Mixing Discussion

[-mjk-]

Welcome! This thread is all about approaches to mixing. Feel free to tell us about what works for you!

 

[-mjk-]

@Bambi, thanks. We're getting away from the main topic, and we really should have a topic dedicated to mixing as well, so.... here it is.

I find a lot of parallels with cooking and mixing . My favorite type of cooking is outdoors with Dutch ovens (as well as wood smoking). I noticed that many of my dishes had the same flavor and I began to get really bored with my own food. What was going on? I had forgotten about flavor differentiation and matching. You see, as I was cooking, I was unconsciously adding spices to the dishes. The problem was that I was adding all the same spices to every dish, lol. There are so many spices that I love, and I couldn't help but want them all.

The next time I went to cook several dishes in different Dutch ovens, I sat down and really took a hard look at the main ingredients and I chose spices that I knew would specifically compliment those ingredients. After all, I did that with the ingredients themselves. For example, I had matched any fruits with a particular meat (based on what I learned from the Chinese cuisine chefs). I would mix peaches with beef, and cherries with pork (but those are not hard and fast rules!). I also chose meat and vegetable combinations that would match. So, this time, I flavor matched my spices and I resisted the urge to throw other spices into the pot. I kept it to what I knew would match. The results were excellent. Each dish was distinct and individual.

When you have limited FX (hardware devices) you often have to record them onto the tracks so you can free up the FX device for use later on in the mix. This means making adjustments to the FX so they match the tracks that are being recorded. Even if you have unlimited FX (plugin instances) you still have to make individual FX adjustments to match each track or else it all sounds the same. One has to resist the urge to throw everything at it too (i.e. every FX you have on every track). That will keep your tracks and your mixes distinct and individual.

 

[BazzBass]

some good analogies there @mjk

With damaged hearing I have trouble hearing any change after adding compression and EQing etc. I can spend hours adjusting things without hearing any great improvement, then I go back to no fx and it sounds the same. It's like adding spices to food without having tastebuds lol

But as long as I can capture a performance cleanly, I can rely on other people's ears to tweak the mix. I try to get the band to help me mixdown but after ten minutes they usually say "it sounds good let's go home" lol

 

[-mjk-]

I'm sorry to hear that @BazzBass. You are not alone though. Many people suffer various degrees of hearing loss yet still produce fine sounding musical products. By carefully watching the meters and using Izotope's Tonal Control plug-in, you can usually get things pretty close. And, as you say, extra ears helps a lot even if you don't have a hearing loss.

 

[Mark Richards]

I always begin with the end in mind.

For me, a good mix starts with good tracking. My overall objective is a well-balanced high-fidelity final product with a wide dynamic range and a sound stage with depth and width.

To that end, I use very analytical speakers and headphones, and I use track sheets to plan and carry out the recording and mixing process.

In broad strokes, here's how I set up a recording and mixing session using my DP-24 portastudio:

Tracking

Since the DP-24 track faders don't control the input signal, I set all faders and the stereo master fader to Zero/Unity Gain.​

I set my monitor level to produce a 75 dB SPL at my listening position.​

My goal when tracking is to record each track to a close approximation of the level/balance I want it to have in the final mix. That means paying close attention to where mics get placed; watching for safe recording levels on the track meters; etc. (this is where track sheets can be very helpful).

My tracking levels top out at -12 dBFS peaks. To retain maximum flexibility when mixing, I don't apply any FX or EQ when tracking, and use minimal, if any, compression.
​

Mixing

The monitor level stays set to produce a 75 dB SPL at my listening position.​

The track faders start at Zero/Unity Gain, the stereo master fader stays at Zero/Unity Gain.​

Where needed I tweak the individual track faders down from Unity Gain to get my final balance for the stereo master. The individual track fader positions become my visual points of reference, making the stereo bus meters and my ears the only other references I need.​

Once levels and balance are set I solo individual tracks to confirm how the track fader position correlates to the soloed track as shown on the stereo bus meter. I aim for peaks of -12dBFS for the overall stereo mix.

I add/adjust effects, EQ, etc. to the individual tracks as desired, and tweak the balance as needed.

I do a dry run (or several) to practice the Mixdown if it's necessary to make tweaks to individual track levels as the mix progresses, relying on memory for the changes (or track sheet notes if things get complicated).​

 

[Bambi]

Hello everybody, I hope you're doing well. MJK, it's a great idea to start a thread about mixing. Very illustrative your cooking examples. I can tell you really like details when it comes to mixing and cooking. I agree that each songs has to have it's own personality. However, it must be difficult sometimes to achieve it in the mixing process, as you depend so much on the own variedness and richness of the original written material. Although perhaps having very similar songs from where to start gives you more creative freedom to differentiate them through the specific mixing process.

BazzBass, I'm sorry about your hearing loss. I'm glad you can get good recordings in spite of it. I guess that you must find visual reference in your gear very helpful.

Mark, very interesting to read about your working process. I'm following several of your recommendations in my own practice and it works great.

Regarding my mixing preferences, my ideal would be just to balance, and not apply any kind of processing to the mix (EQ, compression, effects, etc). I hate my original instruments sound being altered by any change. I'd like to keep my mix as bare as possible. However, I'd also like to get a higher sounding final master, but limiting or loudness-normalizing severely affects my dynamic range and my own inter-instruments balance. So I'm in a crossroads trying to clear my priorities and thinking which way to follow. I know there are intermediate options, like finding a balance between a completely bare and unmastered mix and a completely processed and mastered one, but I still have to clear up my mind and learn some more to make the best decision

 

[-mjk-]

Back in the day (when I was a broadcast engineer in the Boston area) WCRB, the only classical FM radio station in the region, was having difficulties balancing those very issues. During morning and afternoon drive, the engineers had to change the dynamics of the audio to punch it up more and limit the dynamic range because people were listening in their cars. They would never have been able to hear the quieter parts of the music if they hadn't done that. So, it was either having people complain about the station being "too soft" or having them complain about it being "unnatural" sounding. Late at night, the station was in full dynamic range of course and it sounded very good for a FM station and it's inherent limitations.

Interesting is that recently a studio client gave me a reference track to listen to. He said that he wanted it "bare" like that. On my studio monitors I could hear that everything was washed in a beautiful, brilliant plate reverb. I pointed that out to him and he said "oh yeah." It was obvious when pointed out but it was the nature of the song and how it was played and sung that made it sound "bare." If it literally was bare, no one would listen to it. The instruments and voice need to be in some kind of space. That space is created with reverb. Like it or not, if you don't utilize some kind of space, your records will always sound like an amateur recording.

I've done some mixes for clients who immediately complained about the reverb. But they we willing to put that mix into rotation on their daily playlist. Invariably they always are amazed when their tune comes on and it holds up to the commercial releases when their home recordings don't. It's no secret why that is.

Many people have "demoitis" and that's why they need a producer.

 

[Bambi]

Hello MJK, thank you for your examples. Very illustrative. I've just taken the chance to see the video you linked of Rick Beato talking about demoitis, and the story about the band and the preamplifier is very embarassing and comic. I think my demoitis doesn't reach that level, hahaha. However, in my case it's not exactly a self delight of listening to my own recordings but a much deeper feeling. It has to do with a leaning towards truth, naturality and simplicity. I understand it more as a philosophical issue rather than a merely aesthetic or professional one. I'm fed up with the artificiality and lack of naturality of many things on this world, not only in music, but in every other field of human activity. I don't want to be misunderstood, as if I thought that careful musical or artistic production weren't very valuable. I actually think they are essential in most of the artistic productions in any field. 99% of the music, movies, books, or art I like, is processed, edited and polished somehow to get the best results possible.

I've always found the specific case of audio engineering very beautiful, complex, and somehow atractively misterious, like being able to fly an airplane and manage the control panel. And musical production a very beautiful and integral ability, too. However, I think that naturality and bareness also deserve a place in this world. Think that 95% of the people who like music hasn't ever had the chance to listen to a real drumset live in a rehearsal place, and notice how loud, raw, or horrible it can sound depending of the quality, the adjusments, or the playing. I want to offer people a closer reference of what music sounds like without any processing, although it will never be as close as hearing a jazz or blues band live, without any external amplification or intermediates.

I know it's just and idealistic goal, and that the way I understand authenticity may not be shared by everybody, but that is my quest in life and in music, and I'm not thinking of giving up so easily. However, I recognize and agree with you in many things. Having an unprocessed, unpolished, and unproduced song will always sound much more amateur than not. Nevertheless, I'm just learning and trying to clear up my mind. It always takes me a lot of time until I mentalize and open myself to new possibilities. But I keep pushing and sometimes get to do it. So I appreciate very much your advice and, even if takes me some time to assimilate it, I always consider it

 

[Bambi]

Anyway, I don't want to go too much off topic. I'm sorry I wander so far sometimes. So let's get the conversation back to mixing. Only let me say that, although being a little bit dense from my part, all these conversations with Arjan, Mark and you about the process of audio production are really helping me to understand what options I have and what I get with each option. So I really appreciate the guidance of you all on these subjects

 

[-mjk-]

Mixing is an art that takes years to develop and one never stops learning or improving. If I could only go back and remix some of the stuff I did in the 80s!

Speaking of the 80s, back then we never heard anyone talk about "glue" or "gluing" a mix. We mixed on consoles and with analog tape, so naturally we didn't hear anything about adding "warmth" either. So, all these things that you hear about gluing a mix together or advice to "bus everything" stems more from the use of a DAW for mixing, than from the art of mixing itself (did you see what I did there?).

One thing that all of my mixes have in common is that I design a soundstage as early as possible in the process. One of the great things about those incredible 80s records is the use of reverb to create a space in which the song resides. This is something you almost never hear anyone talk about today. It's all about plugins, parallel compression, dynamic EQ and basically a bunch of crap if you ask me. Those are just tools in the toolkit, but to say that you need all of them to get a good mix is ridiculous. I use a hammer only when it's needed.

If you listen to the recent post of a mix I did for @robert monahan you will hear a mix that actually has a surprising amount of reverb thoughout. Recently I was working on a mix for another studio client and he sent me a reference track by a fantastic country artist (I mentioned that in more detail in a post above). He said he would like the sparce sound on that record. I listened to it, and the whole mix was awash with a lush plate reverb but, my client didn't hear it that way, initally. The reverb was the "glue" in that case, and this is what I am taking about when I say that I design a soundstage.

The first thing I do when setting up a mix is to listen to the drum tracks along with the bass and several of the "bed" instruments. I try to listen for the decay time of the drum hits and how it overlaps the instruments. Then, using a Vintage Room reverb, I design a room for all the instruments to live in. Back in Boston, I had a huge room to work with and I often set up multiple Neumann U-67 condenser microphones across the studio to capture the room ambiance. These days, very little is recorded that way and you have to build that room out of digital effects. Most of the track I get for mixing are totally dry and devoid of room sound. That is preferable to a bad room sound though.

I start by purposefully setting the reverb decay time longer than I know will work. I shorten the delay time until it "fits" with the sound of the instruments. Let me say that this part of the mix build is not for the vocals, which I do separately (I'll talk about that in another post later on).

Before I go any further, please listen to this incredible mix so I can make reference to it:

All the instruments are in a room, but the vocals are not in the same virtual space. The vocal soundstage design is very different from the instruments, and that makes the vocal stand out more without having to be louder.

In the example above, there is a lot going on. There is some kind of short-tailed reverb, possibly gated reverb that was very popular in that era. The guitars are also in the room with the drums and there is a wonderful cohesion to this mix that goes far beyond just bussing everything or slapping an SSL buss compressor across the mix with a 4:1 ratio and the threshold barely tickling the level.

Notice that the slide guitar solo has a slight slapback echo on it. That is another technique to make it stand out from the band in the background. This brings me to the next part of the mix build: working in 3 dimensions.

AC/DC and some other guitar bands would always have something like 20 to 30 ms delay on the guitar tracks to push them back behind the drums, just like the amp line at a concert. Adding a bit of a very short delay and carefully balancing the dry mix can add a front/back element to your mix. Try to place your instruments front to back as well as left to right.

OK, so once I get my digital room built for the band, I adjust the most critical aspect - predelay. Improperly applied, reverb predelay sounds horrible. It sounds like it's out of time with the song and "lagging" behind. That means it's set too long. The primary use for the predelay is to give the instruments or vocals an opportunity to hit the ears initially dry and then have the fill of the reverb come in under it. This prevents vocals, in particular, from being obscured by reverb. Proper use of reverb predelay also allows you to put that "wash" sound across the mix and yet have incredible clarity. Michael Jackson's Thriller album has some of the best examples of reverb predelay ever used. There is so much "air" in those mixes and yet none of the air is masking anything.

So, while listening to my virtual room and the instruments all put into that room, I adjust the reverb predelay until I am satisfied with the sound. You'll know that when you hear it. And, btw, I generally put all the instruments into the room at equal levels to begin with, except the bass. I will look at the send levels for the drums, guitars, keboards, etc. and make sure they start out at the same level. If I push up the drums a bit more into the reverb I will push up the guitars too. Over time, I make individual adjustments to those levels until I like how it sounds. In the video above, the drums are more affected by the room sound than the guitars are, but that's how real drums sound in a real room anyway. The drums project all around while the guitars project straight out of the cabinet towards the mics. Understanding how sound travels and is picked up by microphones is crucial to getting a realistic sounding mix (if that is your thing). Of course, these are not hard and fast rules, etched in stone. These are general principles that I use to start out each mix. I normally do end up with different send levels on different instruments. In the end, my ears are the boss, not the numbers.

In a subsequent post, I will detail how I mix vocals. That topic is just too big to put into this post.

 

[-mjk-]

 

[Bambi]

Very interesting. Since I started learning things on the Forum, the idea of getting the best sound from the beginning, previous to recording, has helped me a lot to improve the quality of my audios. Even for people with little recording experience like me, it's easy to confirm how having a good sound and performance from start gives you a better result than trying to fix later what sounds bad from the beginning. I like the cat, too

 

[-mjk-]

I used to work with "artists" who were somewhat cavalier about what they wanted to put down on tape. Sometimes I would be referred to as "too slow" when getting sounds. When it came to mixing though, it was usually apparent that the time spent up front made for a better product in the end.

But I'm afraid that the DAW revolution has made things worse. People seem to think that plugins are tools to fix problems. I heard one guy in my neighborhood say "just autotune everything" and he wasn't kidding. Apparently he doesn't even try to sing properly.

 

[BazzBass]

autotune is an effect to the younger people. They like the warbling sound,oblivious to the fact it's there because their fave singer can't sing.

When I was ringing studios to get quotes, they all gave a price, then said, of course if we need to autotune and time shift a lot it will cost more. When I told them we don't do either of those,we actually want a live recording with no correction of time or tuning they sound disappointed and happy at the same time. Happy that there's less work, disappointed at the loss of extra money lol

 

[Stutz]

Bazz. A sad reflection of the modern times we live in.

 

[-mjk-]

@BazzBass imagine how disappointed they would have been if you told them you wanted to cut basic tracks and take them to mix yourself.

 

[-mjk-]

A lot of studios outsource their Melodyne work. I don't even own it.

 

[Stutz]

Many people have "demoitis" and that's why they need a producer.

mjk, I watched that Rick Beato video, very amusing. I agree with you and I think I can be guilty of "demoitis" to a degree. A great video and topic

 

[BazzBass]

@BazzBass imagine how disappointed they would have been if you told them you wanted to cut basic tracks and take them to mix yourself.

I'd like to record the stems myself at home,but take them to get mixed and mastered.

 

[-mjk-]

My Biggest Issue In Mixing Today's Music

I have been thinking about this a lot lately, so I want to share those thoughts because it will apply to very many people on the forum.

This post is probably not about what you are thinking. Let me phrase the title another way:

What Is The Biggest Difference Between Mixing in the 20th Century and the 21st Century?

This topic would never have been conceived if I had not started taking outside mixing jobs.

So,
1. Is it that my console is no longer analog, but digital?
2. Is it that my analog tape machines have been replaced by a computer with DAW software?
3. Is it that I have access to all of the hardware I used to use in the 20th century via software simulations?
4. Is it that I have access to software simulated hardware I never even dreamed of before?
5. Is it that I can have very many instances of those hardware simulations, not being limited to physical hardware at all?
6. Is it because I now have virtually unlimited tracks?

Here are the answers:

1. No. This doesn't even phase me (pun intended). I have always mixed on a console and the fact that I'm working primarily with a digital console (I do still have an inline analog recording console) doesn't throw me off in the slightest. The digital console is very intuitive and I love being able to save and recall all the parameters with a button press. I wish I had that 40 years ago!

2. I do not miss the associated problems with analog tape machines. My DAW is a multitrack recorder and a master stereo mixdown recorder, and its synchronized with my multitrack. I wish I had that 40 years ago too!

3. No. My plugins act how I expect them. I have a ton of experience with the LA2A and the plugin sounds just like the original, for example. Most of them do what I expect them to do.

4. No. (see #5).

5. My analog studio would handle whatever I could hook up to it, and yes, on several occasions I did use every patch cable we owned. Again, multiple instances are nice and while its amazing to have access to so much stuff, we used to rent what we needed and hook it up. So its not an alien concept.

6. While really nice, that is not really a major difference on the order of what I'm going to reveal below. There have always been analog and digital studios with ganged multitrack recorders.

OK, here is it: The biggest difference between mixing in the 20th century and the 21st century is:

Everything I've mixed in the 21st century has been recorded by amateur (non) recording engineers.

In 1985, if someone came into my studio and wanted me to record their vocals tracks, or mix their record, they brought a 2" tape with them. That tape came from a studio somewhere. Even a novice recording engineer in those days could record clean tracks at good and consistent levels. The tracks on that tape were recorded by a professional.

Please understand that I am not complaining. I am identifying issues and I want to help people make better recordings.

When I get outside tracks, there is a workflow that I go though on every track before I even start to mix them. I rely heavily on Izotope's RX9 Advanced [At the time of this writing, RX9 was the current version] to clean things up. Here is the basic cleanup routine that I use:

1. File conversion: I work at 24b/48kHz. If the files are anything other than that, I convert them, with noise shaped dithering.

2. Leveling: Taking the genre and dynamics into consideration, I do instrument leveling for consistent sounding performances, and I also do leveling on the vocal tracks for thickening. This really thickens things up and I apply the leveling while keeping the natural dynamics of the playing and singing.

3. Noise removal: I make a noise profile for the track and digitally remove background noise, like computer fans, undesirable room noise, reverb, etc..

4. Guitar Cleanup: I reduce guitar pick noise, string squeaks and amplifier hum as needed.

5. Vocal Cleanup: I reduce mouth clicks, smacks and plosives from the vocal tracks.

6. Drum Cleanup: This depends upon whether or not the drums are real and recorded with microphones, or whether they are digital. I have to try to rebalance the digital drum tracks because almost invariably, I get a stereo pair of tracks with the drums already mixed. Typically the hats and cymbals are out of control, and also since there is no separate kick and snare, its very difficult to put FX across the entire drum mix that actually works on everything.

Now is the part where I share some tips and techniques to get better recordings out of your Model 12 or DP-32 whatever.

Microphone Nulls: People tend to think of microphone "pickup" whereas engineers think of microphone "rejection." Example: when you mic a snare drum, where to you point the back side of the microphone? Assuming a cardioid pattern, you should mic the snare in such a way that you point the null at the most offending instrument - that is the instrument that bleeds into the snare mic the most. Typically that is the hat, which is right next to it of course. Reducing that bleed even a bit can make a huge difference when using a gate to isolate the snare hits. Mic up everything you do with a view to keeping other things out and not just picking up the main instrument. I've seen guys point their guitar amp right at the drummer so that every single drum mic has guitar all over it. Point the amp away and then put the amp mic with the null toward the drummer. Problem solved. Experiment with different mic placements before you do your takes. You can erase any of those test takes before you start. I often made mic placement diagrams for future reference. That's really handy for duplication successful recordings.

Drum Tracks: Separate tracks is better. In 1985, my standard drum track lineup was like this.

Track 1: Kick
Track 2: Snare
Track 3: Toms Left
Track 4: Toms Right
Track 5; Hat
Track 6; Overhead Left
Track 7: Overhead Right

Sometimes I would add a snare bottom mic with the polarity reversed. Often I would have 2 room mics since the room was huge. No matter how many toms where were, I submixed them all to 2 tracks, panned as I wanted them.

Without MIDI or some way to synchronize your drum machine, I realize that it may be difficult to impossible to make multiple passes with your drum machine to get separate tracks. Some drum machines do have multiple outputs however. If I only had 4 outputs to work with, I would put the Kick on 1, Snare on 2, and everything else on 3 & 4. I would really watch the hat and cymbal levels too. I recommend that you reduce whatever you think is a good level, by about 3 dB until you get good at drum mixing. Almost all stereo drum tracks I get have far too much hat and cymbals on them and it results in some kind of a drum sound compromise in the end.

Resolution: Please record your tracks in at least 24bit/48kHz resolution.

FX: If you are sending your tracks out to be mixed by someone else, those tracks should be dry. Work with the mixing engineer to apply the FX you want. Studio grade FX are better than the internal FX on your standalone recorder.

Dynamics: Learn how to apply compression and limiting during the recording process to make the recordings more consistent and uniform sounding. Pay attention to things like pick noise when recording guitars. Use a pop filter in front of vocal mics.

I could go on for days, but these are the main points. I'm sure that I'll be editing this over the next few days as I realize what I forgot to cover.