Understanding mixing

Chamber005

Chamber005

Carbonite Member
So I've got my little audio set-up, and while I've always had a pretty good "ear" for music and dialog, all that means is that you can mimic well -- it doesn't mean you have any idea how to actually play the piano.

I can listen to the voice audio tracks and hear that something isn't quite right (or, natural) about them, but Im not sure exactly how to go about finding the best timber.

As most of you know I'm using shoddy equipment, but it's what I've got to work with so there we are.

My mixer is a Xenyx 502. I can pretty much play with the Trim (+10-+60), High Eq 12kHz (-15 - +15), Low Eq 20 hz (-15 - +15), and obviously volume and pan.

In addition, my CM 217 CAD mics have a swtich that takes them from 0 db to -10db.

SO all you audio fanatics -- how would you adjust these levels for the best possible voice recording? Everything is shot indoors (if that matters). I'm getting some decent headphones for the actual shoot, but until then all I have are crappy ones that don't block out enough noise -- which means I have to actually record tracks and play them back on speakers in an enviornment that is not the quietest to begin with.

I'm doing what research I can, but I'm pretty swamped in every other facet and ANY help would be hugely appreciated!!

I promise I didn't mean to skim over audio. I was supposed to have help and no longer have it so I'm scrambling to do the best I can in the few days I've got left.

Thanks!!!
 
by the by, one thing I've definately noticed is that bringing the Low Hz down to about -12 seems to suck out all of the extra..."shinning". Whatever that noise is that sounds like "shinning" when you speak, bringing the low hz down to pretty much nothing seems to eradicate it.

Trim and just plain old volume are confusing, though. I think trim is considered "gain", though, right? And that's different...?
 
The best way to record dialog in the field is flat. Even with high-end portable mixers for film/TV, such as the SD 442, you won't find any EQ controls outside of a bass roll-off (aka hi-pass) filter.

Bass roll-off/hi-pass is useful for taming low-end mechanical noise (AC, refrigerator, etc.), handling noise from a cheap boom pole, and some wind rumble. But you need to have a roll-off that starts at about 80Hz at the least. Some even start at 120Hz. Since your mixer doesn't have that, I'd leave any bass adjustments to post and be careful about mic placement, handling, and wind screening.

You'll want your level on the mixer to peak at 0dB. Unfortunately, you only have a 4-segment LED meter, which won't tell you much. That 0dB peak on the mixer should equal -12dB on your camera.

Since you don't have a tone generator on the mixer, here's a cheat. it requires that you have an iPod or similar with AV breakout cable, or a CD player with line-level out (i.e. not affected by the device's volume adjusments). Go into your NLE and create a reference tone using the color bars and tone tool. 1kHz tone at -12dB, and make 60 seconds worth. Export it as a 16-bit/48kHz AIFF (16/44.1 if you're going to CD) and import that to iTunes or other media program. Dump it to your iPod, or burn it to a CD.

Plug your iPod into the mixer using one of the stereo line inputs (you'll probably need some very inexpensive RCA>1/4" TS adapters to connect the breakout cable, since the mixer has 1/4" ins on 2/3 and 4/5). Set your MAIN MIX control on the mixer to the reference mark (U). Then, starting with the input volume all the way down, play the tone and bring it up slowly just until the 0dB LED on the meter lights solid. You can now bring the input level controls on your camera or audio recorder up to -12dB. You're now calibrated.

The next step is to connect your mic, set it's volume level to U, and then bring the trim control up slowly until the mic, at normal dialog level, peaks at 0dB. With the 12dB of headroom on the recorder/camera, you should be okay if the dialog level occasionally tickles the +6dB LED on the mixer.

"Trim" on this mixer is the same as "gain" on others... it's the strength of the signal coming into the input channel. Volume for the channel is the strength of the signal being sent to the output controls.

And the -10dB switch on your CAD mics is a built-in pad. If you are in high-SPL situations where the mic is just too hot coming into the mixer, hit the pad on the mic. For normal dialog, you probably won't need that.

Lastly, since your mixer's meter is low-rez (only 4 segments), trust your ears and watch the meters on your camera as well. And keep your headphones plugged into the camera/recorder unless you hear a problem. Listen to the mixer as you calibrate it, and check it every once in a while, but your biggest concern will be the signal as it goes to tape.
 
Last edited:
Man, I think that whole iPod thing might be a little out of bracket for me. I'm pretty sure that I'm going to run into an issue with adjusting between the recording device (not the camera used for the shoot) and the mixer. So I'm going to pass on the iPod/CD strategy (although I'm sure it's an awesome cheat and something maybe I can use in the future!).

I was a little lost on a few audio terms. "Bass roll-off"? Are you speaking to the Equalizer settings? While the Low (80hz) setting being brought down to almost nothing seems to eliminate background noise and make things more crisp, taking the High hz up too high or down too low seems to skew the sound. That seems to operate best at zero-ish. But maybe I should be bringing the High lower as well -- and maybe there's some other adjustment to compensate for it?

Also, for the actual mic itself, you said I shouldn't need to use the -10db switch unless I'm in a "high-SPL"? Not sure what that is, but would lowering the db on the mic and increasing, say, the trim to +30 or thereabouts provide additional aid in blocking out any noise?

I guess I don't understand what the difference is between the db on the mic and the db in the trim. Are they accomplishing identical tasks? Or does the db on the mic act as a kind of buffer -- which could help in bring down any kind of a "shinning" sound in the voices.
 
Chamber005 said:
Man, I think that whole iPod thing might be a little out of bracket for me. I'm pretty sure that I'm going to run into an issue with adjusting between the recording device (not the camera used for the shoot) and the mixer. So I'm going to pass on the iPod/CD strategy (although I'm sure it's an awesome cheat and something maybe I can use in the future!).
Click to expand...

I'd highly recommend that you do this. It's the best way to make sure that your levels are within safe boundaries. And once you've calibrated your mixer and recorder (which recorder, by the way?), you're set until you move your setup or change to a different mic. And honestly, once you have it on your iPod, it's really simple to use.

Chamber005 said:
I was a little lost on a few audio terms. "Bass roll-off"? Are you speaking to the Equalizer settings? While the Low (80hz) setting being brought down to almost nothing seems to eliminate background noise and make things more crisp, taking the High hz up too high or down too low seems to skew the sound. That seems to operate best at zero-ish. But maybe I should be bringing the High lower as well -- and maybe there's some other adjustment to compensate for it?
Click to expand...

Bass roll-off/hi-pass takes the bass (low) frequencies, starting at a set frequency (some are 80Hz, some are 100Hz, some even 120Hz) and rolling it off a few dB per octave below that point. This feature usually shows up as a button, rather than a knob, on field mixers. If your mixer's "Low" EQ is set at 80Hz, you can use it as a bass roll-off. But use it carefully, since you want to be able to get rid of ambient rumble but you don't want to get rid of anything that isn't actually there. I wouldn't mess with the High EQ at all.

Chamber005 said:
Also, for the actual mic itself, you said I shouldn't need to use the -10db switch unless I'm in a "high-SPL"? Not sure what that is, but would lowering the db on the mic and increasing, say, the trim to +30 or thereabouts provide additional aid in blocking out any noise?

I guess I don't understand what the difference is between the db on the mic and the db in the trim. Are they accomplishing identical tasks? Or does the db on the mic act as a kind of buffer -- which could help in bring down any kind of a "shinning" sound in the voices.
Click to expand...

SPL = Sound Pressure Level. Sound is just vibration transmitted through particles in the air. The higher the pressure level, the louder we hear the sound. Mics have maximum SPL ratings, above which point they distort (and can even be damaged, in very rare but extreme cases). The built-in pad, which reduces the mic's output level by 10dB, is useful for drum kits, fire works, construction sites, etc. For regular dialog, using the mic's pad and then making up that gain on the mixer just means that you use more of the pre-amp's power in the mixer. Depending on the pre-amp (which isn't the best in those Xenyx mixers), this also means introducing more noise (hiss).
 
Trim/Gain - regulates the amount of signal introduced into the subsequent electronics of a device.

db - Decibels - how loud a sound is in the real world. Normal conversation at 3' to 5' is about 45db to 75db. Hearing damage begins at about 120db.

SPL - Sound Pressure Level - without getting into all of the mathematics and why it's needed, a different way of measuring sound levels. Sound actually moves the air (air pressure). Your ears and microphones measure the pressure levels to determine the loudness of the sound. Speakers do the reverse; they pump in and out to move air.

Hz and kHz - Hertz and kiloHertz - Sound moves in waves and is measured in how many waves are generated per second, or how frequent they are, hence frequency. On a piano the "A" below middle "C" is 440kHz.

The Inverse Square Law - In simple terms, this means that a sound twice as far away is only one-fourth as loud. A sound four times as far away is only one-sixteenth as loud. So a human voice that records with adequate volume three feet away is just one-sixteenth as effective at twelve feet.



Except in the situations described (taming high wind, mechanical noise, etc.) you should not EQ on the set. Most manmade noise coming from machinery generates a hum centered at 60Hz, known as a 60 Cycle Hum. That's why most bass roll-offs are at 80Hz. Do not reduce or increase the highs on the set, you can adjust that in audio post.

The -10db pad has nothing to do with reducing the amount of noise in a given situation, it reduces the mics sensitivity to loud sounds, making it less prone to distortion. You are better off adjusting the sensitivity of the mic when in high SPL situations. Engaging the -10db pad reduces the sensitivity of most budget mics so there is potentially less clarity. Increasing the gain on the mixer introduces more noise from the mixing device.
 
C2V said:
I'd highly recommend that you do this. It's the best way to make sure that your levels are within safe boundaries. And once you've calibrated your mixer and recorder (which recorder, by the way?), you're set until you move your setup or change to a different mic. And honestly, once you have it on your iPod, it's really simple to use.



Bass roll-off/hi-pass takes the bass (low) frequencies, starting at a set frequency (some are 80Hz, some are 100Hz, some even 120Hz) and rolling it off a few dB per octave below that point. This feature usually shows up as a button, rather than a knob, on field mixers. If your mixer's "Low" EQ is set at 80Hz, you can use it as a bass roll-off. But use it carefully, since you want to be able to get rid of ambient rumble but you don't want to get rid of anything that isn't actually there. I wouldn't mess with the High EQ at all.



SPL = Sound Pressure Level. Sound is just vibration transmitted through particles in the air. The higher the pressure level, the louder we hear the sound. Mics have maximum SPL ratings, above which point they distort (and can even be damaged, in very rare but extreme cases). The built-in pad, which reduces the mic's output level by 10dB, is useful for drum kits, fire works, construction sites, etc. For regular dialog, using the mic's pad and then making up that gain on the mixer just means that you use more of the pre-amp's power in the mixer. Depending on the pre-amp (which isn't the best in those Xenyx mixers), this also means introducing more noise (hiss).
Click to expand...


Gotchya, so:

Leave the high eq alone (that felt right to me as well in the testing)

Leave the -10db switch alone (also sounded right in testing)

I can play with the Low hz but just watch to not eliminate all (natural) ambient noise.

And the higher the trim (gain) goes, the more noise comes in. So would you suggest then that I keep the trim as low as possible and just use the volume controls to beef up the message to the recording device? Do people usually go with a happy medium between the output volume and the input gain?

As for the recording device, I'm actually using an old professional cam that doesn't operate any longer as a video camera. I record the audio via its two XLR inputs and then strip the audio and synch in post using slates. I've never heard of it being done before (maybe its a unique situation where someone would have such a cam), but so far the results have been lightyears better than the audio I get from my HF200.

I'll think about your iPod idea, but again -- the camera doesn't work -- no viewfinder, etc. I can only see the volume levels on the side of it which tell me where I'm at. I think the "tonal" iPod step is going to have to be ignored on this go-round.
 
You need a meter that tells you something real. Always record 24 bit so you can safely set the levels a bit lower if you have any doubts. Bad meters will not show spikes that go "over". Record a short at three different volume level. Then go into your audio editor and find the real peak. Adjust accordingly. Better yet, just get a Zoom H2!
 
Chamber005 said:
Gotchya, so:

Leave the high eq alone (that felt right to me as well in the testing)

Leave the -10db switch alone (also sounded right in testing)

I can play with the Low hz but just watch to not eliminate all (natural) ambient noise.
Click to expand...

Yes,

yes,

and in some instances. Again, unless you hear something in the low end that's impeding on dialog, record with the EQ flat and adjust everything in post. It's much easier in the post mix to take out something that's bothersome than it is to add something back in that is missing.

Chamber005 said:
And the higher the trim (gain) goes, the more noise comes in. So would you suggest then that I keep the trim as low as possible and just use the volume controls to beef up the message to the recording device? Do people usually go with a happy medium between the output volume and the input gain?
Click to expand...

Again, set your input channel's volume level at "U" and set your master ouput level at "U" (these may be marked "0" or simply with an arrow or hatch mark) and bring your trim up to where dialog peaks at 0dB on the mixer. This is called gain staging. Pre-amps may add noise, but most of them are mostly clean up toa point. By making sure that your output levels are set at Unity and then bringing the input level up, you're theoretically adjusting the mixer to its sweet spot of signal-to-noise ratio.

Chamber005 said:
As for the recording device, I'm actually using an old professional cam that doesn't operate any longer as a video camera. I record the audio via its two XLR inputs and then strip the audio and synch in post using slates. I've never heard of it being done before (maybe its a unique situation where someone would have such a cam), but so far the results have been lightyears better than the audio I get from my HF200.
Click to expand...

So what kind of camera (format)? And what kind of meters are on the sides: LCD or VU? (VU meters will have the needles that bounce back and forth.)

If this is an analog format camera (SVHS, Hi8, Betacam), then you'll want 0dB on the mixer to equal 0dB on the camera as well.

Chamber005 said:
I'll think about your iPod idea, but again -- the camera doesn't work -- no viewfinder, etc. I can only see the volume levels on the side of it which tell me where I'm at. I think the "tonal" iPod step is going to have to be ignored on this go-round.
Click to expand...

You have meters for your audio recording. That means that the reference tone will be of great assistance to you.

But honestly, this half-broken camera idea isn't ideal. If you can swing it, get the aforementioned Zoom H2, or better the H4n. It'll be a lot less of a hassle to deal with in so many ways, and the record quality will be much improved.
 
C2V said:
Yes,

yes,

and in some instances. Again, unless you hear something in the low end that's impeding on dialog, record with the EQ flat and adjust everything in post. It's much easier in the post mix to take out something that's bothersome than it is to add something back in that is missing.



Again, set your input channel's volume level at "U" and set your master ouput level at "U" (these may be marked "0" or simply with an arrow or hatch mark) and bring your trim up to where dialog peaks at 0dB on the mixer. This is called gain staging. Pre-amps may add noise, but most of them are mostly clean up toa point. By making sure that your output levels are set at Unity and then bringing the input level up, you're theoretically adjusting the mixer to its sweet spot of signal-to-noise ratio.



So what kind of camera (format)? And what kind of meters are on the sides: LCD or VU? (VU meters will have the needles that bounce back and forth.)

If this is an analog format camera (SVHS, Hi8, Betacam), then you'll want 0dB on the mixer to equal 0dB on the camera as well.



You have meters for your audio recording. That means that the reference tone will be of great assistance to you.

But honestly, this half-broken camera idea isn't ideal. If you can swing it, get the aforementioned Zoom H2, or better the H4n. It'll be a lot less of a hassle to deal with in so many ways, and the record quality will be much improved.
Click to expand...


I've already spent what I'm going t spend for this project, so now it's just about doing the best I can with what I've got.

The camera just has little digital readouts on the side. They'll spike and go red if the volume is too high, so it's a pretty easy comparison. I just adjust the mixer's trim or output volume to about 75% of the cam's read out in normal speaking and that gives me a volume I like with little noise (especially with the trim at about +30 and the low hz at about negative 5. I'm not positive what it means to keep the mixer at 0db. Is that to do with the trim or the output volume? The output volume has a zero slot, but the trim only goes between +10db to +60db.

Also, I'm not sure that the recording quality is better from the H2 or H4N. I've iistened to recordings on both of them and the audio quality doesn't seem much improved than what my cam records. Of maybe I just don't have the right kind of ears for audio yet.

I wanted to mention also that although everyone here keeps talking about changing audio in post, I don't think any of the low level programs I'm working with have any sort of compex audio adjustment software. I'm just using Corel Video Studio, and while it has a few audio options, I'm sure it's nothing on the level of prosumer audio editing stuff, so whatever I record needs to be quite close to whatever's being put out, see? I've had to invest a lot of time in lighting and in-camera stuff for the same reason in regards to picture as Corel's color correction software is pretty subpar.
 
Chamber005 said:
The camera just has little digital readouts on the side. They'll spike and go red if the volume is too high, so it's a pretty easy comparison.
Click to expand...

Can you give a specific make, model, and format for the camera? That'll give us a much better idea of what you're working with here and help to answer questions in a more useful manner for you.
 
So I think I've got it to sound as good as it's going to sound. I took the cam's "gain" down to nearly zero. I kept the mixer's output, pan, levels in and High EQ all at neutral, and I took the mixer's gain (trim) up to about +40 and dropped the Low Hz to near it's lowest position of -15. The resulting sound is very "studio" sounding, but I get absolute silences on the camera's levels, even when I shift the mic around (which will certainly help in the case of the inexperienced boom person), and I think the voice sounds extremely clear and crisp, especially considering that my entire audio set up only cost 150 bucks, including mics.

I've definately lost a lot of the surrounding sound, but I figure I can do it how they do in animation and simply create the ambient stuff through recordings of the rooms and any SFX (e.g walking, typing, etc) that the piece needs.

In the end I think I'd rather build a natural performance from the audio rather than trying to capture what's actually there with a bunch of inexpensive equipment -- especially since I have no high level audio software to tweak things with in post. It won't sound at all natural, but at least it will be specific -- every sound and nuance a distinct choice with an ambient blend that (hopefully) ties each indoor location together. And using the mixers gain and bringing the Low hz all the way down also eliminated the tinny sound I was getting.

Bottom line, the cam's levels never go red and the mixer's level never goes above 0db, I'm getting pure silence and crisp-voice tracks at 4 feet. Can't ask for much more than that on my budget!
 
Before you shoot the real thing, do some full test runs - shoot, edit, watch / listen through several systems / setups to be sure you're getting a result you're happy with.
 
JordanBlock said:
Before you shoot the real thing, do some full test runs - shoot, edit, watch / listen through several systems / setups to be sure you're getting a result you're happy with.
Click to expand...


Yeah, I'm going to do what I can with the time I've got left. I have to do a huge make-up app tomorrow as well as build this damn contraption -- and we're shooting this Saturday. So hopefully everything will do what it do.

Also, someone gave me Sony Vegas Movie Studio Pro 9. And while it's nowhere near Vegas Pro, it does give me some help with audio and CC, so that's nifty -- as well as letting me edit in AVCHD, which is the real plus.
 
C2V said:
"Trim" on this mixer is the same as "gain" on others... it's the strength of the signal coming into the input channel. Volume for the channel is the strength of the signal being sent to the output controls.
Click to expand...
Trim or gain on a mixer refers to the sensitivity control on the channel. ie: matching the instrument (mic, line, lav, etc.) for the proper electrical working level of the channel. Think of it as having a 'multiplying effect'. The goal is to make the input device match up to 'unity' or 0db. Further processing will change this level. ie" eq, compression.
Volume is the linear gain output of the channel. It should also line up to 0db (-20 digital) as indicated on the channel strip.

VERY IMPORTANT: 0db analog = -20db digital

Sensitivity is the single most important setting on your mixer as it decides your S/N ratio, distortion, dynamic range and quality of signal. A good way to adjust this level when you are just learning is to raise your channel 'fader' or output to zero and then adjust your sensitivity so that the end result measures zero on your meters. (sort of working in reverse) After you have added your processing, then once again adjust your sensitivity to meter zero.

This is called gain structure and it is very important to maintain zero db throughout the entire chain on your mixer.
 
Last edited:
Sound Chaser said:
Trim or gain on a mixer refers to the sensitivity control on the channel. ie: matching the instrument (mic, line, lav, etc.) for the proper electrical working level of the channel. Think of it as having a 'multiplying effect'. The goal is to make the input device match up to 'unity' or 0db. Further processing will change this level. ie" eq, compression.
Volume is the linear gain output of the channel. It should also line up to 0db (-20 digital) as indicated on the channel strip.

VERY IMPORTANT: 0db analog = -20db digital

Sensitivity is the single most important setting on your mixer as it decides your S/N ratio, distortion, dynamic range and quality of signal. A good way to adjust this level when you are just learning is to raise your channel 'fader' or output to zero and then adjust your sensitivity so that the end result measures zero on your meters. (sort of working in reverse) After you have added your processing, then once again adjust your sensitivity to meter zero.

This is called gain structure and it is very important to maintain zero db throughout the entire chain on your mixer.
Click to expand...

Not to be nit picky but SMPTE alignment levels for analog to digital are tone at 0dBVU = +4dBu signal output level = -20dBFS. But many mixers meters these days, including SD, are calibrated to 0dBVU metered = 0dBu output. That would put tone at 0VU analog on the mixer meter at -24dBFS digital if you were to align strictly by the SMPTE specs. For practical work that might be a bit low so no argument with your statement of alignment levels for real-world recording. EBU specs call for 0dBVU = 0dBu = -18dBFS and that might even be a better standard to use on the set.
 
Do any of you use the technique of shutting off your limiters on your mixer then setting tone on your mixer (using an outside source as CV2 suggested) so that it shows just below or at 0dBFS (+20dBVU) on the mixer and the same on the camera/recorder using a blend of levels that maintain some unity on both. Then, when that has been done, turning the limiters back on the mixer and running the dialogue peaks in the -6 to -8 range on the mixer scale. I have been doing this on cameras that I am not confident of the -12 or -20 dBFS reference mark. I haven't had any complaints with the results but I sometimes worry I might be setting my self up for a fall. Any comments?? Thanks.
 
Sorry, I'm not explaining well. Yes and no. Tone is from outside source sent to mixer, turned up enough to hit the +20 dBU (0dBFS) on the mixer and the camera/recorder. Limiters off. This way I know that what I see on the mixer, especially at the critical top end, reflects what is happening on the camera/recorder. Then when I turn on the mixer limiters I should be protected from blowing over the 0dBFS limit and yet be right up at the acceptable top end of the recorder. So far my SD mixer limiters have held true to their settings and I am able to push up against their set limits (+17+/_) with out losing too much quality. The advantage is that it gives me confidence to run hot and keep the noise floor on some of the less than pro cameras further down. That said, I wonder what down sides there might be to this line of attack.
 
what camera's are you working with that you arent sure of where -20 is?
The only one I used to work with that didn't have clear markings was the dvx
and for that it was pretty much 2 bars below the first red block.
 
You must log in or register to reply here.
Back
Top