Composers and arrangers - how do you mix?

Cheez

Moderator
For those who compose/write songs/sequence, how do you do mixing? All along, I've always been mixing directly from my midi tracks - ie do all my mixing from my softsamplers/softsynths via midi, and mix down directly to audio. I didn't realise that this is not the ideal - as there are a lot of processing and post-production that will be missed in this process. However, there's a sort of workflow for composers and writers that we follow to make things easier while we write. How do you overcome that? It will also be great to have sound engineers to chip in on what you think is the ideal from your perspective. I recall a musical I arranged for 10 years ago - I rendered everything to a stereo audio track. Then during recording with live vocals, the studio engineer was really upset because he said he can hardly do anything to the "damage" I've caused in my 1 hour track!!!!

Here goes. My "new realised" method now becomes:

1. Set up template of instruments within DAW. At this stage, each instrument's volume/panning is already fixed to a listener's perspective. I sequence as I write and layer as I go, so hearing how it should sound like right from the start is crucial.

2. Sequencing and writing step. Will not go into detail (as there's an old thread where we discussed extensively about this)

3. Mixing step. I realised that the actual (or final) mixing should really be done in audio, not midi. Step 1 is hence not real mixing - but getting a sort of perspective that helps me to write. In the mixing step, I would:

4. Reset all panning to center. Reset all volume for every instrument. Each track is played solo and volume set so that masters meter would be at least -12dB for each track.

5. Bounce individual midi tracks into individual audio tracks.

6. Mix using audio + processing.

7. Listen very carefully. Sometimes I would discover after processing (which becomes more evident sometimes) that certain parts require changing (e.g. lack of tightness of bass guitar and kick drum). I would go back into the midi track and make changes, and re-bounce to audio track. Or if I find certain discrepancies with volume - so instead of just applying compression, I can go back again to the midi to edit certain midi note velocity, then re-bounce.

8. Master.

For sound engineers, when somebody sends you an OMF file for mixing and mastering, would each track be how it is as described?
 
I'm an amateur in music, mixing and mastering. However, I'm exploring good practices with mixing and mastering.

I would also like to read more on this topic by fella softies, adept in mixing and mastering.

I do not know how to use software DAWs very well. I rarely bounce tracks to audio for mixing. I reckon I have enough processing power to mix relatively small number of tracks and MIDI soft synths and effects (less than 10).

I do not fully understand the meaning of mixing in audio and MIDI. I assume we are not exactly mixing MIDI data, since it'll be converted to audio in the mixer track for mixing to buss or master channels. That's how I see things in FL Studio, which really isn't a DAW. Some people call it pseudo DAWs. Perhaps FL Studio has made it easy for amateurs like me to use it.

Regards.
 
For me, from the start, I don't care too much on mixing.
It is important to drop all your layer.
During improvising on each layer, you could pan instrument away from it getting messy on the centre.
You could put some effect if it is to change the character of the instrument.

Something you should not do at the start:
1. Find the right EQ and spend timing adjust little by little.
2. Dial compressor to get the right punch, but rough one can.
3. Putting reverb to create space(unless it is too messy).
4. Precise pan on track, unless you need to do left right bouncing for that track.
5. Do not do volume automation unless for writing expression(sometime I don't care about writing expression at first).

The important is to drop all your layer idea first.
Of cos, if you can't stand the outcome of the song/music.
Do at least 75% of your song/music and start the mixing process.

Take a break when you wanna start mixing.
Drink coffee, watch TV, call girlfriend/boyfriend.
It is better to start on the day next if you are not press for deadlines.

When you start the mixing process, first listen through the whole song/music.
Get the feel.
I alway break my instrument in Primary and Secondary instruments.
Primary instruments are the skeleton instrument which bring out the song.
Secondary instrument are those you can do without it, it just add colour to your song.
I alway image the set of instrument place right in front of me.
This help to determine on how loud and placement of instrument should be and start to mimic that.
Take a break for every 1/2 to 1 hour.
Listen some song, play Facebook, call your mom to buy toto.
Then continue to mix.

Lastly, a good monitoring setup is vey important. Take a understanding on how place speaker to create a good soundstage. Do take a walk at Adelphi and listen to some of the great audiophile system setup. Get the feel and listen what a pair of speaker can do.
 
Some very good points there. The reason I panned my strings, for example, right from the writing process (not mixing yet) is so I can hear it from a conductor's perspective. That help me in my melody/harmony/counterpoint writing. During this time, I use a lot of staff view in the DAW, and sometimes I need to write down the music of certain parts just to get the harmony right. Sometimes, I just straight sequencing into the DAW. The music is already in my head. I just need to make sure that it is captured in the right style and mood, and that the harmony makes sense. If everything is in the center, I find it very muddy and distracting to get it clear.

Zerox, I agree with Kongwee that mixing is (and should not) be done during the writing process (ie in midi). It's just something simple to get the right hearer's perspective. The key at this stage is to get the music right musically. No good mixing is going to help if the piece is not well written.
 
Heyo,

I think mixing is essential in arrangements process more than the composing process.
Mixing enables me to see if the arrangement is phat enough or still need more layers.

So I would say I usually:
Compose -> Arrange -> Mix -> Re-arrange if need to -> Mix -> Master

Don't flame me for saying this, but I experience that mix-down is not enough to tell if song is ready, i would pseudo-master just to hear how it roughly sounds in the final product before deciding to finalize arrangement stage. Maybe not work for you, but it has been for me so far. :)
 
For me, mastering in a single song, is just final overall effect enhance in the master bus(EQ,Compression,Reverb,stereo widening...etc). If need to bundle songs together and make it sound like coming from one place, I will do mastering. But very rare I need to do it. Most I do mastering is for MMOs in live situation. This easier to dial in a setting in mixer board, adjust the knot or fader in smaller step during the action. If it is very hard for me to dial in setting in every MMOs without doing anything.
 
The struggle for composers and arrangers to do mixing is that we need to change our hearing perspective at each stage. I would:

1. During composing - take on the perspective of the conductor and composer. I hear from the standpoint of the conductor - how it should sound with me conducting right in front.
2. During sequencing - take on the perspective of the player. When I sequence each part, I imagine I'm playing that instrument with the rest of the band/orchestra etc. This is crucial to make the sequencing as realistic as possible. This also means constant re-sequencing, since the first sequence done will be "solo" with no other instrument playing, and therefore the least realistic. As more tracks are laid down, it gets easier to imagine playing with a band. So the first few tracks will almost definitely have to be redone.
3. During mixing - take on the perspective of the audience. What I hear as a player or conductor will be different to how the audience hear.

Each of these perspective are not only different, but can conflict with one another. In the end, it's about how the audience hears. But that is built on the foundation of good music (how it's written) and good playing (how each instrument is being played). And hence during each step, I find myself going back to re-do things.

How about the others?
 
Something that is off topic.
Doing for people, the hardest is to get what the person want(especially people who are highly trained musician).
Even the people that giving you the info can misled you.
I once need to do two arrangement for a musical.
They send me the info and direct to go.
I arranged and send, but they reject it.
After sometime, I get to the listen to the same piece arranged by other people.
It is total different style from what I being told to do.

Next scenario, a highly respected classical pianist who can do some arrangement.
She learn about DAW from the same instructor that teach me.
She just ask my instructor to do an album for her but never agreed.
So I get to do a demo arrangement for her.
Never able to do it right for her.
There is no news about the album she plan to do.

These two case, I never really get to talk in person.

In most of cases, I will send an initial arrangement and ask about feedback.
Quite often, I need to rearrange when I started doing arrangement.
Nowaday, I just need beef up my arrangement two three time as I learn to interpret what the client want from me.
 
For arrangements, which is indeed a different topic, I would actually spend quite a bit of time talking with the director. The musical I arranged, I think I must have spent hours with the director before I started each scene. I made sure that I understand the exact style he wanted. I played it on a keyboard for him to listen to make sure that it's more or less what he wanted to hear. Only then would I start arranging a scene. And after it's arranged, the director would listen again, and almost all the time, I find I had to re-arrange certain parts to get the feel he wanted.

And I did that for every single scene. I did one scene at a time, complete it to his liking, then get it sent off quickly so he can start getting the choreographer to work on the dances and get the singers/actors/actresses to start practicing. And the scenes doesn't always fit chronologically - it all depends what the director wanted done first (some had complex dances and needed more time - so those get completed first). Then I would start the next scene following the same pattern. Sometimes, it's really hard to understand what the director wanted. He had to get me other musical CDs and let me listen to the "feel" he's describing. I did 8 scenes in total for that musical, plus the overture and the curtain call (both of which must be done last so snippets of the entire musical is included in them, especially the overture; not necessary for the curtain call). A lot of time is spent talking with the director. I need to get into his head, which is not always easy.

Another scenario - working with a songwriter. This was writing music to lyrics. Even before I started writing the music, I spent hours talking with the songwriter to get into her head - to understand exactly what she's thinking about she she wrote the lyrics (the background, the feeling etc). Only when I got the feeling would I attempt to start getting the music down. And in the process, I realised that some words needed to be changed or removed as they didn't fit musically. We met again and spent hours discussing whether some words could be changed/added/removed without changing the meaning. Then it's changing the music again. Took a lot of time. But the end product was something both she and I'm were happy with, esp since I felt the music brought out the meaning and feeling of the words. Some people heard it sang and cried - so it's mission accomplished for me!

The easiest, of course, is arranging my own compositions. This way, the arrangement is always done "as the composer intended"! :)
 
Back
Top