your position prevents you to have emotions, just forget about it but try to remember that the less you move away from the microphone, the better the quality. Nevertheless, if you have to really scream all of a sudden (I think about Frank Black from the Pixies, who goes from whispering to howling in an instant), you should then step back from the microphone in order to avoid clipping.

Just like the guitar recording, you have to set the recording level of your vocals before the take. Have some tries and check once again that the level doesn't go beyond 0 dB. Now, if you plan on having very different volume levels, using the howling moments as a reference will render the quiet moment barely audible. So, either you live with that (but you could bring some noise when compressing the quiet sounds during mixing), or you make several takes, with different level settings for the quiet and loud moments. You can also get the help of a friend who knows the song and will manually adjust the levels on the fly.

I also advise you stand up when recording vocals. Singing while sitting is a bad habit, as it prevents you from breathing deeply and could compromise your vocal amplitude. When standing, singers can better "live" their song. Look at studio recording videos: singers stand. Not only because it looks nice on the video, but for breathing reasons. Also think about carrying your voice. I am not saying you should scream (unless it is required in the song), but you should carry the voice. If you're mumbling, people will hear it. Think of Mick Jagger, David Bowie, Elvis Presley or Freddy Mercury, they carry their voice. You can clearly hear when they refrain, then just let go. So let go, let your future listeners feel that you gave it all, that you lived your song. But once again, unless it is voluntary, do not force it. You are not taking part in a shouting contest!!!

Unless you have a natural gift (and even then...), be aware that singing requires work to be mastered. Nobody becomes a great singer without practice. Do not overestimate your capabilities. Each of us has a singing range (tessitura) that can only be broadened through training. This is the range of notes one can sing, from the lowest to the highest. Needless to sing a note that is too high-pitched for you. If you cannot reach it (not yet), forcing won't help you much. At best, it will sound awful, at worst, you'll damage your vocal cords.

Top of page

No need to go on and on forever, recording is rather easy. As long as you pay attention to your recording levels and take care over your takes, you should get a satisfying result, good enough to finalize the song

Next: Mixing in theory
Top of page


MESSAGES

(leave a message)

Messages page # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35



lgwn
le 08/05/2012 à 18h42

Merci pour tout, super boulot.



Malandain
le 24/04/2012 à 14h37

Bonjour,

J'ai un dr4d akai qui enregistre en 16 bits (44/48khz).
Je relie numériquement mon dr4d à ma carte son (spdif) qui est reliée a mon ordi en usb. Ma D.A.W. est Mixbus qui enregistre en 32/24/16 bits.
Pour un bon enregistrement, faut-il mieux mettre tout en 16 bits ? (dr4d + carte son + D.A.W.)
Peut-on enregistrer le 16 bits du dr4d en 24 bits carte son + D.A.W. ?
Ou 16 bits du dr4d vers 16 bits carte son et 24 bits Mixbus ?
J'ai l'impression que tout en 16 bits, c'est mieux.

Merci d'éclairer ma lanterne.

A. Malandain

* * * * * * * * * * * * * * * *

<em>Bonjour,

Bienvenue sur mon site !

Puisque le dr4d enregistre en 16 bits, il semble logique de penser que tu ne gagneras a priori pas en qualité en réglant carte son et séquenceur sur 24 bits.

C'est vrai si tu te contentes de "transférer" les sons enregistrés avec le dr4d vers ton séquenceur sans les traiter ensuite.

En revanche, si tu fais du mix depuis Mixbus, en utilisant des effets, là il vaut mieux être en 24 bits, car les calculs effectués par le logiciel seront plus précis.

Il est très probable que la différence sera imperceptible, et travailler en 16 bits de bout en bout ne va pas dégrader le son obtenu.

Je ne pense pas que beaucoup de gens soient capables d'entendre la différence entre un traitement effectué en 16 bits et le même traitement effectué en 24 bits.

Néanmoins, la différence existe au niveau de la précision des calculs, et l'accumulation des effets sur différentes piste peut entraîner en 16 bits un risque (très faible) d'avoir quelques petites approximations qui vont se traduire par une légère dégradation du son. Encore une fois, c'est théorique, ça peut arriver, mais ça ne sera pas forcément quelque chose qui va s'entendre.

Donc si tu es pointilleux, je te conseille de régler ta carte son et ta D.A.W. sur 24 bits.

Grebz</em>



Franck
le 02/04/2012 à 03h34

Bonjour Grebz,

J'ai vu que tu utilisais Sonar X1 Producer.
J'aimerais bien si tu veux que tu me donnes quelques truc pour ce logiciel car sur Internet, il n'y a que des vidéos anglo et je ne les comprend pas toutes.

Merci d'avance.

* * * * * * * * * * * * * * * *

<em>Bonjour,

Pour ceux que ça intéresse, j'ai fait ce petit guide rapide pour savoir comment intégrer et utiliser un VSTi (synthé virtuel) dans Sonar X1. Ce ne sont que les bases, mais ça pourra servir à ceux qui sont perdus.

Télécharger "<a href="/Downloads/tutorial_homestudio/Synthés virtuels dans Sonar X1.pdf">Synthés virtuels dans Sonar X1</a>" (Fichier PDF de 1,20 Mo)

Grebz</em>



Molten
le 13/03/2012 à 01h27

Merci mon ami : Excellent !!!



Alain6465
le 08/03/2012 à 00h54

Merci pour ton site qui est très utile pour un novice comme moi dans la MAO, même si je fais de la musique depuis longtemps.

J'ai l'intention de revendre mon matériel hardware et d'acheter un séquenceur en soft. Je penche pour Sonar car j'aime bien l'interface graphique de studio instrument, mais je ne sais quelle version choisir, Sonar Studio ou Producer ou X1 etc.

De plus, je n'ai pas vu la VS-100 de Cakewalk dans tes choix de carte son, qu'en penses-tu ? Serait-ce un bon choix, car elle a été conçue pour être utilisée avec Sonar je crois.

Merci pour ta réponse et pour tes conseils très pédagogiques !

* * * * * * * * * * * * * * * *

<em>Bonjour,

Merci de ta visite sur mon site.

Pour l'interface V-Studio 100, je ne l'ai pas listée car je ne l'ai jamais utilisée et j'ai simplement référencé quelques marques connues, mais ce n'est pas du tout un choix exhaustif. Je suis un simple passionné de musique, pas un professionnel du test, et je ne me permets de parler en détail que du matériel et des logiciels que je connais !

J'ai été voir divers avis d'utilisateurs de la VS-100 et ils sont globalement très positifs, tant en ce qui concerne la qualité des pré-amplis que de l'aspect pratique des différents boutons présents sur l'interface, qui donnent un accès direct à de nombreuses fonctions du séquenceur utilisé. Effectivement, la carte a été conçue par Roland, qui est aussi l'éditeur de Sonar, et donc Sonar reconnaît immédiatement la VS-100 et attribue directement les bonnes fonctions aux bonnes touches de l'interface. Mais la VS-100 fonctionne aussi très bien avec les autres séquenceurs, avec lesquels il est tout à fait possible de configurer les boutons pour leur attribuer les fonctions voulues. Conclusion : même si je ne l'ai jamais utilisée moi-même, cette interface me semble être un bon choix.

Pour le choix d'un séquenceur maintenant... En tant qu'utilisateur de Sonar depuis plusieurs années, je vais évidemment te dire que c'est un très bon choix. L'interface me convient, je le trouve assez intuitif à utiliser et largement assez riche et puissant pour moi. Point important : Sonar est livré avec beaucoup de plugins (VST, VSTi) et outils qui permettent de travailler sans avoir impérativement besoin d'acheter des outils complémentaires. Mais restons objectif, les Cubase et autres Reaper ou Presonus Studio One sont aussi d'excellents logiciels. Je crois que c'est surtout une question de feeling, de coup de cœur... et de prix ! Tous les séquenceurs sérieux permettent de parvenir plus ou moins aux mêmes résultats, chacun avec sa méthode et son organisation.

Dans la gamme Sonar, le dernier en date est Sonar X1, qui existe en 3 versions : Essential, Studio et Producer.
La différence se trouve dans le nombre d'outils livrés avec le logiciel.

<strong>Essential</strong> est la version la moins complète et la moins chère. Elle est disponible uniquement en version 32 bits et est limitée à 64 pistes audio (ce qui suffit à la plupart des gens, moi y compris). Elle ne propose que les fonctions de base (pas de synthés, batterie logicielle, outils de compression évolués, etc.). Environ 90 euros.

<strong>Studio</strong> est disponible en versions 32 ou 64 bits, n'a pas de limitation du nombre de pistes, et possède des outils complémentaires (divers VST, VSTi, V-Vocal pour corriger les problèmes de justesse des voix ou instruments, Audio Snap pour caler des enregistrements sur le tempo). Environ 180 euros.

Enfin, <strong>Producer</strong> est la version la plus complète et la plus chère bien sûr. Elle contient en plus différents outils d'excellente qualité comme le ProChannel (compresseur, EQ, saturation), Session Drummer 3, un logiciel de batterie plutôt bien foutu, les VSTi Dimension Pro (très bon synthé, très complet) et TruePianos (version allégée d'un piano virtuel de très bonne qualité à mon goût). Environ 390 euros.

À sa sortie fin 2010, X1 n'était pas super stable et plantait un peu trop souvent à mon goût. Du coup, en attendant les correctifs, je continuais à utiliser la version précédente, mais c'est très frustrant et décevant de payer pour un logiciel qui ne marche pas aussi bien que ce que vous attendez. Mais heureusement, les correctifs ont rapidement amélioré la situation et la version de X1 actuellement en vente est nickel.

J'ai encore quelques plantages de temps à autres, mais j'ai pu constater que c'était toujours à cause de certains VST que j'ai ajoutés. La faute donc aux programmeurs de ces VST qui perturbent Sonar dans certaines conditions (pas toujours très claires à déterminer, d'ailleurs). Ça n'arrive d'ailleurs qu'avec des VST gratuits. Je n'ai jamais eu de plantages à cause de VST du commerce. Normal, puisque les conditions de tests ne sont pas les mêmes... Entre un bidouilleur passionné qui développe son VST dans son coin et une société qui peut tester sur tous les séquenceurs du marché, il y a une différence. N'empêche que beaucoup de VST gratuits sont non seulement utiles, mais aussi performants que certains logiciels du commerce.

Grebz</em>

Top of page