Mixing and Reference Tracks

[LFO.juice]

I’m still getting to grips with the art of mixing. First of all, it’s hard. Really hard. I’ve watched endless on-line content on the subject, I try my best and sometimes it’s great and other times it’s shite, even though I follow a similar workflow each time; the only consistency seems to be luck and inconsistency.

My question to you good people is this - if you are making music that doesn’t fit into a specific genre, how can you use reference tracks to polish the mix? And is matching to someone else’s results a good idea in the first place? I’ve tried mixing in mono against pink noise with limited success. It’s quite useful for a quick level balance half way through the creative process but I struggle with it on completed tracks. I find that because I have hearing loss, (anything above 7k I can’t hear properly,) and snares, claps, etc., gets drowned out by “hiss".

So I’ve reached this impasse. My workflow produces inconsistent results, pink noise is OK but I find it has limitations and I don’t have enough experience or skill to make it to the next level.

Any suggestions? I’d like to know if anyone else has been at this point on their production journey and how you managed to take your mixing skills up a notch?

C.

[bfk]

Gotta check your mixes on multiple mediums…

Me personally I go by combination of numbers and what the track calls for.

When I create musical sketches I just see what sticks

But for my more serious efforts.

I treat every sound and sound layer like a Lego and build a track LEGO by LEGO in the frequency spectrum. I use an eq for each sound layer and filter out the extra frequencies that just take up room in the spectrum…and leave the meat of the sound as a Lego to be used to build the rest of the track

[morphic]

I tried the reference track thing and that didn’t really do anything for me. IDK, maybe you have to find something that sounds exactly like your music to make it work? When I listen to other folks’ mixes, I try to understand why they work and go from there.

@bfk

Gotta check your mixes on multiple mediums…

Me personally I go by combination of numbers and what the track calls for.

When I create musical sketches I just see what sticks

But for my more serious efforts.

I treat every sound like a Lego and build a track LEGO by LEGO in the frequency spectrum. I use an eq for each sound layer and filter out the extra frequencies that just take up room in the spectrum…and leave the meat of the sound as a Lego to be used to build the rest of the track

^^^ This, exactly. All of the above.

I have learned that, instead of trying to make things sound good altogether at once, you need to first make sure the tracks sound good together in random pairings, or when there are 2-3 tracks happening in a similar frequency spectrum, so you can look at the masking that may be happening and clean it up.

I also like very rich compositions, with many layers. So I often remove notes here & there when I do the final mix to make it all sound cleaner.

Like you, @LFO.juice , my hearing is not exactly great. To remedy the shortcomings, I pay a lot of attention to individual levels in the mix console, see how they fluctuate in different parts of a song, so I can tell if something’s too loud/low.

[Artificer]

I’ve never had much luck with reference tracks, except as broad inspiration. The feel and the vibe is what I get from them, not the technical details. Like you, I’m not making music for a broad consumer base; if I was trying to mix Beyonce or whatever, I’d probably have to play by the rules a little more.

Years ago, a friend told me “mixing is the art of turning things down, slowly”. I still think that’s a pretty good encapsulation for how I approach it. You make sure frequencies in different tracks aren’t stepping on each other’s toes, then you set levels to your liking.

I think the biggest leap I made with mixing was when I started to think about the frequency spectrum from the minute I started a track. EQ, for me, starts with sound design. If you have a track, whether it’s a mic’d guitar, a softsynth, or some busted-ass sample you recorded with your phone, it’s going to live in the mix and the time to start thinking about how it’s going to play with others is now, not later. Given the wonder of non-destructive editing, nothing you do up front has to be final, but I always start early EQing and getting things to sit in whatever mix I have and adjust as I go with a final pass at the end.

After that, it’s literally just turning things up or down. If you’re EQing from the get go, there’s little danger of mud in the mix and you’re pretty free to be creative with the levels and just make things sound good to your ear.

Optionally, once it’s just about perfect, slam it with a brick wall compressor undoing weeks or months of work to make sure you’re the loudest goddamn track on the playlist.

[ddsot]

All great points and ideas above.

I have evolved from a trial and error, hit and miss type mixing journey to a point were I’m a bit more comfortable with it and pleased with my results. But I still bounce out mix revisions and listen in the car or through some cheap ear buds to see how its working. Make some notes to make adjustments when next in the session. This can take time and If i’m getting above 5 or 6 mix revisions. I know im over doing it. And have to ask myself if this time an effort is making any difference at all.

I can’t really offer sound advice from this viewpoint, rather just share some of my experiences that I have used to try and avoid the revision procrastination cycle…. Based on using computers based DAW’s like Logic, Protools. Using inexpensive headphones and monitors in an untreated spare room at home.

I too have some high frequency hearing loss and constant tinnitus from years of playing drums in rock bands. But even with these limitations, I still really enjoy mixing and mastering. And getting things to sound good. But if I had the money, I would be better off getting someone else to do it.

I found treating the mix session as a separate process helped me get better at it. Sure you want to get things in a good spot when composing. Balancing volumes, pans and Eq or processing. Clean up and organise the session. Even bounce out each element as audio tracks to use in a new session. And then create sub mix groups (Aux channels) . Drum elements, (kicks, snares, toms) Percussion Elements (high hats, cymbals, shakers) Bass elements, (bass synths, bass guitars) Mid range elements (guitars, leads synths, pianos) Other elements (pads, samples, sound effects). Some room for being creative with this. Also FX and automation on the groups can also be applied with this approach. Nothing really new with this approach TBH. Just a technical set up process. Creating template helps.

Having the control over the group elements can help find the right space for each. Finer work can take place on the individual tracks in these groups. Applying heavy processing and stacks of plug ins. In my mind, tells me something isn’t right with the source sounds. Might need to go back to the source here, unless it’s a creative decision you want to stick with.

Regarding reference tracks for mixing. If you had a track that you like the sound of, that you want to emulate and try to match. Consider this.

Is the track you want to match Mastered or a professionally released track? If so its near impossible to try and match this via mixing. Mastering changes the audio into a glued together, frequency rich, detail popping out sound.

Mixing is getting your awesome track to a point ready for mastering. Using your creative vision to shape the compositions into a musical story. And knowing that it will enhance further with a bit of mastering processing. This is important to factor into some of the decisions and technical aspects you make with the mix. Again, that’s part of my thought process.

Metering tools, EQ and frequency analysers, goniometers & correlation meters can all assist on the Stereo Output, to guide some of the decision making. When your eyes can help with some decisions. Also, Putting a mild limiter on Stereo Out can also give the perception of how it may work when you get to the mastering stage. Turning off when ready to bounce out.

I recently downloaded Sonar works Sound ID. (was free 14 day thing) and created an EQ curve for my headphone model and monitors. Copied this EQ curve by running white noise and a frequency sweep through a match EQ plug in. Saving the curve and then using to referencing on/off the Stereo Out. As it pushes the mix through this EQ it can be perceived with a flat EQ response. That will help highlight potential issues when played back on different sources. Eg the bass is lost or to much top end in the snare, piano hurts my ears….

I use this EQ curve to reference on/off when mixing and mastering. It sounds like shit when its on. But it does help identify the frequencies that are out of balance or really sticking out.

Not sure if any of this makes sense or will help. There is plenty of info online to trawl through. Some of it useful, most of it not… Trust your decisions and creative vision. Back yourself. Take a break if it is getting frustrating. Having others input (other artists whose judgment you value), and feedback is valuable. Mix when alert and awake. Ear fatigue is a real thing. This might sound weird, but if I’m going to mix or master a song that I’m really trying to get a good result on. I might wear ear plugs for most of the day before doing some mixing. I do work in a factory environment though.

Sorry for the wall of text.

Peace

[Mecha_Twitchy]

Lots of complex paragraphs… Good, but simple is all that is needed

There is no “right way” to do stuff - just listen to the sounds and drop the level of unpleasant frequencies until you like it.

Mix in the morning, and make sure your volume isn’t too high or too low

Try and shift your focus around in the track/s, but do a bulk of your listening on the master channel.

Always cut unnecessary lows to leave room for your sub (and don’t double up on sub-bass pitch unless it’s a perfect fifth or octave up!)

Make sure the key instruments you want get the most in both the resonant/fundamental frequency and overall volume. If there’s two snares with the same fundamental frequency, lower the level of the fundamental of the LEAST prominent snare, for example.

On that, layer a bunch of shit if you think something is weak. I always blend snares for dubstep tracks.

Remember: Mixing is effectively blending the sounds, regardless of if they collectively sound tinny/etc. Mastering can get rid of that annoying timbre if necessary.

[LFO.juice]

A big thanks to everyone’s contribution; there are a lot of nuggets here - I’m very grateful.

I try to listen to my mixes on lots of different consumer products and it never ceases to amaze me how much variation there is. I came across a guy recently that recommends concentrating hard on mixing the mids. If you EQ out the highs and lows, work on the mids and mix in mono he claims that you stand a better chance of the result sounding good across multiple mediums because that’s where the manufacturer’s put in the work to reproduce those frequencies. Admittedly he was using a reference track but I wonder if the methodology could be modified to work with pink noise? Worth an experiment.

Like @morphic I struggle with reference tracks. I suppose if you’re cranking out sound-a-like clones then it’s OK but if your trying to be genuinely different and creative then it’s probably not a good move.

C.

[White_Noise]

@Artificer

I think the biggest leap I made with mixing was when I started to think about the frequency spectrum from the minute I started a track. EQ, for me, starts with sound design. If you have a track, whether it’s a mic’d guitar, a softsynth, or some busted-ass sample you recorded with your phone, it’s going to live in the mix and the time to start thinking about how it’s going to play with others is now, not later.

This is huge. Part of the reason genres coalesce into certain sounds is that when you have a big four on the floor kick drum for house music, there are certain frequencies that are just taken up at certain times now, and so you’re physically limited to what else you can do. I’ve grown to appreciate that more as every project I open with grand intentions falls back into a house beat these days. Same would go if I was doing rumbly techno, fast DNB, atmospheric ambient, etc. You just grow to appreciate what sounds do and don’t work in certain genres from both a technical mix perspective and an aesthetic perspective.

@ddsot

Mastering changes the audio into a glued together, frequency rich, detail popping out sound.

Mixing is getting your awesome track to a point ready for mastering.

Yes and No. Agreed, if you learn mastering or you are using one of the services (I’ve consulted on the development of some of them and they’re not inherently bad, but you need to understand what they are doing to get the best out of them) or best case if you want to send your work to someone like me… then yeah the master CAN change the sound of the track. BUT, it is best if you get the sound you want in the mix as much as possible. This isn’t always possible, but I find I get the best mastering results for my clients when it’s possible to have them fix issues in the mix. Sometimes that comes from things they didn’t realize were wrong until everything gets turnt the fuck up in the master, sometimes that comes from things they weren’t sure what to do until they heard other tracks mastered, and sometimes I catch things that they can’t hear. In all those cases while it might be possible to patch things up at the master, the results will always be better if we go back and tweak the mix to do what is desired. 100% of the time that is true. It’s also true that the master changes the sound of the mix, 100% of the time - so you’re not wrong to say a reference track that is mastered is unfair to compare yourself against.

I don’t use references a lot of the time personally. I tried the pink noise thing for a while and I never really liked how that sounded either, I always had to stray pretty far from the pink noise target (especially for my drums) to get things to sound really good.

I think a good place to start is with the mids in mono, as your most recent post said. I personally use ISOL8 now, which is free IIRC, and adds the benefit of being able to solo other bands as well (helpful for checking the lows). If you make this a habit with your mixing for a year or two, you’ll internalize what stuff should sound like and not need to reference things in soloed bands as much. I wouldn’t write with the whole mix bandpassed, but get into that habit with switching to a bandpassed signal early in your mix and you’ll find things come together more easily later on.

[ddsot]

Its good to the hear some knowledge from someone who works in audio for actual clients. Thanks

I use ISOL8 also. Usually for the band pass purpose. But sometimes creatively also.

I need to get my head around the mix in mono and try research the pink noise style referencing

[KvlT]

@Mecha_Twitchy

Lots of complex paragraphs… Good, but simple is all that is needed

There is no “right way” to do stuff - just listen to the sounds and drop the level of unpleasant frequencies until you like it.

110%.

If I start dialing in specifics, there goes my creative juices (not to be confused with over-engineering the shit out of my rigs - that makes it fun ). If the track sounds good to me and it was fun to make, I’m happy. Boring and sterile workflows kill my creative energy more than anything and make me want to stop making music.

Whether or not that makes a good mix is debatable, but I think the sacrifice is sometimes worth it

[morphic]

@White_Noise

This isn’t always possible, but I find I get the best mastering results for my clients when it’s possible to have them fix issues in the mix. Sometimes that comes from things they didn’t realize were wrong until everything gets turnt the fuck up in the master, sometimes that comes from things they weren’t sure what to do until they heard other tracks mastered, and sometimes I catch things that they can’t hear. In all those cases while it might be possible to patch things up at the master, the results will always be better if we go back and tweak the mix to do what is desired. 100% of the time that is true. It’s also true that the master changes the sound of the mix, 100% of the time

Exactly. The first time I worked with a professional mastering engineer, it became instantly very apparent to what extent mastering will “discover” any present mixing fails.

I liken mastering to a magnifying glass that brings to the surface what you either didn’t realize was wrong in the mix or what you ignored, thinking it would be OK. Personally, I won’t release anything that wasn’t professionally mastered or at least mastered by someone who has the qualifications to work as a mastering engineer.

I only “pre-master” stuff myself with Ozone to get an idea of what my mix could sound like. I guess this is my own reference track, which I then send to the mastering engineer alongside the tentative final mix, to give them pointers regarding what I’m trying to achieve. After that, there is a certain back & forth to fix the mix according to the engineer’s recommendations and/or what is clearly not right.

[carlos_pG]

Hi @White_Noise, are you a mastering engineer? I sat in on a mastering session once, a while back, and it was pretty interesting to hear the results, both subtle and more drastic.

[White_Noise]

It’s not my day job, but I have paying clients. I think of it as a “side hustle” that I enjoy doing and it pays for a few high end plugins every year.

It’s definitely eye opening when you see how we actually think in mastering, but that’s off topic