Having trouble getting the mix just right!

Kewlpack

New member
Hi guys -

Brand new to HR BBS. Gotta question I hope you can help me with. I'm relatively new to mixing/home recording and ripping hair out by the handfuls trying to get this stuff to sound great.

I have been working on mixing a song recently that has 3 guitar tracks and a simply piano/drum backing track. I have attempted to mix about 10 variations but none have the instrumental distinction or clarity that I want (I hope for!).

Would you please listen to the song (it's short) and try to identify what is causing the problem(s)?


What things can I try in order to get this to sound "great" (I've only been playing guitar for about two years so please forgive any goofs)?

Hoping for some great insight! Please let me know how to fix the mix!
~A~
 
Oh... lots of no-no's...

Mmm - good stuff.

This part jumped off the page at me:

I strongly suggest you adopt the subtractive EQ approach - cut instead of boost. If there are too few highs, remove some mids or bass to shape it .... and.... Overuse of effects results in muddy, poorly-defined mixes, so much like EQ - less is usually more.
:cool:

I laid down all of my tracks using the amps effects! Hmmm guess that is a NO NO for recording (i.e. do it afteward)! After that I was using the Parametric EQ to bump various frequencies (doh). To be honest, I'm MUCH more comfortable with a graphic EQ than this Parametric beast (I don't understand it much - HELP!)... :eek:

Guess I gotta a whole lot of playing to redo... :(
 
I really don't think it sounds all that bad. My suggestions (and they are just that, everyone will tell you something different) would be to try moving the piano and the guitar to more opposite sides of the pan. If you still want stereo images for both of those instruments, try putting a mono guitar on the right, touching it with a little verb, and shooting that verb to the opposite side. Likewise with the piano, but in the other direction. Also, sometimes effecting the entire mix with a very minimal reverb or compression can help everything kind of sit down in the mix just right.

Practice with your parametric. I have a huge preference for parametric for channel eq-ing. I like it better. Graphic eq's to me, are more useful for whole mixes.

I would be inclined to bring up the snare a bit, and maybe make the whole drum kit a tad wet. Make the drum kit kind of frame the mix and put everything else inside the drums (if that makes any sense) Overall, I think you did a good job. Again, I really don't think it sounds nearly as muddy as you say.
 
Thanks PP,

I'm a n00b on all this mixing stuff so please bear with me.

There are two specific tracks that just aren't making it to my ears:
-- Bass EQ'd guitar (because I don't own a bass)
-- Rhythm Acoustic

I will more than likely retrack the "bass" as I found I get a MUCH fuller tone using my acoustic guitar.

The rhythm track just gets completely lost after the intro and before the outro.

Two problems with moving piano and drums - they are a backing track (ala MP3). The track itself is pristine and well defined... but my staging options are sorta limited.

I don't know how to send only an effect to L or R in ProTracks Plus. I normally double the track, apply the effect, and then move it 100% L or R and drop the level of that track by 4-5db...but I don't have a clue other than instinct here. Could be another tone killing error on my part. What is the 'normal' process for doing what you suggest with the Reverb effect in the opposite channel?

After reading that article on mixing/recording - I definitely broke some rules. I may try retracking everything dry and then subtractively EQ it... and only then think about effects... :rolleyes:

Is there a good online guide that goes into how to properly use the Parametric EQ without burying me in too much technical mud (I"m a programmer and designer, and can deal with tech, but I'd love a simple primer to get started)??

When you say "make the drums wet" what exactly does that mean?

Okay - lots of questions here... please let me know. And thank you for your help!

~A~
 
By "making the drums wet", I basically mean a tad bit of verb or something. They sound excessively dry to me. But, as always, be sparing. It's too easy to overdo effects in general.

What is the 'normal' process for doing what you suggest with the Reverb effect in the opposite channel?

Well, I don't know your exact software, but in PT LE, you can create an aux send. Send the original signal into your designated plug in (again, I don't know what you're using) and set the effect to completely wet (nothing but the verb). Then, you should be able to mix back in the completely wet signal with the original dry signal and pan however you like.

I've been stuck without a bass as well and been able to somewhat successfully mimick one with a guitar and some eq. It works somewhat until you realize that...doh! Basses are one octave lower than guitars. So by this reasoning, if you can find a way to drag that sound down an octive, it should sound even better.

I'd say do some google related research on parametric eq's. Sometimes it's easier to get at what you're trying to fix better with parametric. The basic steps that one could do are...boost a frequency (if you were looking at a graphic, it would be the same as creating an upwards hump), sweep it around until you find the annoying spot, then if you want to take it out, just cut that frequency range. The "Q" function you'll see on parametrics is how wide that "hump" is. That's basically it.

If you want some real specifics, I'd say the lead guitar could spare some of the low end and general volume to maybe make a little more room for the other stuff. The acoustic might stick out if it was a bit brighter (more highs), but not brighter than the cymbals. Another thing I try to make myself do is realize (when it's my music Im working on) that if a certain part isn't really bouncing out of the track (in the mixing stage mind you), or seemingly doing much in there, maybe it's a musical problem in nature. Listen to your track without the acoustic and see how you feel about it. Or maybe, keep it in the beginning and the end. Never be afraid to pull out parts of your own stuff. And as a matter of fact, I think if you pull something completely out and it doesn't feel as if it's missing, then you've definately done the right thing.

Again, you'll have a hundred different people tell you a hundred different things. These are merely little things that I do to help my mixes. It's always best to experiment for yourself. That's the best way to learn most of the time, and it's half the fun anyway. Best of luck, and sorry Im so long winded.
 
This is my little mix method:

1.) First thing listen to your recorded material and make some decisions. What needs to be up front? What needs to be in the background? What are the important parts of the mix? Have you recorded everything you need in a clear, high quality manner?

IF YOUR RECORDED TRACKS AREN'T UP TO SNUFF GO BACK AND REDO THEM!

Nothing slows a mix down faster than tracks that have a lot of issues. If it's noisy, pops, bad performance or whatever you owe it to yourself to fix it before you mix it.

Unless you are getting paid by the hour you don't want to play the "fix it in the mix" game. Trust me, I've polished as many turds as a toilet at an overeaters anonymous seminar, and it is never fun. You will kick yourself and end up re-tracking it anyways, so why wait?

2.) Set levels manually for a rough mix in **MONO** (don't stereo pan yet). Don't touch any eq or compression at this point. KEEP IN MIND THAT YOU SHOULD MIX AROUND YOUR *VOCAL* LINE OR MELODY (if an instrumental song). At all times remember that songs are to sell a vocal performance and everything should be subordinate to it.

KEEP IN MIND MIXING IS EASIER IF YOU START WITH THE "CORE" ELEMENTS OF A SONG AND NAIL THOSE FIRST.

Thus, start with the main percussion (which may be all of it), the bassline, the main melodic instruments, vocals, background vocals, primary guitars--anything that is the strong parst of the song. Things like background noises, samples, special effects, random noises and the like should be *MUTED* and put on the backburner until after you complete this entire process.

The reasoning is that if your core material sounds great, you can fit the 'support' stuff *around* it and the song will still sound good. After all, it is bassackwards to have the greatest sounding pad that just rules if the vocals and drums are totally buried by it.

Thus--MIX FROM MOST IMPORTANT ELEMENTS TO LEAST. LESS IMPORTANT STUFF MUST WORK AROUND MORE IMPORTANT STUFF. THE VOCALS ARE ALWAYS THE MOST IMPORTANT ELEMENT.

3.) Once you've gotten a rough mix going on, listen to it again and note any deficiencies--is the low end tight? Is it muddy? Is something that is important not popping thru? Is something popping thru too much? Does a part go from too loud to too soft? How's the high end balance and clarity? Does the midrange sound cluttered?

4.) Now that you've mentally assembled your laundry list of mix complaints, it's time to do something about it.

5.) First clean up your low end. Run an EQ on all the tracks with nothing but a HIGH PASS FILTER on it. Stuff that should be part of your low end like kick drums, bass and so on get a HP filter around 35 or 40hz; stuff that shouldn't be cluttering your low end much like strings, sweepy pads should get a HP cutoff around 70 to 200hz. Vocals can be cut off around 150hz pretty safely. Guitar gets cut off around 70hz in general. Remember: you don't want excess garbage cluttering your low end--this is one of the main sources of audio mud.

IN GENERAL THE MORE BACKGROUND A TRACK IS THE MORE YOU SHOULD REMOVE ITS LOW END!

I've had pads start rolling off at 400hz before because all I really wanted was a little midrange color and some upper harmonics (so I boosted them around 11khz or so later on). Heck, on high hats I typically roll off starting at 500hz for that crisp, clean and transparent high hat sound.

6.) Now that things are looking clean on your low end re-examine your VOLUME issues, which means listening and start grabbing for the compressor.

7.) Stuff that still seems to pop in and out of the mix need compression--target these and compress them so that their volume stays put. (Read my compression tutorial for additional details.)

IN GENERAL I COMPRESS **EVERYTHING** IN MY MIXES AT LEAST A LITTLE BIT.

I am a big believer in fairly low compression ratios though. 2:1 on a lot of things. I always lightly compress analog synths because they are very erratic; if it's an analog synth doing a bassline I will squish it pretty good. In general VA, softsynths and digital synths need **LESS** compression than analogs, but let your ears and mix decide.

8.) Stuff that should be prominent rhythmically like kick and snare definately get some compression as well. Make them slam hard as hell.

9.) Get your low end instruments thumping be it bass guitar, synth or whatever. Make that low end steady, yet punchy. Try not to have more than 3 "low end" elements if you can.

10.) Now that you've gotten levels to be pretty consistent re-listen to the material critically and ask yourself--what needs more seperation, and what needs more integration?

11.) Now it's time to EQ. A lot of mixes sound tinny and thin because of overuse of EQ. If you've gotten your volume levels sounding great manually, and then used compression to make it even more tight, there shouldn't be a whole lot of EQ that you need to do.

12.) First thing--listen to the mix and try to identify weak sounding areas that sound BAD. Is there a little fizz to the guitars? Kick drum a little muffly sounding? High hats sound clangy? Prepare another mental list....

13.) Now use *subtractive EQ'ing* to locate and eliminate these discrepancies. Use the narrowest and smallest cuts you can get away with to bury the offending freq's in the *mix* (not solo'd by itself--always, always look at things in the context of the mix). When you have eliminated these frequencies (and there will probably be a few, perhaps none if you're lucky, sometimes on poorly recorded stuff there will be some in almost everything) we can move on.

14.) Now that the shit frequencies have been zapped listen to the song again and listen to see if the seperation/integration issues have been taken care of. Sometimes you can get lucky and a few problems will work themselves out; if not, the overall quality should have gone up a few notches.

15.) Now it's time to EQ for *SEPERATION*. Listen to the mix and figure out what elements are fighting for space in the low freqs, low-mids, midrange and high frequencies. Choose the one that you want to be more dominant in that frequency band--now go back and slightly cut that track in that band, while (sometimes) applying a slight boost (we're talking 1-2db's) to the dominant track. Keep doing this until you've gotten them all. Re-listen to the track.

16.) Now you want to integrate some of the elements so they work together more. An example is bass and kick drum. But how do you integrate AND seperate these sounds? Easy--give them boosts that are close on the lower end of the spectrum on or near the same frequency (for example: kick drum at 80hz with a boost, bass synth at 100hz with a boost); next move up into the midrange and boost one element someplace and the other one someplace else (such as boosting kick at 4khz and bass synth at 2khz). Play around with these techniques until you have things really cooking.

17.) Now listen to the WHOLE mix. Focus on the different frequency bands, paying special attention to the high end. Does the bass sound tight and clear--with the bass and kick working together yet with distinction? Does the voice mix well in the midrange with the other instruments? Is the high end crisp and clear, but not domineering and tinny? Can you hear the "air" and upper harmonics of the instruments in the over 10khz range?

18.) Now use EQ positively to *add* any of these missing characteristics... such as boosting some cymbals at 12khz, or a string synth at 9khz or wherever there is a bit of a pocket that needs filling, or place for something to shine a bit more without queering the mix.

BE CAREFUL WITH SUB 1khz BOOSTING. Too much boosting in this area can mess you up... too much cutting will give you a thin sound. This is a difficult area to master. When in doubt, leave it alone for the most part.

19.) Now, at long last, STEREO PAN your tracks. Try not to weight any one side more than the other. Keep low frequency or primary instruments centered, or close to center. Bass and kick should always be centered... and snare as well. Give a nice panorama of sound but don't get carried away. Panning over 50% is often too much. Panning less than 30% is what I do most of the time except in specific circumstances like mic'd drum overheads (due to stereo bleeding) which I'll put at 60-75% or so.

20.) Correct any deficiencies that may have arisen from the stereo panning. 80% of the time if you've done the steps pretty good you won't have any correcting to do. The song will suddenly have "mixed" itself when you stereo pan everything.

21.) Now go back and fit the less important elements into the mix. DON'T TOUCH THE CORE ELEMENTS--make the less important ones fit around them with compression and eq.

22.) When you're done, put the mix down for a day or two and go back and listen. Correct anything you don't dig. Compare it to CD's you like and see if it measures up. Make sure it's not too bright of a mix, make sure there is good low end, make sure it doesn't sound muddy, make sure the midrange is well defined, punchy and clear.

Most of all--have fun. There is no right or wrong way to mix YOUR songs.
 
let me gues....., You are the lead guitarist on the song right???
It's waaaaaay to loud(the lead guitar)

Pan the rythem guit to the left and piano to the right (or other way around)
After 16 sec's the lead guitar is kicking in and eating up al the bandwith
maybe a bit of reverb here and there.
I think the song itself is ok ....
Welcome to the club ;)

Remco
 
Cloneboy Studio said:
...
21.) Now go back and fit the less important elements into the mix. DON'T TOUCH THE CORE ELEMENTS--make the less important ones fit around them with compression and eq.

I enjoyed reading your article Clone Boy and thought your approach to setting up the Core Mix first then sprinkling in the other elements last to be pretty interesting. I think it would help me keep the focus of the song in perspective a little differently than I normally do - I'll give it a shot next time. :cool:
 
kewlpack, great playing on the tune. Also, sounded very good to me, but you know what you are looking for in it. Just don't be too hard on yourself.
Cloneboy, great article. I'm really new to this, and have been having the same trouble, just don't play as well. This info will REALLY help me out. Thanks so much.
ED
 
Dreaming of Heaven - final mix - please give it one last listen.

Okay - I studied up on mixing (I had no idea there was so much to it).

Please have another listen - this is the last one on this song. I think it finally has balance and instrumental distinction - at least on my PC 4.1 system and fairly good headphones.

Thanks for being my guinea pigs ;)
<-final version (for now!)

<-previous version (if you want to hear the difference).

Comments are encouraged (mostly looking for comments on sound quality).

Thanks a billion for all the good tips and process!

Blessings.
~A~
P.S. I played the Rhythm Guitars, Lead Guitars and the "Bass". The piano and drum track were something I found out on the web. I know the lead tone is a little squashed, I didn't feel like redoing it yet - coming down with a cold. :(
 
About "cutting" via EQ

When you guys say "cut" - what do you mean as far as "amount"? Do I drop the entire frequency (Param EQ) or just lower it a smidge? What's the rule of thumb?

Thanks again - man you guys are GURUS!
~A~
 
Cutting means down basically. You can either cut a specific frequency, or you can cut an area of frequencies. As far as how much, there is no set solution at all. Highly dependant on the situation. It can be minimal for recording for the most part, and severe if you're say....pulling out feedback in stage monitors.
 
kewlpack, much better sounding mix. A lot fuller, and even. You could probably play with it forever, but I say it is sounding very good. These threads help immensely.
Have you burned it to a cd, and listened to it on a home stereo system, with quite a bit of volume? That's when I really hear what I need to work on.
Anyway, good job.
Ed
 
Thanks guys. Really. This is my first attempt to do anything I would ever let anyone listen to that isn't some "cover tune". :o

I have to mix on my PC (on a TIGHT budget) through a fairly nice Creative Labs 4.1 via SoundBlaster Live! card. I throw on the headphones when I think I have it close to listen for weird things.

One of the very first things I do once I have a mix I think might work is to burn it to audio CD and throw it into my big 5.1 home theater (Yamaha, Paradigm, Jamo, and Kenwood components) - medium sized room, vaulted ceiling.

I play the song through the Dolby Pro Logic setup, paying particular attention to the center channel (Paradigm CC350 - handles all frequencies really well).

Next, I turn off the DSP and listen to straight stereo signal through the mains (Jamo - uh can't remember model) for overall sound quality.

If everything is pretty balanced and can crank up without something going whacky, I'll jump in my van or car and see how it sounds - one has a decent stereo system, the other is only fair.

This mix seemed to do okay in all the above. It isn't "wow", but it works.

What would really be handy now are answers to the following:
1) How do I analyze a mix once I "think" it's okay (some guy did a spectrum analysis on it and asked me a load of questions about certain frequencies that had problems)?

2) When I do an analysis like that - what in the world do I look for as a rule of thumb??

3) Once I know what the frequency issues are and what my "ideal" is supposed to be, do I just jump into the Parametric EQ and cut? (This was so hard for me to deal with because I don't know how to cut multiple frequencies at one time with the Param EQ in my software... I have to do it one "guess" at a time - is there a way to do several at one time?)?

Again - your wisdom and experience are greatly appreciated.
~A~
 
chadsxe said:
Do you ever pan to identical guitar parts past 50%?
Actually the acoustic is an identical cut/paste of the rhythm guitar panned hard L/R. I applied a slightly different reverb and flange to one side (I think - it was getting late).

I have also had to dramatically drop the level on the bass track (which I'm using my Ovation Acoustic for with bass cranked and the mid/high cut 100%)... is that normal when doing this?

Thanks,
~A~
 
Cloneboy Studio said:
Unless you are getting paid by the hour you don't want to play the "fix it in the mix" game. Trust me, I've polished as many turds as a toilet at an overeaters anonymous seminar, and it is never fun.


I don't care who you are... THAT was funny right there.... :D
 
Kewlpack said:
1) How do I analyze a mix once I "think" it's okay



Go away from it for a while. Take your mix, put it aside for a week or two and work on some other stuff, don't listen to it at all. Then you'll get to practice on other stuff, honing your skills, and with the new skills you've gained on your new song, you can revisit your old song and hear the problems with it, and hopefully you'll be better equipped to deal with them from practicing with another song.

I don't know about spectrum analysis or anything like that, it sounds like mixing by sight, and I would'nt do that. If it sounds good, go with it, if not, tweak it, if it sounds good once, and you listen to it a month later and you don't like things, fix them. Practice practice practice, I know it's a totally overused saying, but it's true, that's absolutely how you learn.
 
Back
Top