Question for Pros, How to remix ya track whilst creating it

[dominicw78]

haha that's so easy, how come you don't know that, are you retarded or something ? obvisously you're not a "pro producer" !

All right, sorry.

There is a lot of different ways to do, and they are all right/wrong.

Keep in mind though that some way push you to commit yourself and work faster, and some other open more possibilities, and more adjustment, but then you never go to the next step.

For example, I tend to work with audio every time. If I do something in midi, I'll sample it as soon as I can. Cause once I have the audio, there is no way back, and at the idea of coming back and work on it again is out of my head and I can focus on the next step.

So there is a right balance between committing too fast and keeping to many options.
Find this right balance to suit your workflow.

So to take an example for your problem, I think I will not care of all the EQ/compression. I will program some basic paterns first. Then I bounce them on every track. Then I can work on the EQ/compression. Then I can copy/past this basic patern and edit them manualy by cuting/copy/etc... I can always modify an EQ or compression, but my midi part is gone and I work with audio so the midi programing part is done at least.

You were wrong in assuming that my parts were still in midi, I'm always immediately bounce to audio. I agree with you sentiment, however I don't believe you fully appreciate how it relates to the scenario I have explained. If you read my post again, considering that all the parts I am referring to are already in audio you will no doubt see the situation in a different light and understand what I'm getting at. You can't always modify the eq and compression as you have suggested in this situation otherwise it will not match your earlier drum bus bounce.

[dominicw78]

Catch 22, couldn't have put it better myself, the second option of committing & bouncing down seems the more logical one however it seems to require a massive amount of organisation, unwavering commitment and forethought that seems very daunting to me.

Commit, and move forward. There's always more music to make, the more time you spend second-guessing yourself the less time you have for making new tracks. Think about the "olden days" when people worked with magnetic tape, and you actually had to physically cut stuff up with a razor blade - talk about committing to your process!

If you're really worried about working yourself into a dead end, try using "Save as" as a version control system - every time you open a track to work on it, immediately save it as a new version - that way you can always revert to your last version, and all you'll lose is one session's worth of work.

- sounddevisor
(pro license #135791113)

I think that's the best way to speeding up production, it's just something I'll have to get used too. Just to clarify, I already commit to audio early though, this is different it requires another layer of commitment which is committing to you mix, eq, compression and panning reverbs etc before you probably have all the sounds in your song, so it's blindly mixing in some ways.

[dominicw78]

(My pro license expired too... )

The answer to this might come down to using sends as alternate submix busses.

For let take some basic stems just for example sake -

Stem 1 - synth rhythm/stabs/arps, leads and pads
Stem 2 - vocals
Stem 3 - bass mix
Stem 4 - rhythm mix
Stem 5 - Kick

You have the main mix going to master or whatever - personally I send this to an internediate mix down track so I can apply global inset effects to the mix and keep the master track for strictly post mix (wht you might call mastering) processing.

Now lets assume you have a heap fo sends used for various things - sends A and B with different reverbs, Send C with a basic delay, send D with a tape/dub delay etc, send E with a reverb/delay was for enhacing the odd snare clap etc or whatever.

Now lets assume that at some point in the track you want to do some kind of mush (fade to grey ie reverby dub and all of sort shit type thing) effect on everything except vocal and kick.
And somewhere else you want to do a filtered beat mash over rhythm, and vocal, but keep the bass and kick clean.

So you create two more return tracks and stick the mush effect on one, the beatmash and filter on another.

I would have my mix tracks as follows:
Stems 1 ro 5 submixes as above, these all mix into an audio track I might label as 'main mix' - you must disabale all sends on this!
The mush return has its audio going to another audio track call 'mush mix' - you must disabale all sends on this!
The beat mash return has its audio going to a track call 'mash mix'. - you must disabale all sends on this!

All of these have their audio to another audio track called 'final mix' - again you must disable all sends on this track else it does wierd thing to Abletons latency compensation.

I might use the cross fader assigned as follows:
Main mix on A, 'mush mix' and 'mash mix' on B so i can use the cross fader to transition between main mix and the effect submixes.
Ill use the mutes of the mush mix and mash mixes to select which one I want at any given moment.

In case it aint obvious - sends and return tracks are actually mixers in their own right - the send level from each track is like the main track level, so when you send from some instruments to a reverb, you are actually create a specific submix for that reverb anyway - this seems to be something that generally only people who have used hardware consoles pick up on - but works the same in software.

The downside fo doing multi-submixes to multi effect bussed is CPU use - you might need a pwerful machine for this - often Ill just bounce submixes to audio and use them for special partial mix fx - which also has the advanantage that you can take a submix, drop it into sampler to create a tape stop type effect with a pitch envelope for eg.

Thanks for all that info, your last paragraph relates to what I mean in when you are working with bounced audio of a stem. My point being that you have too predict/guess how to mix it early on or do it right at the end of the production line or your parts might not match due to being processed differently.

[dominicw78]

Just had a thought, chuck BT's Stutter Edit of the drum bus, problem solved!

T

[Khazul]

Thanks for all that info, your last paragraph relates to what I mean in when you are working with bounced audio of a stem. My point being that you have too predict/guess how to mix it early on or do it right at the end of the production line or your parts might not match due to being processed differently.

TBH - thinking back to some stuff ive done mroe recently - I have right from the outset taken the view that I was going to keep the core track simple and mix it down to a finished simple track but with the full intent of having got to finished stems, I would then treat it like a remix and work purely with audio with the complete freedom to chop tuff up, manglle it etc. Its actually quite a liberating approach - its stops you from getting over obsessed by effects when just trying to get the musical arrangement right, and once its time to mess with heavy fx etc, then it stops the temptation to go back and tweak the midi etc and instead just live with/mask the flaws etc.

The more times I approach a track this way, the more Im liking it, but then its an aproach that is probably well suited to dance music cross over type tracks where the scond pass is about taking somethig that isnt quite a dance track (ie largely rock, soul or even striongly classical inflenced or whatever) and making it some flavour of house/dance music track.

Naturally ive done thing tthe heavy way before too - a great example of this is somethign I did 4 or more years ago and in my head the damn thing still isnt done and dusted, and yet in the favours allmost done forms it perhaps the track that people like the most, so there is alot to be said for just setting up live like I outlines above, but without the different effect submixes and having got a track working as a bunch of stem - calling it a day, bouncing it, leaving it for a week or two, then going to work purely on the audio - ie actually remixing it.

[dominicw78]

Thanks for all that info, your last paragraph relates to what I mean in when you are working with bounced audio of a stem. My point being that you have too predict/guess how to mix it early on or do it right at the end of the production line or your parts might not match due to being processed differently.

TBH - thinking back to some stuff ive done mroe recently - I have right from the outset taken the view that I was going to keep the core track simple and mix it down to a finished simple track but with the full intent of having got to finished stems, I would then treat it like a remix and work purely with audio with the complete freedom to chop tuff up, manglle it etc. Its actually quite a liberating approach - its stops you from getting over obsessed by effects when just trying to get the musical arrangement right, and once its time to mess with heavy fx etc, then it stops the temptation to go back and tweak the midi etc and instead just live with/mask the flaws etc.

The more times I approach a track this way, the more Im liking it, but then its an aproach that is probably well suited to dance music cross over type tracks where the scond pass is about taking somethig that isnt quite a dance track (ie largely rock, soul or even striongly classical inflenced or whatever) and making it some flavour of house/dance music track.

Naturally ive done thing tthe heavy way before too - a great example of this is somethign I did 4 or more years ago and in my head the damn thing still isnt done and dusted, and yet in the favours allmost done forms it perhaps the track that people like the most, so there is alot to be said for just setting up live like I outlines above, but without the different effect submixes and having got a track working as a bunch of stem - calling it a day, bouncing it, leaving it for a week or two, then going to work purely on the audio - ie actually remixing it.

Wise words, me thinks. I think it's definitely a great idea to commit to your audio parts asap and in this case your mixdown. Unfortunately for me I have really been enjoying leaving my mixdown to the end and approaching in a traditional way. I have found that I have really been improving my sound that way and become much more pro sounding. However both these methods are opposing. If I can truly really on my mixing skills straight then it shouldn't be a problem, but I still like to be heavily influenced by all the sounds in the track as to how I mix them and I can't know how I want to mix them together until there all there waiting to be mixed!

I kind of jokingly mentioned BT stutter edit being put on the instrument busses as the solution. I have just downloaded it and OMG it's amazing and I think it might actually give you the best of both worlds.

[Khazul]

It was actually beats me's post in this thread (http://forum.ableton.com/viewtopic.php?f=1&t=157067) that reminded me that Ive kind of been doing thing simpler these days and getting on better with it in many ways perhaps.

[timothyallan]

I still have a valid pro license and have nfi what is going on here.

[macmurphy]

I still have a valid pro license and have nfi what is going on here.

i had mine revoked for being anti and bloody minded.

i couldn't be happier