peter47
ok guys some interesting advice here, i am a little confused, i understand there is "normalize" found in the process menu under "apply effect" and there to is the "gain" function, but there is also a "gain" knob at the top of the tracks strip and there is a "gain" automation function by pressing t in the highlighted clip then choosing clip automation-gain.. and moving the gain automation with ctrl up or down with mouse...so which "gain" do we use? also for reference i am looking at a groove 3 vid with Kenny Gioia, mixing in reel time..he is using reaper, he normalize the track in question then fine adjusts with gain to get a level in a rough mix..so this is a bit complex..it seems there is no right or wrong way???
Peter...learning as i go along....

Peter - Great addition to the conversation. I like the way it's going because it sort of getting all the cards on the table. I am hoping someone with some technical training comes along and adds their feedback. 
 
greg_moreira - I appreciate you getting involved as well. At the end of the day, this is art and we will all have opinions. I want to be sure we all gain knowledge and understanding from this topic.
 
I appreciate everyone getting involved in the conversation and putting out their ideas.
 
I am looking forward to more ideas, tips and advice to help us all progress in the mix engineering world. I really want to step up my game so that I can eventually step up what I charge for my work. I have to earn those stripes first. At the end of the day, I love what I do. I just want to sharpen the sword and have more knowledge to back the decisions I am along the way.

exactly.  Im not trying to say my way is the right way at all.  Just trying to explain why I do it, and its entirely possible that its the wrong way lol....or at least maybe there is a 'better' way.
 
as far as doing things the 'right way' goes....  I believe its pretty subjective.  In other words....  if you really love the quality of mixes that a certain guy turns out, then he is doing it the right way for you.  It would make sense to learn what you can about his approach and ultimately add your own thing to it.
 
there would be no good reason for you to emulate what im doing, for example, if you didnt particularly care for the quality or sound of something I put together.  Even if I could cite a scientific explanation for why I do XY and Z....still, its not right if you or your clients dont like it.
 
 
 
 
 
 

greg_moreira
exactly.  Im not trying to say my way is the right way at all.  Just trying to explain why I do it, and its entirely possible that its the wrong way lol....or at least maybe there is a 'better' way.
 
as far as doing things the 'right way' goes....  I believe its pretty subjective.  In other words....  if you really love the quality of mixes that a certain guy turns out, then he is doing it the right way for you.  It would make sense to learn what you can about his approach and ultimately add your own thing to it.
 
there would be no good reason for you to emulate what im doing, for example, if you didnt particularly care for the quality or sound of something I put together.  Even if I could cite a scientific explanation for why I do XY and Z....still, its not right if you or your clients dont like it.
 

greg_moreira - I guess the deal breaker with what we do is.... Does the client like it? if they do, then you have delivered them a quality mix.

DJ Darkside
 
 
greg_moreira - I guess the deal breaker with what we do is.... Does the client like it? if they do, then you have delivered them a quality mix.

no doubt!  I tend to ask folks what artists they like and who's record they would like to sound similar to.
 
Then when it comes time for them to listen to their mix....  I can queue up something they like as a reference just so they can see that their project is in the same vein of what they said they wanted.  Especially with younger guys who are newer to the whole game....and each member of the band wants to be louder than every other member....  it helps to ease their concerns.  

Definitely! I also ask for the client to send reference songs that I can work towards the sound they are after or at least get close.
 
The reason for this reason post is to get fine details about how people set up their mixes:

• Their gain inputs and outputs
• How to prevent clippings from clipping
• What levels the meters should be peaking at 
• When and when not to use normalization
• Bus routing ideas
• Panning techniques
• Reverb techniques
• etc...

The more technical side of things...

DJ Darkside
Definitely! I also ask for the client to send reference songs that I can work towards the sound they are after or at least get close.
 
The reason for this reason post is to get fine details about how people set up their mixes:

• Their gain inputs and outputs.  - If I am doing the recording, each individual signal is trimmed as hot as possible wthout ever clipping at the loudest points.  I'll ask folks to sing/scream/play as loud and as hard as they are going to play during the course of the song and set their trim at that point.  
• How to prevent clippings from clipping - if step one was done right, I get no track level clips anywhere.  "If" there were ever a point where there was a clip....as long as its not extreme, it can be fixed.  A quick digital pop from a drum hit peaking for example can easily be removed without notice with a good crackle reducer plugin.  Lots out there that seem to work well.   Each plugin has the input or output gain adjusted to make sure that the track does not get any louder or peak any higher from one to the next.  So, after 0 plugins or 10 plugins, its still going to sound as loud as the source material.  It may very well peak lower now though after some compression has squashed some of the transients.
• What levels the meters should be peaking at - as I was stating earlier, Im not looking for an initial level balance between instruments.  I want every individual track to peak at -1.  All those peaks are the transient material.  those initial hits and impacts and what have you.  Im primarily a rock and metal guy, so I try to preserve those transients to get all the punch and snap out of each individual piece.  RMS levels/audible volume is going to vary from track to track but a combo of compression and mixing with the faders is where I create balance.....   hopefully leaving the transients and keeping them loud.
• When and when not to use normalization - I only use it to get each track peaking at the same place as a starting point.  Again..  just my method and not entirely necessary just depending on how you like to approach things
• Bus routing ideas - lead vox goes to a lead vox bus with one treatment.   Harmony vox goes to a harmony vox bus so I can do something different(a different delay or reverb or eq or whatever).  Both lead and harmony busses output to an overall vocals bus where I can glue them together.  that overall vox bus outputs to the master.  Often on the overall vox bus, I use a send to run to a reverb or delay bus so that all vox summed can have a consistent verb or delay added.  I might sidechain compress in those sends to duck the level of the delays and verbs when the voice is singing.  
• guitars are simpler for me.  I treat each individual track and output every guitar to a single guitar bus.  Guitar bus outputs to master, and I use a send on the guitar bus to run to a reverb channel too if I want to add just a tad to every guitar.
• For bass, I usually duplicate the bass line track 3 times.  One is the original and is pretty well preserved as is, minus light compression and maybe some EQ to shape.  The 2nd bass track I parallel compress and squash the life out of it.  Kill all the transients and make the bass sound very round and full(and unnatural too).  I blend it in to taste.  The 3rd track, I high pass it around 250hz or so, so im getting out most of the lows.  I run it through and amp sim and add some dirt and midrange.  I blend this track in just to add enough dirt to give the bass some cut and growl.  I buss all 3 to a single bass buss and use a light compression to glue it all together.
• Drums are where I invest most effort if the whole kit was mic'd properly and I have enough sources to play with(I usually use 9 or 10 mics when doing a 5 piece kit myself).  Each track, drums and overheads and room(s) get outputted to a master drum buss. compression is light with a 30ms attack and moderate release to keep transients and keep it sounding lively and natural.  I use sends on the individual kick, snare, and tom tracks to send them to a parallel compression bus.  I squash the life out of them, just like the bass guitar, in that bus to make the drums sound really round and have depth.  I keep cymbals out of this bus so that they dont get splooshy with the excess compression.  I then output the parallell compressed drum bus to the master drum bus and blend in to taste just to make the drums sound rounder.  That master drum bus has a send to a drum verb bus also if needed to add some roominess.  Lastly....  if the drums arent present enough, I use drum replacer and create midi events for each individual drum itself, then run those into addictive drums and lightly blend kick, snare, and tom samples outputted to the master drum bus.  It doesnt take much.  Blending just a bit of each sample in brings the drums forward.
• Panning techniques - I usually keep it kinda simple.  I use track lanes to automate pans.  Like during the verse, guitars are 70% left and right.  During the chorus, they open up to 90% left and right.  Same with the overheads.  70% during the verse opening to 90% in the chorus.  Its subtle but it makes the song seem like it grew once the chorus hits.   If I use sends, I automate the send pans the same way
• Reverb techniques - also kept simple usually.  If I need it on an individual track(like only a snare) sometimes I just put it directly on a track.  If I want the whole bus of instruments to be affected by the same verb, I use sends and on the reverb bus I dial out all the dry in the mix and only leave the wet verb and dial in to taste.  Sometimes maybe side chaining so that the verb ducks a bit when the instrument or voice comes in.  Really only adding space between phrases etc.
• the next to last step is to mix to taste, then check the output on the master.  If the output on the master is peaking too high I select everything and hold CTRL and drag every single fader down the exact same amt simultaneously untill its somewhere between -4 to -6 on the master.  then export the whole thing at 24/44.1 with no dithering anywhere and open a new session to master the thing.  In the mastering chain, the last plugin is usually a limiter that can apply dithering.  Ill set the sample to 16 bit in the limiter and add dithering, then export the master at 16/44.1(with no dithering because it was done with my limiter plugin).
• the last step is to listen to the mix in a dozen different sources and then start over because I hate how some things sound in the car or in the home theatre or whatever lol.
• it sounds like a lot but it isnt when you have a track template built with all the plugins and busses and sends pre - added.  Make it a big enough template to support more tracks than you typically have and you're never having to stop and add any new tracks or anything.

The more technical side of things...

 
Just to give a more thorough explanation of what I do and why I do it..  I answered above

greg_moreira - Fantastic insight into your process. This might just become the holy grail for some of us. Thank you.
 
I think I will take notes from people's feedback here and start morphing my own process. Again, it is art! But we do need some direction.
 
Cheers again!

Another extension to this tip which I think could help me. I recently received some multi-tracks from a client and the audio is very low. Are there any tips anyone can provide for dealing with this? I don't want to boost the audio and add noise. Am I not sure this client knows about the levels they need to record at? From the looks of it, not!
 
This is a hip-hop beat that they want mixing. I don't think re-sending the audio files at a higher volume is an option. Has this ever happened to any of you? If so, what was your approach to dealing with it?
 

It would depend on the source material and how much 'noise' there really is.  Sometimes the noise is never noticeable when mixed with the audio, but it pops out during quiet parts.
 
If its only really noticeable when quiet...  I would try to cheat it first with a gate with a lookahead(like the sonitus one has I believe in sonar).  Put it first in the chain and adjust the gate so that it doesnt eliminate any audio, but promptly closes when the instrument sound goes away... effectively eliminating line noise during the 'silence".   That lookahead will keep the gate opening sounding natural.....allowing it to open just before the wave starts(so its not unnatural sounding because it pops in from a fast attack....and/or cutting off too much transient because you've set a slow attack to keep the popping out).
 
That alone might do the trick.  If you struggle to get a gate setting that works naturally, you could spend a little more time doing the same thing with track volume envelopes.   
 
Another thing to try is a noise reducer plugin.  Plop it right on the track that is the problem and let it study and learn the noise profile during periods of silence.  The suck the noise out of the whole track.  Careful though.  this can be destructive to audio and actually hurt your sound.  Its basically a try it and see if it works kinda thing.
 
If the noise is simply too obvious and you cant reduce it without somehow hurting the audio...  I would let them know and see what they have to say

greg_moreira - I guess I can try and normalize and see what results I get? Some of the audio is so low, I can even see it peaking on the meters in Sonar. I can hear it, or if I change the audio meter scale, then I can see it.
 
I need to educate the client on getting better volume per track. It seems as if they mixed the song themselves and then applied, or bounced the tracks and then exported them and sent them over to me that way? I was expecting much hotter levels.
 
I made the mistake of not mentioning it earlier and worked myself into a corner. Now they are expecting a mix and I don't know how to deal with it?
 
I'll try and normalize to -1dB and see how much noise gets introduced and go from there.