quote:

Originally posted by Shahar Audio editor like 'Wavelab' ?! how can I maximize it there?

if you use wavelab, select whole file (ctrl+a), and press 'y' to analyze the file (maximum level, average level, etc). or 'n' to normalize. do not ever normalize to 0 dB, it is better to normalize to -0.5 dB or some other figure, just not 0.

quote:

Originally posted by thecYrus 1. never use normalize (use compressing and limiting instead) 2. if the master-channel clips you have to lower the single channels until nothing clips anymore! 3. the master channel can go higher than the single channel-levels because sound in the same frequency will be added together. if you have a huge differnece it's most due the lack of bad equalising. just my 2 cents cYrus

1. why not to normalize?
2. is it bad to lower master fader and why? if it is bad, then why is it there in the first place?

quote:

Originally posted by Shahar My compressors\eqs look like this: (first comp\eq go to the first melody channel, second comp\eq go to the second channel)

the settings look pretty much ok. but, there is an issue here - if both these channels play at the same time, there is a problem, because they are both equalised the same. it is never good to play two sounds with similar frequent content in the same time - that's why almost every mixing tutorial gives you advice to eq kicks and drums complementary - boost the low bass on bassline and cut the low bass on the kicks, for example. most probably this boost is making your fader jump too high. and another thing - posting comp settings is completely useless. these setting might work great with some material, and can suck hardly with other input material. it depends on audio you are compressing.

quote:

Originally posted by Shahar when you say 'add limiter or compressor to the master' u mean to the mixer master or to the nn-19 master? btw, every synth im using has a compressor..it's ok?

most probably not. you don't need to compress EVERYTHING, use compression only when needed (for example, the melody is playing, and on some notes it is too loud or too quiet).
it looks like you use compression only because someone said it sounds cool, or it is mentioned as 'the tool professionals do'.
basically, what you are doing here is overcompressing each element of the track, what sucks life out of it. you must have some dynamics left, this is just squashing it down.
so, do the following - first, bypass ALL your compressors, play the main loop of your track, and try setting the level (in the mixer) for each channel so that it sounds good, instead of compressing. after spending some time setting this up, THEN try turning on your compressors, one at the time, and leave only the ones that do good things to your track.
in one of my tracks, i only used compressor on a snare (to add punch), and on a kick-bass group channel (to get that nifty flat bass foundation), and a master limiter - that is on an arrangement with 15 stereo outputs. and it sounds real nice, except on crappy 2x1w pc speakers.
oh, and another thing. if you think your bass, kick, leads, whatever need really heavy processing (eq, compression, distortion), try changing the sounds. you cannot make a pie out of mud.

quote:

Originally posted by h.vox 1. why not to normalize?

Because if you have a peak in a track it will only raise the volumn of the track up to the tracks highest peak volumn ie:




What you want to do is hardlimit or compress it to take those peaks away

Something I agree with Rob.

Infact looking at one of ur channel. The channel is way too high. U might consider to lower even more. Or adjust ur Eq setting of ur gain. U pump up at least to +10dB(I reckon) Try minmize everything on ur volunme channel. I reckon that this is ur starting doing of song, right? Adjust the level now. Whenever you layer more tune, the sound will be cover more and more. From there, U do ur mixdown again. Just make sure you dun go beyond 0dB on ur master dB. U can try normalise it provided you do have alot of peaks. Try Rob way. Just finish the song. For me, I will try to do compressor to reduce the peak and also for other reason.

Good luck mate!

___________________
Lets !!!

Production Site
-=:::Currently in the stage of new tune:::=-

quote:

Originally posted by Rob Because if you have a peak in a track it will only raise the volumn of the track up to the tracks highest peak volumn ie:

i get the point, but i never had any problems with peaks like that. if mix is done properly, those peaks will not occur.

quote:

Originally posted by Rob That's why you need a compressor on the master and possibly even on the lead. Why do you need to compress a lead? Well trance is definitly not a genre where the lead has alot of dynamic range, so you'll want to decrease that dynamic range and most importantly stop it from any having abrupt spikes or peaks(especially the case when you layer lots of leads together to make 1 phat lead). Anyway, back to your problem. Try and set this up(wait a while for it to load as it's animated ): Notice how when I increase the volumn of lead1, the compressor(really a limiter as it's set to the highest ratio) kicks in and compresses the signal source. Result = well set it up yourself and listen to what it does.

You don't have any problem with the compressor with it's automatic gain?
For me a normal compressor and a normalize is different then a compressor in reason on the mix which will give some volume dips.
But I haven't tried your setting so you'll hear about that later.

Last edited by MadThijs on Jan-24-2005 at 11:05

quote:

Originally posted by h.vox 1. why not to normalize? 2. is it bad to lower master fader and why? if it is bad, then why is it there in the first place?

1) It takes it to the level of the highest peak- if you are close to clipping or do actually soft clip, then it will sound shit. Its also another process of signal processing- and each process will add noise and degrade the signal further.

2)Gain structure. And noise.

Your gain sturcture is terrible. You may lower a fader to say -10, but by adding a total +15 of boost you have then pushed it past where it was before, meaning you have to drop it even further. No problem with this on one or two channels in a full mix, but when it is on every channel and you drop the master to compensate, all you have done is increased the noise floor by quite a lot.
Cutting the remaining frequency band has the same effect as boosting the one band.

I cant help with any diagrams- i work with analog sound in real studios, not virtual studios, so no screen grabs to show you but the principals are exacty the same. In fact they are more critical in digtal/virtual studios due to the nature of digital clipping and headroom and the lack of soft headroom you get on an analogue set up.

the whole point of good gain structure is to reduce any potential undue increase in the noise floor in the signal at every stage of the chain.

ghetto diagram:

!
. is the signal out of one piece of equipment, where '!' is the music, and '.' is the signal noise

it then goes into the desk where you boost it in eq and gain and it then looks like this:

!
!
.
. Not only have you increased the signal level, you have also

increased the noise level as it is part of the signal. Now imagine this happens with every input and gain stage you go through- think how much noise vs signal is present at the end.
And then, when you drop the master, you are decreasing the signal/noise level even further

like i say, ghetto digram (not accurate by any means) but hopefully my point is made


If you fancy googling it its called 'unity gain' or 'gain structure'

Last edited by Freak on Jan-24-2005 at 10:48

quote:

Originally posted by Freak 1) It takes it to the level of the highest peak- if you are close to clipping or do actually soft clip, then it will sound shit. Its also another process of signal processing- and each process will add noise and degrade the signal further.

it will sound shit because you changed gain? that's crap. it will sound quiet, maybe. but not shit. then again, RMS-based normalizing will get rid of those peaks. btw i once made an analysis of the noise it adds to the process - the noise it adds (if it adds it at all) is about -100 dB, which is far below audible. and that's in digital world. in analogue, that is just riding the master fader of your desk.

quote:

2)Gain structure. And noise. Your gain sturcture is terrible. You may lower a fader to say -10, but by adding a total +15 of boost you have then pushed it past where it was before, meaning you have to drop it even further. No problem with this on one or two channels in a full mix, but when it is on every channel and you drop the master to compensate, all you have done is increased the noise floor by quite a lot.

so? when lowering the channels volume you lower the noise floor, but by raising the master fader you still raise the level of mixed noise from all channels. basically the same thing, the only difference is that you do not get noise from master fader. in a good quality desk, this master should not be too much anyway, right? btw, what we are talking about here is reason, not an analogue mixing desk, which is somehow being forgotten. anyway, you must do SOMETHING to prevent the sound from clipping, right?

quote:

Cutting the remaining frequency band has the same effect as boosting the one band.

pretty much, yes, when you compare those two after getting the same rms levels. but what happens to noise in that case? anyway, you must agree that equalizing is a must-have process in getting a good mix, at least one eq is always needed. personally, i would rather use 5-6 eqs (or how many is needed) to balance the elements and introdude eq noise than making a mix without them.

quote:

I cant help with any diagrams- i work with analog sound in real studios, not virtual studios, so no screen grabs to show you but the principals are exacty the same. In fact they are more critical in digtal/virtual studios due to the nature of digital clipping and headroom and the lack of soft headroom you get on an analogue set up.

that headroom is exactly the reason he is being advised to lower the master fader. i do not think he likes the sound of digital clipping.

quote:

the whole point of good gain structure is to reduce any potential undue increase in the noise floor in the signal at every stage of the chain. If you fancy googling it its called 'unity gain' or 'gain structure'

i just as well might do that, since that is one area of my interest. thanks.

quote:

Originally posted by h.vox that headroom is exactly the reason he is being advised to lower the master fader. i do not think he likes the sound of digital clipping.

I think your master fader should sit at o db on the mixer. In reason that's al the way up. Then you have to adjust the channel faders to keep from clipping. What do you guys think?
Lowering the masterfader will shorten your dynamic range, the amount of different volumes you can use, or do I miss something here?