Lost in the mix, caught in a trap, no turning back...

[Dr. Varney]

But I don't quit my 8-5. I'm completely stuck... Please help?

I am trying to mix a radio play using a DAW. I have loads of VSTs and don't understand how and where to use them.

The project is taking months because everytime I open the project, I look, listen, try out some new VST mastering thing, then close the project and go away confused.

Now it's got to the point I don't want to look at this project at all.

But I need to make it sound right.

I have downloaded articles on compression, limiting, mastering etc but I am struggling with all of the technical terms and concepts.

There do not appear to be adequate explanations for the functions of controls in instruction manuals, rather the same gobbledegook seems to be rehashed over and over, assuming I am familiar with the language. I am forty three and passed my English A levels at 16, yet I struggle to understand half of what people are saying to me these days. Yesterday someone asked me if I had a blue tooth... I still can't work out why anyone would be so personal. That's between me and my dentist, surely.

Anyway...

When I load up a limiter/compressor thing, I can't hear differences when I turn the attack & release knobs. I only hear anything different when I turn up the gain. There are way too many knobs and switches bearing labels I am not familiar with and they confuse me. I can't work out what's actually happening when I employ any of them. I'm not even sure whether they are the right sort of tools I should be using in the first place.

I need help - from the ground upwards.

What sort of stuff should I have in my signal chain, to make voices and effects sound right, for radio/audio book presentation?

What kind of volume levels should I be aiming for?

I've staved off coming in here with loads of annoying dummy questions like "What do I do now?" and tried to research this on my own butnothing is making sense to me and I can't tell if the mix is right or wrong because I'm too familiar with it.

It's an amateur project, so there is no question of money. It gets done on my computer or not at all, unless someone is willing to help me out. But I want to learn, I don't really want anyone to do it for me, unless it helps teach me for the next time.

I figured perhaps the friendly and practical approach I get from you guys might help give me an idea of what to do next.

I have stripped out all the EQs and compressors and god knows what and I'm left with a blank slate again.

What do I do now? What comes first?

Many thanks in advance

Dr. V

 

[TetraFish]

I don't know enough to start throwing out a step by step but I know you're on the right track. Listen to what you have in the raw recordings and try and decide what's wrong with them in the first place. Once you find something that needs fixing look for a solution for that problem.

As I've heard others say here before a really good recording wont require much tweaking at all.

I think some of my best mixes are the ones I did before I knew about compression etc.., Just a little EQ and I was good to go.

 

[gecko zzed]

I have stripped out all the EQs and compressors and god knows what and I'm left with a blank slate again.

This sounds like a very sensible thing to do.

How does it sound when you just play the unadulterated files?

If they dissatisfy you, what is the nature of your dissatisfaction?

On the other hand, do the plain, untouched files sound that bad? Could they actually be acceptable for the purposes of your project?

With a compressor, the main functions are:

Threshold. This is the point at which the compressor starts compressing. A common technique is to set this so that the peaks get hit, while the rest burbles on underneath.

Ratio. This sets how much the signal is compressed once it kicks in. 2:1 is a light compression, whereas 10:1 is a heavier effect.

Attack: This determines how quickly the compression effect will kick in. A fast attack (or small time) will knock off the high initial transient spikes associated with (for example) percussion. Slow attacks will allow these to pass through.

Release: This tells the compressor when to stop working.

Knee: tells the compressor how to apply its effect (whether to immediately compress, or whether to do it more gently) once it kicks in.

What settings you specifically need depend on the nature of the problem you're trying to solve. Recommending a setting is difficult without knowing this.

However, if you are struggling with it, there's no harm in trying a 'vocal' preset (most compressors come with a range of presets of variable usefulness) and see whether you like the sound of it better. If not, fiddle with the settings, or just leave it out.

 

[Dr. Varney]

How does it sound when you just play the unadulterated files?

'Essy'. I don't like the sudden rushes of highs and upper-mids. I'd like to address those without losing good, overall treble.

Overall, nearly all feel like they are in need of a mid-range cut. The question is, where to place the EQ - before or after the compressor?

On the other hand, do the plain, untouched files sound that bad? Could they actually be acceptable for the purposes of your project?

Many are but the main characters (especially the women) are sibiliant.

With a compressor, the main functions are:

Threshold. This is the point at which the compressor starts compressing. A common technique is to set this so that the peaks get hit, while the rest burbles on underneath.

Thank you. My first question is how do get to see/understand what's happening? I can use the Maximus which shipped with my FL package to actually see a visual graphic but when listening to the effects of the compressors in the Antress Modern package, for instance, I really can't tell enough about what's going on, to know where to set the threshold...

Also, there's some confusion about threshold. I can only guess that by turning it clockwise I make it higher. Does making it higher allow it to work at lower peaks or vice versa?

And the same with attack. Lower means faster, right?

Ratio. This sets how much the signal is compressed once it kicks in. 2:1 is a light compression, whereas 10:1 is a heavier effect.

Attack: This determines how quickly the compression effect will kick in. A fast attack (or small time) will knock off the high initial transient spikes associated with (for example) percussion. Slow attacks will allow these to pass through.

Release: This tells the compressor when to stop working.

Knee: tells the compressor how to apply its effect (whether to immediately compress, or whether to do it more gently) once it kicks in.

...Or the ratio, attack, release and knee.

Since attack and release make some sense, I don't quite understand the purpose of the 'knee'.

However, if you are struggling with it, there's no harm in trying a 'vocal' preset (most compressors come with a range of presets of variable usefulness) and see whether you like the sound of it better. If not, fiddle with the settings, or just leave it out.

I have tried a few vocal presets. I'm not sure which sounds best.

I have tried this, the 'Sweetboy T-Sledge' as it's been recommended:
http://forum.cakewalk.com/tm.aspx?m=1391345

I am wondering if you might be able to guide me on what the controls mean?

I've also tried the 'Pushtec 5 1A Leftover Lasagne' from the same page. I'm not sure what kind of unit it represents (is it a compressor?) but it makes for a kind of richness in the lower mids and bass region but I really can't fathom out how to actually use it or what all the 'push-pull' terminology could possibly mean. Otherwise I have to say it sounds smooth in operation, even though I haven't the foggiest what I'm doing with it. I'd like to get to know this creature a bit better.

Thanks for your input

Dr. V

 

[miroslav]

What sort of stuff should I have in my signal chain, to make voices and effects sound right, for radio/audio book presentation?

It appears that you are in a "fix it in the mix" frame of mind.
You mention how everything is sibilant and in need of mid-range cut when you remove all the plug-ins you've slathered on.

I hate to say it...but it sounds like your problems are not in the mixing and which plug-ins to use. They are in the tracking that you did.
For typical spoken voices...that stuff should just fall into place without a lot of processing/FX if it was recorded well in the first place.
Granted, you CAN probably fix a lot of sibilance and other EQ issues in the mix...but again, you seem to be searching for plug-ins that will just “make it so”.

Yeah...strip everything away...start clean...keep the FX/processing to a minimum...IOW...apply what is REALLY needed to fix things, rather than stringing a bunch of plugs on everything and hope they will improve it.

And here's another ugly reality...
Sometimes you just need to start from the beginning and use your current effort as a learning experience rather than "throwing good money after bad" as they say.
Yeah...it's a lot of work that appears like wasted time...but take from it what you can and then apply it to the new effort. It's all a matter of how important this is to you whether it bums you out or you take it in stride and move forward.

 

[Chibi Nappa]

'Essy'. I don't like the sudden rushes of highs and upper-mids. I'd like to address those without losing good, overall treble.

Overall, nearly all feel like they are in need of a mid-range cut. The question is, where to place the EQ - before or after the compressor?

I don't see why you need a compressor at all. Just EQ.






...and before anybody tells him to side-chain a compressor to de-ess, first he would have to know how to use a compressor and that would confuse things even further.

 

[Dr. Varney]

I'd rather not use compressors if I can help it. That's something I can learn later on.

Yes, I think I can take it on the chin and move forward. Get this one out and learn from it. Mind you, I'm not at all sure what I can learn from this. I can't think of anything different to what I've already done.

I don't want to use a compressor because I just don't get it but some advice on other mastering tools would be useful.

When I have played around with a multi-band compressor, it has been like using a great big equaliser. I'm told this is not how they should be used.

I don't understand what you mean by 'tracking' though. I've already tracked everything by placing the wavs on the playlist. That won't make any difference to the sound, only the sequence of events. Right now, I am mixing... right?

As for the recording - my mic and recording environment are non-negotiable, fixed conditions unless I get a recording machine in future so I can record in the bathroom or a cupboard full of clothes and books.

Perhaps the one thing I would do different is not gate noise on the recordings. The lesser of the evils is a constant hiss, rather than this stuttering, 'breathing' effect it's produced. You probably would notice a bit of hiss but it would not be as annoying as hearing it switch in and out, as it does. I can't change that now because all the wavs were saved over after I gated them.

My other main question is - what volume levels am I aiming for? Do I need to boost the volume from -24dB upwards or should I leave it at that?

Dr. V

 

[Dr. Varney]

...and before anybody tells him to side-chain a compressor to de-ess

I don't know what side-chaining means. At this point, I'm not sure I want to.

Dr. V

 

[Steenamaroo]

i think stripping it back to butts was a good idea, and i think the idea of eq only (for now) might be a good one too..


if you don't understand the workings of a compressor and you're finding it difficult to hear the differences when you turn the knobs, perhaps you should take a very short test track,,like 4 snare hits or something and loop it,,then play with a compressor..

the effects will be more apparent on a repetitive sound.


with your threshold as high as it goes, the comp should do nothing.

if you bring the threshold down, the compressor should act kinda like your turning down the volume briefly every time there's a sound which is above the threshold.


if you keep the threshold down and the snare sounds squashed,,,then you can move the attack time to hear its effect...

an attack time of 10ms, means that the snare will hit, and the compressor will allow 10ms of sound before it starts squashing the volume.

this means you get the snap sound of actually hitting the drum, without squashing it, but the 'body' of the snare is squashed.

release time is how long the compressor continues to squash for, each time it's triggered.

if you're compressing a tom with a heavy ring, or a sustained bass note, you would hear the effects of release time being adjusted, where you wouldn't necessarily hear it on a short hit like a snappy snare.

is that helpful?


if you imagine riding a fader instead of compressing.

threshold is how loud something has to be before you turn it down.

ratio is how much you turn it down (based on how loud the sound is)

attack is how quickly you react to the sound (long attack means you missed the start of the loud sound)

and release is how long you hold the fader down for, before you let it back up)

 

[Dr. Varney]

perhaps you should take a very short test track,,like 4 snare hits or something and loop it,,then play with a compressor..

the effects will be more apparent on a repetitive sound.


with your threshold as high as it goes, the comp should do nothing.

if you bring the threshold down, the compressor should act kinda like your turning down the volume briefly every time there's a sound which is above the threshold.


if you keep the threshold down and the snare sounds squashed,,,then you can move the attack time to hear its effect...

an attack time of 10ms, means that the snare will hit, and the compressor will allow 10ms of sound before it starts squashing the volume.

this means you get the snap sound of actually hitting the drum, without squashing it, but the 'body' of the snare is squashed.

release time is how long the compressor continues to squash for, each time it's triggered.

if you're compressing a tom with a heavy ring, or a sustained bass note, you would hear the effects of release time being adjusted, where you wouldn't necessarily hear it on a short hit like a snappy snare.

is that helpful?

VERY!!! Thank you very much, mate.

I can now see I was barking up the wrong tree by trying to fix vocals with a comp. This sounds like a very educative exercise... and I'm gonna try it.

I'm sure it will come in more useful when I'm mixing music.

Kind regards

Dr. V

 

[Chibi Nappa]

Perhaps the one thing I would do different is not gate noise on the recordings. The lesser of the evils is a constant hiss, rather than this stuttering, 'breathing' effect it's produced. You probably would notice a bit of hiss but it would not be as annoying as hearing it switch in and out, as it does. I can't change that now because all the wavs were saved over after I gated them.

I hate to say this, but start over. Re-record the entire thing. Gated voices will never sound right.

 

[miroslav]

I hate to say this, but start over. Re-record the entire thing.

Yes...that's what I was saying too.
It sounds like there's already too much mess to dig out of.

By "tracking" I was referring to the initial process of recording the individual tracks (not about placing recorded files on tracks).
IOW...start from the beginning and focus on getting the sound you want during tracking....rather than trying to manipulate everything into shape during editing/mixing.

Spoken word shouldn't need tons of FX/processing to sound right.....

 

[Chibi Nappa]

Spoken word shouldn't need tons of FX/processing to sound right.....

Hell, it really shouldn't need any.

 

[Dr. Varney]

I hate to say this, but start over. Re-record the entire thing. Gated voices will never sound right.

Under most circumstances I would but it took a very long time. Waiting until there is no one around, to start recording; finding a day I could take off work in the first place. Then getting into it and finding that by the time I'd rehearsed and set up it was time to switch off and start again on the next available day... then having to remember where I was last time.... And that's only MY parts... The people who read the other character parts are simply not available for another recording session this year.

Basically - NO WAY!

I know its the right answer but it definitely ain't gonna happen. No way, Jose.

If it breathes, it breathes. It's either that or not have it done at all.

Having said that, I have managed to cover up the gated imperfections for most part, using the PushTEC thing coupled with some really simple EQ on the upper mids... so for what it's worth, I think I'll just about get away with it (this time).

But I have definitely learned (perhaps the hard way?)

The good news is that other readers who submit stories do not go anywhere near the lengths I do, to create a decent sound. We're literally talking five dollar mics into standard onboard sound, running Audacity. I'm not going to get run out of town by presenting a little bit of distortion...

However, my aim is to get REALLY GOOD at this game - for my own satisfaction. Even through the frustration, I enjoy it and I feel I've achieved something that is a transferable skill to my music making hobby.

The next project will be better.

Dr. V

 

[Dr. Varney]

Those who said "less is more" are absolutely right and since I've downloaded a new bunch of DSPs I find I'm using less of them. Some of these 3rd party plugins seem very well designed, not too heavy on CPU and certainly smoother than some of the ones that shipped with my package. Except... I have not yet found a reverb that comes anywhere near as natural and smooth sounding as the Fruity Loops native 'fruity' reverb. It neither looks nor sounds particularly exciting, but I'm pleased to say, it does its job without unecessary tinkle or any other annoying artifacts. And since I have enough problems as it is, I count this as a blessing.

Do you find this? That verbs come in so many different flavours and it's not what or how you dial 'em but they just all have a different character?

So I've got the 'fruity' basic handling my narrator and I've put one of those Modern Flashverbs on a send, with only the slightest touch, just to handle the big room scenes.

Dr. V

 

[oretez]

I can now see I was barking up the wrong tree by trying to fix vocals with a comp.

Dr. V

not necessarily . . . Every pressure wave has a distinct ASDR, crudely Attack Sustain Decay Release . . . individual character of most audio is rooted in the transient, attack, portion of the wave, controlling and balancing dynamics will effect 'how' we hear certain frequencies without appreciably modifying relative balance of frequency. A spectral graph of frequency response might not be (or depending on parameters it might be appreciably effected pre & post compression, but how humans interpret, i.e. hear, those frequencies might be. Frequency; Time; & dynamics which can be thought of has to how a specific fundamental pressure wave is expressed over time, and how fundamentals interact with partials over time, how clusters of harmonics interact over time; are the three primary variables we play with to balance individual pressure waves in a mix. To keep things interesting they are not independent variables.


no, multi band compression is not specifically for bullying frequencies . . . but modifying the dynamics of frequencies selectively rather then one size fits all will impact balance of frequencies in a mix . . . but as general rule conventionally effective multi-band EQ (as opposed to specific egregiously applied FX) can wait till you have better control of dynamics processing and how to balance frequencies in a mix

'gates' are simply another form of dynamics processing and typically if you are hearing unwanted artifacts from the gating process you need to adjust the settings. But effective gate, when recording theatrical performance can be a powerful tool, but completely eliminating ambient sound from an audio play is seldom as effective as it might at first seem . . . ambient @ threshold of human hearing provides significant, necessary, contextual info about what we hear. One does not listen, pleasurably, to anything, for any length of time in an anechoic chamber.

Initially dynamics processors (compressors, limiters, gates, etc.) were purely hardware and implemented which parameters and how they could be controlled differently. Different circuits had different 'tapers' as to how the actual compression was applied (somewhat akin to the slope of a hardware EQ) As dynamics processors evolved a feature toggling between a sharp and gentler taper or knee was added and the knee was added to marketing flyers as distinguishing spec.

in a full featured digital, software dynamics processor determining the 'knee' is simply part of selecting which parameters you modify and specific settings. If you use an emulation then the 'knee' is simply another preset and you have to toggle between soft and hard until you hear the difference then do this enough times on enough different material to get a sense of when it does and does not make a significant difference . . . hard knee percussion, soft knee acoustic guitar but that is not even general guide line. . . juch something that with certain specific hardware compressors I might or might not apply . . . again specific ASDR of a specific wave in context of everything else in the mix determines what and how dynamics parameters are applied (a pointer to why 'presets' are deadly)

but back to the original speculation . . . with spoken word material, depending on how it was 'tracked' (i.e. recorded, captured, etc. in the first place) it is possible that little or no compression needs to be applied to individual tracks. It is quite possible that simply adjusting volume via faders during mix, with some light limiting of the master track might be all that is needed . . . and it is quite possible that even limiting might not be required on the master. It just depends.

 

[Chibi Nappa]

I can now see I was barking up the wrong tree by trying to fix vocals with a comp.

Dr. V

(extensive encyclopedia entries on compression, the origins of mixing, patents on the light bulb, and the population of Brazil)...

...but back to the original speculation . . . with spoken word material...it is possible that little or no compression needs to be applied...

Wow. oretez just managed to stretch the sentence "I agree" into 10,847 words.

I'm impressed.

 

[oretez]

but the actual answer to most of Dr V's dilemmas is that if you have to fix it in the mix not at the source do as little as absolutely necessary

fuck eq, compression and verb unless absolutely necessary even if the initial capture seems to suck

but . . . if you have some fucking idea of how and what they do (rather then that they bring charm to the process) modifying frequency, dynamics, how a wave is expressed over time can be powerful tools in creating something that might, possibly teeter on threshold of art . . . particularly if what you have 'tracked' or have been handed is a fucking shoebox full of yellowing news clippings and someone wants you to turn it into something else

so an even simpler, though true, blanket response to every single inquiry in the 'Newbie' section is: (wait for it, roll & rim shot) 'IT DEPENDs'

when in doubt don't do anything at all. if the fix doesn't seem to help don't fall into the rabbit hole of fixing the fix, go back to original and try a different process, a different parameter

the best any individual wave is ever going to sound is at the individual capture . . . adjust frequency, dynamics, time primarily to balance individual waves in a mix

and while I am aware that my posts tend to fly in face of, defy the 'less is more' concept I suppose I could just drop by, laugh at everybody here and figure that if this is the competition then my revenue threads are secure . ..

but unfortunately 'it depends' might not always be as helpful as it might appear

and at the moment I've time on my hands that leaves me bored and restless . . . in about six weeks I expect to be too busy to spit again . . . and would typically, in these conditions go camping out of range of cell phones and internet . . . nor do I think that I am imparting 'wisdom' here as much as participating in an internal dialog . . . challenging what I believe as much as anything else . . . as to whether anything posted is of any use to even a single individual is something about which I am historically indifferent

 

[Dr. Varney]

Um... Ok! Yeah! Rock on...

Dr. V

 

[Steenamaroo]

lmao! i was thinkin the same thing