Mixing

[Donna]

Hey all, one main difference I hear between 4 track mixing and bigger units is a sense of remoteness. It's like the pro mixes have space in them. If you don't know what I mean, it's the difference between cheaper film, when the movie or whatever seems "present" and film-type film where there's a remoteness.

Now I am not using FX at all right now, but I think it's more than that. I've a buddy who can get quite a big sound spatially via a 4 track. Is that because he mixed the bounce onto stereo tracks?

Experimented recently mixing in mono and little eq as possible. The mix sounded great last night. Today it sounded quite bad. The best mix of the one I'm working on is where I just got it how it sounded best - more eq, louder levels overall and the mid range boosted (at the expense of drum hum).

[Prometheus]

Hey all, one main difference I hear between 4 track mixing and bigger units is a sense of remoteness. It's like the pro mixes have space in them. If you don't know what I mean, it's the difference between cheaper film, when the movie or whatever seems "present" and film-type film where there's a remoteness.

Now I am not using FX at all right now, but I think it's more than that. I've a buddy who can get quite a big sound spatially via a 4 track. Is that because he mixed the bounce onto stereo tracks?

Experimented recently mixing in mono and little eq as possible. The mix sounded great last night. Today it sounded quite bad. The best mix of the one I'm working on is where I just got it how it sounded best - more eq, louder levels overall and the mid range boosted (at the expense of drum hum).

The main ways to get good seperation are by EQ and by panning...

EQ first, to start off with I would seriously advise downloading a page on how the human ear perceives the key frequencies of commonly used instruments. These are a general rule of thumb and are a bit different from one guitar or Cello or Kick Drum to the next, but after a while you get used to this and with a lot of practice you can do this yourself when you're ear becomes tuned to picking out the keys of any sound. There could quite possibly be a page about EQ in the Songstuff tech pages...

When I talk about EQ in different frequency bands, I think of it as Sub Bass 20 - 60Hz, Bass 60 - 200Hz, Low Mid 200 - 1 KHz, Hi Mid 1KHz to 4KHz, Sibilants 4Khz - 10Khz, Sizzle at around 10 Khz and Gloss at around 20Khz, and above that you get into realms that only a dog or cat can hear...

One of the biggest mistakes inexperienced engineers make is soloing up all the tracks and trying to make them all sound nice individually... With a lot of tracks, this is not a good approach for the simple reason that no one is going to hear them individually, ever, unless you give them your multitracks to listen to... For an example of what I'm talking about, say you make a multitrack recording starting with a recording of a grand piano, and you solo it up and make it rich and warm, then you record a bass guitar and make it rich and warm, then you record and acoustic guitar and make it rich and warm, what you have is a pile of lower mid range harmonics, bass and sub bass...

Thinking about what you want people to hear, for one thing, the human ear hates midrange... It sounds harsh at the Hi Mid and boomy and boxy at the Low Mid... It stands to reason that you're gonna want the bass end of the mix to be carried by the bass guitar, so as a general rule of thumb, you can roll all the sub bass off the acoustic guitar and the piano. You'd probably want the piano to still carry a bit of power so using bandpass EQ to filter some of the lower mid out to get rid of boominess (bearing in mind all of this should be done very judiciously and carefully, small movements), and often shelfing up the high end a little can improve the sound. Acoustic Guitars often sound better with the mid cut quite savagely (but judiciously) and maybe a little High Shelf from about 2KHz up and maybe a little high end sizzle or Gloss added, again I would stress that all of this is purely rules of thumb. Bad use of EQ is often worse than not using it at all...

The effect of doing the above things properly should be that the Bass, the guitar and the piano should all sound more seperate...

For an example of bad EQ, add a massive boost at 200HZ to 400Hz on a guitar and Shelf off all the High End, and you'll see how Bad EQ can destroy a sound. I found that when I was learning to mix, I tended to drastically over use effects like EQ and reverbs...

Panning is essentially controlling how wide the stereo field is and is done by controlling how much of the signal is sent to each speaker (or speakers in a surround mix), usually with a pot that can be turned from 8 o'clock (100% of the signal to the left speaker) to 4 o'clock (100% of the signal to the right speaker) with the middle of the stereo field being at 12 o'clock (Signal split 50-50). In order to make space in a mix if the middle of the stereo field is becoming cluttered sounding, you can start to shift more of the signal out to the right or left, which makes it sound like your stereo field is wider with more space in the middle. Again, this effect should be used judiciously, or you can end up with a hugely wide spatial sound with a "hole in the middle" of it. This hole in the middle can become apparent when you're double tracking to widen a guitar sound, or when you're recording drums with overhead mikes... Time domain effects like Chorus can also be used to make a mix sound wider and deeper, and if you're feeling really adventurous you can pan things across the stereo field, sudddenly to make them jump from one side of the mix to the other (a great way to grab people's attention) or slowly to create and artificial doppler effect, like you hear when an ambulance passes you. This would be best done in conjunction with a pitch shifter, to make the sound get higher pitched as it reaches the middle of the mix and then lower again as it reaches the other side...

Things like reverb can be used to add depth to a mix, another very useful spatial tool... The Predelay setting in reverb tools is essentially to set how far away you will perceive the instrument or vocals or whatever to be away from you. Each millisecond translates into a foot, since sound travels at 1000 feet per second... If you ever get a chance to stand in an anechoic chamber it's an interesting feeling. Because there is virtually no reflections from the walls, no reverb, your eyes tell you that you're in a room ten feet wide by ten feet long by eight feet high (or whatever) but your ears tell you the walls are two inches from your face. The feeling is quite disorientating. to translate that concept into mixing, if you've recorded singing, for example, in a vocal booth that has little in the way of reflections, it stands to reason that people will feel that singer is very close to them when they play back the sound (unless they listen to it in a very echoic room) so it often sounds far better to put a reasonable pre delay to make it sound like the singer is further away. As a rule of thumb again, if you make your singer fifty feet away (fifty seconds pre delay) this tends to make them fit in the mix pretty well. The reverb length basically simulates a very reflective room, and the reverb size a very big room... The current fashion in mixing is to generally speaking have big reverbs that are gated pretty short length wise. Obviously, you want to think about the effect you are trying to achieve... The other day I was paid to create an artifical orchestra out of 200 Violin tracks, so it stands to reason that a lot of pre delay will be needed (to make the orchestra a realistic listening distance away, and a big reverb will be needed to simulate the feeling in an orchestra house, and then maybe dampen it a little or gate it to simulate the bodies of the audience soaking up the reflections... Again I can't emphasise enough that reverb should be used very carefully...

Anyway, this is turning into quite a weighty tome now, so I'll stop here and hope you find some of it useful...

[Nightwolf]

Great info Prom, lot's of good pointers, thanks for taking the time to write that up.

if you make your singer fifty feet away (fifty seconds pre delay)

I for one see the advantage of having the singer be 50,000 feet away, but I don't think that's what you meant

Couldn't resist, sorry

Here is an article John Moxey wrote about EQ:

Click ME

[Prometheus]

Great info Prom, lot's of good pointers, thanks for taking the time to write that up.

if you make your singer fifty feet away (fifty seconds pre delay)

I for one see the advantage of having the singer be 50,000 feet away, but I don't think that's what you meant

Couldn't resist, sorry

Here is an article John Moxey wrote about EQ:

Click ME

That's a good article by John...

[john]

Hey

Tahnks Prometheus.

Donna, you might also find this useful:

EQ Frequencies

Cheers

John

[Donna]

Prometheus, thank you so much! It is extremely helpful, what you wrote, and I shall save it.

And try some, tomorrow; I'm guessing what I bootsted was really lower than mid, but whatever. I did vocals and mixed tonight. I love the around 10 o'clock/2 o'clock panning but find there's nowhere else to go if I start out that way (vox). It's quite difficult to hear the subtleties of eq (and sometimess pan) on my cheap phones (and ears). Quite understandable when you speak of "hole in the middle", warming all the instuments/what that badly does, and so forth.

Where to now? OK John, I've seen the first article more than once (the 2nd is maybe a variation of what Prometheus suggested to download & looks very helpful)...but I'm afraid I have to back way up here becausse I have never figured out anything technical w/ the board, just go by intuition

I hope you guys will tell me what's what!

First: for each track on my board there is a section "eq" w/vertically spaced pair of knobs - which sit on top of one another - two tiers. There is a label also vertically spaced which reads "gain o frequency." Is the frequency then the top tier of the knob do you think? The fact that there are 2 eq knobs for each track, does that mean it's 2 band?

Second: The numbers around these knobs...this is confusing, cause there is a different set on both top and the 2nd knob beneath it. Also I don't know which numbers correspond to the upper or lower tier of each knob.

a) I can see how to get something between 1-8 k (if the top knob's labeled w/ these numbers, 2.8 is 12 o'clock), but what is this scale on bottom knob - 62 at 8 o'clock, 300 at noon, 1.5 () at 4 o'clock. How can one go UP from 300 to 1.5! Now are these numbers htz or db or ?

the meters (negative 20 black to + 3 red) are THEM db's, Precious?

I feel like a scandal - o well.

And so...how does one boost...just turn the "k" to a higher "k"? But how does one boost or cut them bottom tiers of these doubled-stacked knobs, when it goes from 300 "up" to 1.5?

I did ask someone else this, and since he didn't explain direction of knob turning for boosting/cutting, I figured turn it clockwise to increase and vice versa. This is the 246 tascam portastudio I have; but I think I'm beyond manuals, even if I had one.

gracious thanks.

Edited November 23, 2005 by Donna

[Donna]

PS: Mixing tonight (before I read any replies here) I did drums panned to between 10-11 o'clock and guitar between 1-2. The forth mix experimentation, I panned the lead vocal to about 1 o'clock, never tried that before, cause I's looking for space! Who really knows tonight how it sounds via the phones?

Nothing's a problem w/ unison vocals panned above like drums/guitar, plenty space there and I'm content.

I'll listen tomorrow on speakers...wonder if I'll get my effects running for this MP3-to-be, and if I'll be content w/ the sound copied from analog to digital?

[FinnArild]

Don't know if this can help, but:

Long time since I did 4 track, but what we used to do was to take the 4 tracks and bounce them down to 2 via another tape. Then we had 2 new tracks to record on. We did that 2 times, and the result was a de-facto 8 track (and a little more tapehiss).

Takes some planning though ...

[Prometheus]

PS: Mixing tonight (before I read any replies here) I did drums panned to between 10-11 o'clock and guitar between 1-2. The forth mix experimentation, I panned the lead vocal to about 1 o'clock, never tried that before, cause I's looking for space! Who really knows tonight how it sounds via the phones?

Nothing's a problem w/ unison vocals panned above like drums/guitar, plenty space there and I'm content.

I'll listen tomorrow on speakers...wonder if I'll get my effects running for this MP3-to-be, and if I'll be content w/ the sound copied from analog to digital?

I don't really know the specific device that you're using but -20 to +3 does sound like decibels... Most faders are calibrated to have the origin (zero) at Unity gain (zero decibels), about 3 quarters of the way up the strip, which gives headroom if the signal needs pushed a little for that track...

With panning, I find the best thing to do is think about it from the point of view of someone in an audience looking at a band on a stage, or as a Sound Engineer would say, from an FOH (Front Of House) persepective. That way, you tend to pan all your mixes from the same perspective, so on your album, you aren't going to have a song with a drum kit panned from a drummers point of view, then on the next song it switches around to someone in the audience. This keeps things consistant...

Phil Spector came up with a set of panning laws that are basically still in use today, which roughly speaking are that you start off in the center with your kick drum and bass guitar and lead vocals, and then after that if things get cluttered you can start panning things out a bit, but when you pan something to the left (or maybe right first if you're right handed, which I'm not), then you balance that by panning something else with a similar SPL (Sound Pressure Level) The opposite way. If you wish to add more, you can pan out a little further each way, then again a little further still, and so on, and out away from the center of the stereo field if you are doing a very lush mix with a lot in it you can still seperate with EQ... Using this method you can come up with very huge spatial mixes that are still evenly balanced on both sides, making them more pleasing to the ear... This is exactly the technique I used to create the artificial orchestra I was talking about before... There is always room to be creative, and I've found it can be interesting to have some kind of mad panning, maybe sweeping a guitar slide fast across the stereo field to make something stick right out of the mix, but I wouldn't do this kind of stuff all the way through a mix...

The only hard and fast rule I tend to stick to is that I always have kick drums and bass guitars centered, since they carry most of the energy of the mix and therefore will have most potential tp make it sound unbalanced, and usually I keep Lead Vox centered since that is what attracts most people's attention in a song...

I think honestly you could unlock a lot of effects and panning potential by bouncing your 4 Track recordings onto a sequencer on the computer. I know it sounds daunting, and I remember pulling my hair out when I first did it, but it would free up a lot of tracks for you to play around with and you can buy a Creative Audigy Sound Card with 24 bit Dynamic Range for fifteen quid (just over twenty dollars) http://www.ebuyer.com/customer/products/in...oduct_uid=98938 and download free versons of the software you need... You can even get an 8-Track version of Pro Tools, the industry standard sequencer for nothing from the NET... http://www.digidesign.com/ptfree/ and you'll be able to find plugin effects for it on the NET... Pro Tools is a very powerful package, it's one of the sequencers I served my time on...

[Prometheus]

First: for each track on my board there is a section "eq" w/vertically spaced pair of knobs - which sit on top of one another - two tiers. There is a label also vertically spaced which reads "gain o frequency." Is the frequency then the top tier of the knob do you think? The fact that there are 2 eq knobs for each track, does that mean it's 2 band?

I don't know if it works along the same lines as the Tascam 244 4-track recorder that I have at home, I suspect yours will maybe be a more advanced model, but it is a Tascam if I remember right? The 244 has a knob for low EQ and one for High, but they are dual pots with an inside component and an outside one, one being for gain (boost or cut) depending on whether you go right or left of center, and one being to sweep the center frequency of the boost or cut... As mentioned in Johns article, sweepable EQ is great for indentifying problem frequencies by turning the gain up a few db and then sweeping across until something unpleasant jumpls out at you, then you can do a cut at that frequency... So, in answer to the Question, one tier on each knob will be for the EQ gain and the other to sweep the center frequency...

Second: The numbers around these knobs...this is confusing, cause there is a different set on both top and the 2nd knob beneath it. Also I don't know which numbers correspond to the upper or lower tier of each knob.

I would really need to take a look at the thing, but I'm guessing one will have small numbers on it say 0 at 12 o'clock round to 10 or 12 or 15, and to -10, 12 or 15, which will be the gain. the other will have numbers in Hz or KHz which will be the center frequecy of the sweep?

a) I can see how to get something between 1-8 k (if the top knob's labeled w/ these numbers, 2.8 is 12 o'clock), but what is this scale on bottom knob - 62 at 8 o'clock, 300 at noon, 1.5 () at 4 o'clock. How can one go UP from 300 to 1.5! Now are these numbers htz or db or ?

-62 and 300 are in hertz, where as 1.5 is in KHz (thousands of Hertz) so 1.5 is actually 1500...

the meters (negative 20 black to + 3 red) are THEM db's, Precious?

They sound very much like it... I would be willing to put money that they are a dB scale...

And so...how does one boost...just turn the "k" to a higher "k"? But how does one boost or cut them bottom tiers of these doubled-stacked knobs, when it goes from 300 "up" to 1.5?

I did ask someone else this, and since he didn't explain direction of knob turning for boosting/cutting, I figured turn it clockwise to increase and vice versa. This is the 246 tascam portastudio I have; but I think I'm beyond manuals, even if I had one.

gracious thanks.

Towards 4 o'clock would be the direction to boost, and towards 8 o'clock would be the direction to cut, with 12 noon being Unity Gain...

I hope this helps a bit...

Edited November 23, 2005 by Prometheus

[Donna]

Prometheus, it helps again. I appreciate your time and attention and I think you shine when writing tomes. The panning examples are excellent like the earlier ones.

The htz knob is identifiable by the number scale you used; the other numbers are written in "k" from 1 (8 o'clock) to 2.8 (noon) to 8. I'll just keep reading John's 2 articles and your replies till it makes sense.

A good next major sort of goal would be the sequencing Finn and you speak of. btw Finn, I know these planning things...some of the time away from music I learned to sew and boy it became just like doing songs...funny how one field can keep a person somewhat connected to a different one.

Got an adapter for the effects unit! We'll see if it actually works aside from lighting the LED light.

Edited November 23, 2005 by Donna

[Donna]

Finn...but how did you put those 4 tracks onto another tape? That second tape'd have to be run at the 4-track's speed, right? And then how did you get the 2 tracks back onto the 4-track?

_____________________________________________

Just so I understand....the scale Prometheus you used where you think of sub bass as 20-60 hz and so on, when you got to low mid it was in kz (from 200 hz to 1kz you think of low-mid range as). So the continuation of the hz goes up to kz, right? Is the kz the same as the "k" on my upper knob, think you (from 1-8...the sibilants' range roughly as you think of it)?

OK, so on each knob there is one tier for gain and one for frequency. Do you mean by "sweep the center frequency", the gain stays put and you rotate the frequency around till that something bad jumps out, then make the cut (w/ the frquency tier)?

PS! Hope you see this - just re-checked John's eq frequencies article and see my question of kz being continuation (as it were) of hz has been answered.

Edited November 24, 2005 by Donna

[Prometheus]

Finn...but how did you put those 4 tracks onto another tape? That second tape'd have to be run at the 4-track's speed, right? And then how did you get the 2 tracks back onto the 4-track?

_____________________________________________

Just so I understand....the scale Prometheus you used where you think of sub bass as 20-60 hz and so on, when you got to low mid it was in kz (from 200 hz to 1kz you think of low-mid range as). So the continuation of the hz goes up to kz, right? Is the kz the same as the "k" on my upper knob, think you (from 1-8...the sibilants' range roughly as you think of it)?

I would say more from about 5 - 10 for the sibilants... Sibilants are basically over pronunciations of the letters "s" and "t" and therefore found on vocal tracks, although I have recorded other sounds that can be loud at the same kind of frequencies, like a flute for example. The standard solution to sibilants is to use a Frequency Band Limited Compressor called a "De-Esser" with which again you can set a center frequency and squeeze down the annoying "s" and / or "t" sounds... I've noticed sometimes that over emphasis of the letter "f" in vocals can cause trouble even higher up still... Again whether dealing with sibilants by EQ or De-Essing, it is important to be judicious, because if you overdo it, you give your singer a lisp...

OK, so on each knob there is one tier for gain and one for frequency. Do you mean by "sweep the center frequency", the gain stays put and you rotate the frequency around till that something bad jumps out, then make the cut (w/ the frquency tier)?

Precisely... You're thinking like an engineer now. But very carefully, and make sure you compare the sound before and after the cut a few times to double check that it is an improvement...

Edited November 24, 2005 by Prometheus

[FinnArild]

Finn...but how did you put those 4 tracks onto another tape? That second tape'd have to be run at the 4-track's speed, right? And then how did you get the 2 tracks back onto the 4-track?

First we mixed those 4 tracks over to an other tape machine as 2 tracks. Then we recorded those tracks back to 2 tracks on the 4 track again. This was way before MIDI (around '86, if my old mind serves me right, I was 17 at the time), so there were no syncing to worry about. Usually I would lay the bass track while my pal did the drums on the first two tracks. Then I'd sing while he did some keyboards on 3 and 4. After mixing down we had two more tracks which I would do some backing vocals on - and usually some more guitar or keyboards. We actually have the results of those tapes on DAT somewhere ... turned out pretty good ...

Anyway - if you have no midi sync to worry about, you can just mix the tracks down and back again - thus freeing up 2 tracks. You'd want to save the first 4 tracks in case you want to mix down further.

[Prometheus]

We actually have the results of those tapes on DAT somewhere ... turned out pretty good ...

I like this sentiment 100%... Sometimes, throwing digital technology at a problem is just not the way to solve it at all... There is no substitute for good, competant musicians who work well with each other and make the best use of whatever tech is available...

[Donna]

Good tips re: the s/t etc; Gentlemen, thank you for clarifying. Don't know if this'll be appreciated, but your answers beg more questions.

For Prometheus, how do you decide where the gain should be in the first place? And, how would you compare the sound of the cut before and afterwards, by mixing down or ? Also, I've only been using headphones (we're working on speakers!) if that makes any difference.

Btw, that first round of mixing "that I liked" re: the housewife success thread is STILL the best one and I've no idea how to duplicate it (tho I've been making records of knob placement since)

For Finn...what kind of machine, another 4-track (that yoou mix the 4-tracks-onto-2-as initially)? I'm not understanding that, nor how the 2 tracks are then recorded back onto 2 tracks of the original 4-track. It does help hearing how you/your partner played on the tracks.

I'll bet the results are good and I remember DAT! Wow, what happened to it?

[FinnArild]

For Finn...what kind of machine, another 4-track (that yoou mix the 4-tracks-onto-2-as initially)? I'm not understanding that, nor how the 2 tracks are then recorded back onto 2 tracks of the original 4-track. It does help hearing how you/your partner played on the tracks.

Well - we were lucky enough to have a 1/4 inch stereo reel player available at that time. So we mastered the 4 tracks down to that and then played it back and recorded it onto track 1 and 2 of the 4 track machine (on a new cassette).

Any stereo recording device will do as the intermediate (as long as it's relatively silent).

[hariossa]

exellent thread!!!

thanks a lot

[Prometheus]

Good tips re: the s/t etc; Gentlemen, thank you for clarifying. Don't know if this'll be appreciated, but your answers beg more questions.

For Prometheus, how do you decide where the gain should be in the first place? And, how would you compare the sound of the cut before and afterwards, by mixing down or ? Also, I've only been using headphones (we're working on speakers!) if that makes any difference.

Btw, that first round of mixing "that I liked" re: the housewife success thread is STILL the best one and I've no idea how to duplicate it (tho I've been making records of knob placement since)

The received wisdom on that is that you want your recording input level to peak between -3 and 0dB, although I've found that if you have reasonable quiet pre amps it can make life easier to go quieter than this, to about -6dB. That way, if the performer gets over exuberant, they aren't going to overload the signal chain.

Another method of avoiding overloads are to "Ride the gain" ie. if you are recording a vocal track for example, and you know the singer is going to sing far louder going into the Chorus of a song, you nudgel the fader back a little bit. I try to avoid this kind of practice because I prefer to keep as much of the natural dynamics as possible, but if you have a singer with poor microphone technique or dynamic control, it can be a necessity. What I tend to do with that is put a thin strip of masking tape over the fader strip at the loud point, and another one at the quiet point, then you can flick from one to the other accurately and quickly.

Another surefire way to avoid overload to tape is to insert a brick wall limiter at the end of the signal chain right before the signal goes to tape, and then you can guarantee that no overloads will occur to tape.

What to watch if you are chaining effects units together is that you have to make sure none of them are overloading at any point in the signal chain, unless of course you're deliberately using clipping as an effect, which can sometimes be interesting in the analogue domain.

Edited November 26, 2005 by Prometheus

[Donna]

Ulp, I'm beginning to know what you're speaking of (brick limitor! tho I know not what it is exactly - think I get what it does!)

Thanks for the primer on faders/gain/overloading. Funny, it seems a common problem I've seen perusing here and elsewhere is the mix being too low volume and/or distortion/overloading is there. One little hill atta time. (Do your drums really stay at those lower peakes?)

For general interest re: my machine tascam 246 and it's eq section function, just received an almost definitive reply from a trusted friend (Loo). He's many years upscaled from a 4-track (that's why I say "almost" definitive), and I knew he could help this past month, but the man just became a father, dursn't bother him...yet y'all have been here to walk me through these things.

Here is Loo's bit: "As I recall, the 'bottom' button ( ie tier) is the sweepable portion of the EQ knob (for example, picture a normal "Equalizer" with 30 adjustable

> faders, 10 for bass, 10 for mid, and 10 for high). I'm pretty sure

> the bottom half of the knob on the 246 machine (and my 244 machine) is

> the same as adjusting faders, ie, all the way turned down would be

> dropping the low end, and all the way up would be boosting the high

> end, and everything in between (I hope this makes sense). The 'top'

> half of the knob is the overall volume."

And Prometheus, Loo then went on to pass onto me a mixing trick someone gave him long ago, and it was exactly how you explained to do it: sweep the frequency til one finds something BAD, then cut that.

Loo ended his note: "In

> other words, rather than trying to find 'good' EQ levels and boost

> them, he'd search for 'nasty' EQ levels and drop them."

I hope to be re-recording this week.