I have a theory. Please tell me what you think: In the cue I write there is a point I want to hit with a zillion and one toms, and other percussion, to create a BAM that will make one jump out of his seat. So, one would theoretically take watever sound he had and make it louder, but say I have tom 1, 2, 3 and 4 loaded into kontakt on different channels. on that place I play with all toms. In Kontakt, none seem to clip. All of them playing loud together cause, obviously, a clip of lets say 2db. But when you listen to it, you can't HEAR any clip. Is there or is there no clip? The question is : instead of starting to reduce each tom's volume, and consequently reduce the effect, could one just keep it with that virtual clip, and later add an adaptive limiter that will limit it? How do you guys go about it?

Hi Ben, digital distortion is so different to analog... Some say if you don't hear it, you're ok. But personally I'd keep the levels low and then use an adlimiter and compressor etc to give it the loudness. Also worth thinking about relative levels (against other parts). :)

Hi Ben, Rounik's right in that technically, you should try and get a better balance without anything clipping. If you need to make it louder(est) after the mix then you can master the track accordingly. However, it is worth noting that individual channels can 'clip' without anything really being heard. 32 bit floating point headroom enables that to happen, but you really really really shouldn't let the main output channels 1-2 clip. You might not be able to hear anything, yet, but outside of the potential distortion when you put it all down to 16bit, you also run the risk of burning a very dodgy CD if you use that file. CD players do not like digital overs at all and most consumer players prefer around -0.3db headroom.

Good advice guys, thanks. Rounik, can you elaborate more on relative levels?

Hi Ben, By relative levels... I think I was referring to the idea of mixing at a low volume... not from your speakers (these can be turned up) by in Logic. Some Mixing Engineers prefer mixing and adjusting levels of their tracks so that the output 1-2 never gets above -12 or -6db... then as Yadj says making it louder later :) So, for example say the music track is peaking at -12db and you add the Toms or 'large sound' you might try having it peak at -8db/-10db ... So, it's relative level is louder than the rest of the track. Another way to give the Toms more perceived loudness is to apply an EQ cut on the rest of the tracks to give the Toms more presence. Also panning the Toms slightly to the right or left will give them more space in which to catch the attention of the listener. HTH Rounik

Good morning Rounik, I work with a lot of virtual instruments. Every instruments I add contributes to the overall output, right? So even if the new instrument doesn't peak louder than the preexisting instruments, the overall loudness will increase. So do I go and reduce ALL of the channels' levels every time I introduce an instrument that increases it? I think an important point to make here is that one probably (correct me if I'm wrong) should keep the RMS levels at -12dbs or -6db during the COMPOSITION stage and keep the PEAK levels at under 0db at the MIXING stage (compressing and limiting obviously)

[quote:]I work with a lot of virtual instruments. Every instruments I add contributes to the overall output, right?[/quote] Yes, this is correct. [quote:]So even if the new instrument doesn't peak louder than the preexisting instruments, the overall loudness will increase. So do I go and reduce ALL of the channels' levels every time I introduce an instrument that increases it? [/quote] That's what I'd do. Just swipe over the channel strips in the Mixer and lower the levels there. Another option is to place a Gain plugin on the Output 1-2 and set it to a negative value. [quote:]I think an important point to make here is that one probably (correct me if I'm wrong) should keep the RMS levels at -12dbs or -6db during the COMPOSITION stage and keep the PEAK levels at under 0db at the MIXING stage (compressing and limiting obviously) [/quote] I guess that's one way to approach it. You don't mention the 'Mastering' Stage. Mixing tends to be when you find the right balance between all the instruments and tracks and add volume, panning and fx to give space, depth etc to the mix. Then, traditionally the mix is bounced to a single stereo audio file and sent to the Mastering engineer to add the final sheen and give it the 'radio friendly' loudness / sparkle etc. While Mixing, I tend let the Output channel peak at about -6 or -3. As for individual channels, their levels are not really too important in terms of numbers as long as it sounds good and I'm not overloading my output 1-2. They could be at -14db and that would be fine for Logic! It's not always a good idea to try and mix and master at the same time. So, try leaving -6db or -3db at least of headroom to the finished Mix and then Master it. Have you checked out Steve H's Mastering Toolkit tutorial? [url]http://www.macprovideo.com/tutorial/Logic401[/url] and/or the Mixing series tuts? They are really useful IMHO. :) Rounik

Yes, I meant MASTERING, not MIXING indeed. I might actually have 401, not sure, I have a bunch of the DVDs. Anyway, I think I have a better idea about how to approach it, which is basically what I've been doing before :) Thnx for the brainstorming