Clipping on the master channel? What?!

I guess a small amount of clipping on the master rather than limiting can be as transparent, or add some subtle desirable distortion.
Baphometrix’s video series on his clipping workflow might help explain this better..

In the early days of DAWS’ digital clipping would sound horrendous, but these days it seems to act much more like a brick wall limiter would. Really it depends on the summed sound that’s going into it which determines if it works or not.
There was an interview with someone good on Hospital, might have been Cyantific, and he said he always smashed the master a bit red on his own versions that he’d play out. There will be others too for sure.
However it’s probably better practice to avoid doing it if you can. Limit, clip, saturate your groups, then a final dynamic trim on the master keeping it below 0 will make you build all your loudness into the production rather than trying to wring it out at the very end. But then again, if doing all that and then adding another 1-2db of red line on the master sounds good…. then it’s good.
But yes, if a tune of yours is getting mastered professionally they’ll want it -6 on the master with very little on there, ideally nothing. But your slammed clipped ‘self master’ they’ll use as a guide…
Hope this helps!

I would suggest you to just put a limiter in master and mix your track into it without gaining anything with limiter. Personally I have Pro L 2 with Allround/Agressive setting. 100ms attack 0 release 0 lookahead 0 linking truepeak disabled for starting point. Then I mix my tune until it sounds good and just bounce out and call it a day. There can be exceptions but thats my starting point always nowadays.

So you have to think about it this way…..the end result of a mixdown has to basically not go over 0db….because even though in modern 32bit/64bit float mixdowns you have ENDLESS headroom above 0db on your master mix…if you see the red clip light showing on an internal DAW mix then that will not actually do anything other than being louder….the red light in your DAW is just to show you went over 0db….if you bounce out a mix at 32 bit float for instance and then take it into Audition to look at it….if its clipping and you lower the gain you will see that there is no actual clipping at all…it will look totally un-flatlined.

If you bounce the mix at 16 or 24 bit you will see it is flatlined and clipped

But in order for the track to play online or on some kind or recording medium or just playing back on your phone or speakers….it will have to be limited 16/24 bit and 0db somehow.

I never change the main output level on my DAW mixer….its just another algorithm being run over the whole mix for no real reason as you want to make your mix the right levels before the final channel output….so if you need crushing and limiting and all that stuff then better to either do it in a pre-fader insert in your master level or better still getting your mix sounding good and ‘sounding loud’ by all the stuff you do before the master.

Then you can bounce out the file at 32bit float which will NOT digitally clip the recording and then take it into a mastering program all clear and un smashed…..you can always clip and crush the hell out of the master file afterwards but its so much better to have a weighty and uncrushed bounce to then master afterwards.

Hope I make some sense ;0)

32 bit float and 64 bit float files won’t play back on any consumer devices……so eventually you will either have to make it 24/16 bit by converting or make it an .mp3 which will compress the file into 24 bit anyway….

either way at that point of conversion then all the dynamics of your song will be crushed if they are going over 0db in your 32/64 bit mix….so it will be pretty lofi and not the best way to get a clear dynamic master mix, by getting it right in the first place by trying to stay as close to 0db peak in your mixdown in your DAW and then finally limiting it a tiny bit to stop any short peaks and then converting to 24 bit with a clear and nice sounding final mix

More thoughts from Optical here:

https://www.instagram.com/p/CgMPQ_fqkJA/

This is great, thanks for sharing that!

What about running oversampled compressors / limiters and distortion plugs as a means of reducing intermodulation distortion. Would this negate the need to bounce projects at unsupported sample rates?