I suspect this is the same problem I posted in "Rendering Glitches" yesterday.....It happened across a bunch of tracks for me at hotspots. (Am remixing 7 year old tracks recorded when I thought headroom was something to do with low bridges )
Not sure if I fully understand the render process but I think NR is dropping the gain after checking a peak level & after rendering is reapplying gain. If it is misreading the level at all then when it's reapplied gain based on this it may cause clipping??
Haven't had time to check this, but I'm guessing saving your MWC program with around 4db cut on the output & using this in the render would stop it clipping.
Or use Wavelab batch process & edit Nebula to take the input down & push the output up.(or use a gain plug-in before & after to do the -18 thing.)
in aes17 things are around 3db quieter, so maybe wavelab and nebula are matching here. When you select a level you are rising the hottest rms value, but you could get dangerous peak levels, like in the real world. The console should have an headroom, though. So try to raise levels manually (+10dB), if the signal clips check it with alessandro. Maybe even -18dB is not safe for particular tracks, I think the hottest value should be handled at a lower gain
vicnestE wrote:I might export the drum samples at 3dB quiter before rendering like I used to. It used to be ok when I lower the drum signals.
it will not work, because renderer will try to compensate the missing level. I think you should select something in the range -22dB in nrgui. Anyway try by yourself, maybe the issue is different and not directly linked with gain staging. Add 10 dB to your track, and try to render it in wavelab.Does it clip? and if you add simply 6 dB, does it clip?