Most common rookie mistake when mixing?

[blue2blue]

 

The main musical problems I hear are: Overly-cluttered arrangements Solos that should be cut in half Lack of dynamics and modulation (going up a half step can be very cool!) Lyrical cliches Trying to sound like another band instead of creating a unique sound

 

 

I remember the day I realized some of my very favorite solos were a quick, clean 8 bars... Growing up in the wild, acid-drenched soloing of the late 60s, I'd come some sort of ill-observed and only half-considered notion that only really long solos could fully explore something or other...

 

But when I really listened to a lot of music I realized that a lot of the really long solos were often by people who didn't have a lot to say or a grasp on how to say what they had to say in a coherent manner. (Of course, sometimes that sheer lack of coherence had a kind of half-crazy cool... but that was then and this is... for better or worse... now.)

 

 

Interesting point about modulations... they really can make a song/arrangement come to life. I think part of the problem is that a lot of us old-timers got fed up to the gills with modulations in the 70s... And then there's the band dynamic -- you can just FEEL everyone's eyes rolling up into their heads when you say, "You know what this song needs? A modulation!"

 

And I think a lot of that comes from what I'll call "capo modulations" -- just jacking the whole tune up by a given interval.

 

But a lot of the great songs of the past have more subtle modulations where a bridge or chorus or other subsection will modulate into a different mode or even key in a way that opens the song up harmonically, rather than just sticking a bumper jack under it and lifting...

[blue2blue]

Craig, you mean you'd find fault with my latest lyric??? Here it is: I LOVE YOU I woke up this mornin' Rememberin' all we'd said last night You said you'd found another and we must part It cut like a knife when you broke my heart Girl, you've touched my very soul... Now I've got to swallow my foolish pride... Just give me one more night Just a chance to make things right And I know the lovelight will come shinin' through, And two hearts will beat as one. 'Cause, girl, you know tomorrow We'll both feel pain and sorrow... (Chorus) 'Cause I LOVE YOU, I LOVE YOU, I LOVE YOU... Yeah, I LOVE YOU, I LOVE YOU, I LOVE YOU, Hey girl, I LOVE YOU, I LOVE YOU, I LOVE YOU Until the end of time. :D

 

It needs some la-las or sha-la-la-la-las...

 

__________________________

On another note altogether...

 

 

In discussions about internal gain staging in modern floating point DAWs I think it's important to remember that, just because the DAW can handle internal plus 0 dBfs values which, in the case of tracks can be gracefully compensated for by brining down the output bus, that doesn't mean all the PLUGS in your signal chain can! Some may handle overs gracefully; some may handle them but not be optimized for them -- and some may distort like a barbie mic with a dying battery.

 

So, for that reason, if for no other, I tend to try to keep tracks in the "normal" sub-0dBfs range.

[audioicon]

 

Yeah, I think I'm guilty of that. I think I'm also guilty of using too much reverb. How do you know when there is enough? What instruments do you avoid using reverb on (for a typical rock arrangement)?

 

 

When I first started mixing, I wanted to hear everything in the song, the reverb, the compression, the eq. Today I barely use these things.

 

My musical style is different from yours but I can say it's the same principles.

I do R&B/Hip/Hop/Raggae/Ragga and in most of my songs I add only up to 4 percent reverb, I may or may not add compression at all. but if I do I can barely hear the difference myself. If my bass is more into the 40 to 50kz, then I try to remove or trim those frequencies out of my kick so they dont collide. If my bass is more into the 80 - 90hz area then I'll gave the kick more bottom in the 40 - 60hz area. I prefer the kick to sit on the lap of the bass or get under the sheets with the bass but not the bass on top of the kick.

 

Again I'll solicit any corrections from the pros!

 

As for reverb it's about creating space, It's all about the musical style. Poor arrangement will make things sound like a clogged toilet. It's okay to let only the drums and the bass play during the verse and add a lead guitar during the bridge and the slam the huge rock sound on the chorus. More reverb if the song is spacious and less if it has a thight groove but it's up to the song.

 

Let see what the pros has to say.

 

I use the Rode NTK, it's already a little high and so I don't use that much EQ.

I' m still a rookie but I try to take the knowledge of the folks in this forum and all my other research convert them into practicality and create my own format.

 

Every input in this post is very valuble information.

 

Audioicon.

[UstadKhanAli]

Yeah, I think I'm guilty of that. I think I'm also guilty of using too much reverb. How do you know when there is enough? What instruments do you avoid using reverb on (for a typical rock arrangement)?

 

It's hard to say without hearing the song. The song often "tells" you how much reverb to add. But if your reverb is clouding up your vocals or the mix, it's likely too much!

 

It depends on each song - and I often record very ambient, atmospheric sort of music, which is rather different - but for a lot of rock and pop songs, I add a little bit to the vocals to give 'em a sense of space, and might add a tiny bit to the snare track. I usually add enough to give it a sense of space, but not enough that you literally hear the reverb so you can say, "Oh, there's reverb added to that track!" Does that make any sense?

 

If I want expansive sounding vocals, I also often add an analog delay to the vocals so that you can hear one faint repeat instead of adding more reverb. I find that if I add this delay, I can also back off on the reverb on the vocals more so that they won't sound so washy. When I add reverb to the vocals, I often play with not only the decay of the reverb, but also the pre-delay, adjusting it until it sounds good and does not interfere with the vocals themselves.

[shniggens]

Should I avoid reverb on guitars and piano/keys?

[UstadKhanAli]

Well, I don't know, it really depends on the song. I put reverb and delay on my keyboards/guitars quite often, but that's because we're playing ambient-atmospheric music and - well, it probably wouldn't be very atmospheric if it didn't have lots of that stuff on it!!

 

But when I am doing non-ambient/atmospheric recordings, I usually mic the room so that you are getting the natural room sound (I often have one close mic and one room mic when micing guitar cabinets) and then don't add any reverb. I almost never add reverb on acoustic guitars, preferring to back the mics away if I want a more distant sound. And with keyboards, again, it just depends. I frequently send my keyboards through an amp and mic them, so I would mic this in a similar manner to the guitar cabinets and not add any reverb on it because it already has the room sound (a sense of space, in other words).

[MrKnobs]

There is absolutely no truth behind the idea that recording at lower levels somehow prduces a grainier waveform. NOT TRUE. Digital sampling theory says otherwise. After reconstruction/filtering the waveform loses NO fidelity at all because it was converted at a low versus high level.

 

Really? So if you record something just using 4 bits and then amplify it to mix it in with other sound with 24 bit quantization it sounds just fine?

 

Hey I learned something today. Maybe I'll go back to 8 bit recording now.

 

 

Terry D.

[boosh]

Really? So if you record something just using 4 bits and then amplify it to mix it in with other sound with 24 bit quantization it sounds just fine? Hey I learned something today. Maybe I'll go back to 8 bit recording now. Terry D.

 

He's talking about recording levels Knobster,... not about bits.

[audioicon]

 

Should I avoid reverb on guitars and piano/keys?

 

 

The only way you can get the answers to question is to experiment.

I'm not trying to say Ken is wrong or his respond lacks validity.

 

But when I hear ken discribed his process of adding reverb/micing I'm like wow! This guy has class/his own misical identity.

 

This is where people mess things up a lot trying to asked, how do you do this how do you do that. Asking questions that involves musical principles is better then how to.

 

For example I will not come to this forum and asked how music reverb do I add to my back up vocals? Once again the answer to this question is subjective and should be based on the kind of song.

 

Shniggens, I respect all your questions and I'm also learning from the answers you are getting but you need to experiment, record a guitar/piano, applied reverb/delay/modulation, in music there are no creative boundaries.

Trust me, if you experiment, soon you will have your own unique style.

 

I have not heard Kens work but I can honestly say that he has a style that is very unique!

 

I started recording my own music because I could not find a better way to inteprete my feelings to the producers, I sit at the table when it rains, I'm looking through the window at the rains drops. I have this emotional sound/piano/strings/guitar in my head, I go to the producer and try to explain but it's like I'm loosing my mind. So I learn to record so I can lay down the idea myself and then have a producer work his way from there.

 

To wrap up my speach, experiment, don't be afraid, you can always do it again or hit the undo. But only learning the principles and exprimenting with the knowledge, will give what you are looking for.

 

Audioicon.

[shniggens]

Yeah, I understand that experimentation is paramount, but I'm asking from more of a "rule of thumb" angle.

[Super 8]

I LOVE YOU I woke up this mornin' Rememberin' all we'd said last night You said you'd found another and we must part It cut like a knife when you broke my heart Girl, you've touched my very soul... Now I've got to swallow my foolish pride... Just give me one more night Just a chance to make things right And I know the lovelight will come shinin' through, And two hearts will beat as one. 'Cause, girl, you know tomorrow We'll both feel pain and sorrow... (Chorus) 'Cause I LOVE YOU, I LOVE YOU, I LOVE YOU... Yeah, I LOVE YOU, I LOVE YOU, I LOVE YOU, Hey girl, I LOVE YOU, I LOVE YOU, I LOVE YOU Until the end of time. :D

 

My god that was brilliant! Do you have music to go with that?

[MrKnobs]

He's talking about recording levels Knobster,... not about bits.

 

You have a 32 bit floating A/D converter?

 

Seriously, low input level into an A/D converter causes quantization error. The only "cure" for that is dithering, and dithering is noise. So it DOES make sense to keep a healthy level into the converter inputs, within reason of course. That's the primary reason to use 24 bit audio, to avoid quantization error without having to slam the levels.

 

Terry D.

[EccentRick]

(Chorus) 'Cause I LOVE YOU, I LOVE YOU

 

HEY! :mad: You're ripping off the chorus to "Silly Love Songs".

 

 

Rick

[Kendrix]

Really? So if you record something just using 4 bits and then amplify it to mix it in with other sound with 24 bit quantization it sounds just fine? Hey I learned something today. Maybe I'll go back to 8 bit recording now. Terry D.

 

Well...To clarify:

 

I dont think recording 8 bits worth of signal in a 24 bit system is the same as recording into an 8 bit system. I never suggested that.

 

My statement was assuming/talking about reasonable adjustments of line levels- not a totally anemic signal. I wouldnt suggest recording at such anemic levels. You'll have other analog signal to noise issues doing that.

 

My point is this: Based on digital sampling theory, as i understand it, if you put a pure sine wave into a 24 bit system, after D to A reconstruction and filtering the result is no more or less of a sine wave whether you go do this at 0dbfs or -12db or -18db. Only the amplitude/dynamic range is different.

This result also applies to more complex signals. There is no "stairsteppin effect" introduced. Thats not how the D to A conversion works.

Thats the common myth I was calling out.

 

Practically speaking, there is enough dynamic range in a 24 bit system so that recording at -12 or -18 peaks does not cause any issues. It has the benefit of ensuring plugs or the summing bus maintain adeuate headroom.

 

Cheers

[rasputin1963]

HEY! :mad: You're ripping off the chorus to "Silly Love Songs". Rick

 

 

Oh, gee... and if anything, it was Edwin Hawkins' "Oh Happy Day" I was influenced by...

 

[MrKnobs]

Well...To clarify: Practically speaking, there is enough dynamic range in a 24 bit system so that recording at -12 or -18 peaks does not cause any issues. It has the benefit of ensuring plugs or the summing bus maintain adeuate headroom. Cheers

 

On this we agree.

 

Getting back on topic, mixing mistakes:

 

(1) Thinking that you're going record crap and extrude cupcakes from it simply by mixing well, aka "We'll fix it in the shrinkwrap."

 

(2) Setting the tone of everything with the SOLO button engaged, and then wondering why the sound changes horribly when you hear it all together.

 

(3) Trying to make everything fit in a dense mix using EQ when you know in your gut that some of the stuff just shouldn't be there. As others have said, don't hesitate to automate that MUTE button as needed.

 

(4) You can't mix well if you can't hear well. Solving that problem is job one, before the mix begins.

 

(5) Mixing too loud. Mix at a fairly low level, and change levels from time to time to asses certain components of the mix. I like to lower the volume to a whisper to see how the vocal really sits in the mix. If it sounds right at very low level, it'll be fine cranked. Likewise, turn it down then up for just a moment to see how the bass level's working, just not too loud or for too long until you're done.

 

(6) Mixing for hours and hours. Varies by individual, but you really only have a certain number of hours in you per day where you can mix at your best. Hearing fatigue, or just burnout will cause mistakes. Save the snapshot and automation and call it a day when you're starting to make mistakes or what was sounding right a couple hours ago doesn't sound the same anymore.

 

(7) Not backing up your work in a timely manner. Details depend on what sort of rig you're using, but do you really want to lose several days of work? Or worse, lose a hard drive or erase something important and (maybe) have to ask the artist to do it over?

 

(8) Too much FX. FX has a way of creeping up on you, especially when you first start out in this biz. Set a modest level of 'verb and/or echo, then reduce it. Better to err on the low side than put too much on. It's embarrassing to listen to a mix a few days later you drenched in reverb or slapback echo, especially if you aren't alone while listening to it.

 

(9) Confusion. Sure, everything's on the computer or in your automated console or whatever, but MAKE NOTES. It might be weeks before you come back to something, and you'll probably work on other things in between and even sleep a few nights before getting back to it. You'll forget something and have to figure it out all over again, wasting a lot of time compared to just reviewing your notes from the last session. "What? That track you just used wasn't empty! It had my best solo of the session on it!"

 

(10) Creative substances. It *may* be true that there are various recreational substances that actually enhance performance of a player, but I've not heard anyone claim that mixing works the same way. Hold off on the drinking or toking until after the mixing's done for the day. Then crank up the speaks and enjoy with the band.

 

 

Terry D.

[EccentRick]

rasputin1963,

 

Edwin Hawkins, huh? Hmmm. Don't forget the "Singers"!

 

 

Rick

[where02190]

I am using 24 bits, and now that i aim -18 dBFS RMS in the loudest parts, i dont have audible noise in the quieter parts, and i dont have clipping problems neither.

 

-18dbfs peak is pretty low. I recommend targeting your nominal input level to the 0dbu reference of your converter. You'll still peak well below 0dbfs, leaving plenty of headroom for transient peaks, and capture solid audio quality.

 

But whatever works, better a bit low than too hot.

[UstadKhanAli]

Yeah, I understand that experimentation is paramount, but I'm asking from more of a "rule of thumb" angle.

 

It's difficult to get a "rule of thumb" thing going because it truly depends on the genre and song and taste and style and how it was tracked and all that.

 

But for a lot of pop and rock songs, you could try goosing the reverb forward a little, and when you really start to hear the "Oh, wow, it's got reverb on it now!" kind of thing (hopefully you know what I mean since that's not a very scientific description ), then back off on it a bit. Then mute the reverb and see if it's lending your vocal some body.

 

But really, listen to the song. It tells you what to do. Let the ego go and listen to the song.

 

And remember that thing I mentioned earlier about adding a little bit of delay if you want really expansive vocal sounds. What I sometimes do is when the vocalist is really belting (often, this occurs in the chorus), I might ride the delay slightly so it really comes across and imparts this feeling of a large space, if it supports the song emotionally.

 

And too, again, the song tells you what to do.

 

If the song is a really intimate sort of song, does swamping your breathy, intimate vocals with reverb add intimacy? Often, the answer is emphatically NO!!!!

 

Consider riding the reverb, especially on vocals. Don't let it be static. Really listen to the emotion and feel of the song. That always tells you what to do. Some songs require some reverb, others, just a touch. Others, maybe they should be quite dry. Your job, when tracking, mixing, etc., is to enhance the emotion and artistic statement of the song as much as possible, and part of this involves choosing the perceived sense of space that vocals (and other instruments) go into.

 

There are times in which a few plaintive, sparse piano notes swamped in reverb impart a really ghostly or eerie or lonely quality. And then there's times in which you should really leave your piano alone, or give it a sense of space with all the other instruments to glue everything together and give a sense of people playing in a room.

 

I hope this helps somewhat. And remember, this is only my opinion. Others may have different approaches.

 

~~~~~~~~~~~~~~~~~

 

Audioicon, thanks for the compliments. My web site is full of MP3s if you ever want to hear them:

 

www.blueberrybuddha.com (then go to "Virtual Tour" to listen to MP3s)

http://www.elevenshadows.com/mp3.htm

http://www.nectarphonic.com and then look for our songs

[shniggens]

Wow, that's great!

 

What do I owe ya?

[mobobog]

-18dbfs peak is pretty low. I recommend targeting your nominal input level to the 0dbu reference of your converter. You'll still peak well below 0dbfs, leaving plenty of headroom for transient peaks, and capture solid audio quality. But whatever works, better a bit low than too hot.

 

Well yes, i am not targeting -18 peak but RMS. and my peaks are -12 -10 some thing like that... but you know, when people get in the song i then reach -6 or something like that in really loud passages.

 

Where a question... what happens when you have an interface with unbalance inputs? whats the level to target to?

 

[mobobog]

 

Mobobog: Yes, but to clarify, this doesn't have anything to do with the converters sounding better, it has to do with the mix buss, y'see what I mean? The converter is going to sound like the converter at any reasonable level - it's not going to sound better or worse. What you and Where are discussing have to do with the mix buss.

 

 

Yes thanks, i wasnt sure about that... maybe as Craig said i was distorting the buss even without seeing it in the meters, and that is why i hear an improved sound now.

[where02190]

Well yes, i am not targeting -18 peak but RMS. and my peaks are -12 -10 some thing like that... but you know, when people get in the song i then reach -6 or something like that in really loud passages. Where a question... what happens when you have an interface with unbalance inputs? whats the level to target to?

 

That makes more sense, your original post said you peaked at -18dbfs, but if that is your target nominal, then you're fine. Is that the 0dbu reference for your converters?

 

Balanced or unbalanced, the target nominal is the same, the 0dbu reference of the converter.

 

Automation is yoru friend when mixing, and I prefer it to dyamics, especially when mixing ITB.

 

It is especially helpful for fx such as reverb and long delays. Instead of automating the return, automate the send, using it to build sustain at the end of phrases, or that one big 1/4 note delay repeated fx that you should never use more than once a CD.

[UstadKhanAli]

Wow, that's great! What do I owe ya?

 

 

 

Don't forget about adjusting the pre-delay on your reverb (not just the decay length).

[mobobog]

That makes more sense, your original post said you peaked at -18dbfs, but if that is your target nominal, then you're fine. Is that the 0dbu reference for your converters? Balanced or unbalanced, the target nominal is the same, the 0dbu reference of the converter. Automation is yoru friend when mixing, and I prefer it to dyamics, especially when mixing ITB. It is especially helpful for fx such as reverb and long delays. Instead of automating the return, automate the send, using it to build sustain at the end of phrases, or that one big 1/4 note delay repeated fx that you should never use more than once a CD.

 

I am not sure what the 0 dBu reference point is for my converters. I wrote to emu and they answered with a lot of specs for the converter but not what i asked them they really helpful guys though.

 

I am using that level because i read somewhere that if in doubt that was a good level to point.

 

Thanks