I'm running a 5 track mastering Montage in WaveLab. Each track's output does not exceed 0dBFS. But, the composite signal at the output of the master section is +6dB. The seemingly 'louder than legal' signal coming out of my studio monitors sounds distortion free.

I render that mix to a 32-Bit Floating Point stereo file; & there are spikes at every snare hit w/ 3 or 4 consecutive samples flat-topped; & a 25 sample plateau that had a combo cymbal & snare hit goin' on (ala Keith Moon)...& the file doesn't sound bad… but it does indicate 6 dB of overload on the master section meters. Normal so far, I guess. [& I'm guessing that a FP file doesn't necessarily show clipping…but the overload data is there, somewhere.] I save the 32-Bit FP file as a 24-Bit fixed; & it looks the same (flat-topped in the exact same areas), but playing back that 24-Bit file, the master section meters don't indicate any clipping, because those transients were sawed off at 0 dBFS in the render.

Now…forget about rendering for a moment. Is WaveLab truncating the 32-Bit Floating Point signal to 24 fixed to send to the AI's DAC's…w/ it's plus 6dB peaks also flat topped? The output meters in my Lynx console say it's not...because they're showing 6 dB of overload when I play the 32-Bit file. Why is that? There's no such animal as a 32-Bit converter...so the software has to send the sound card a fixed point, 24-Bit signal. And this 24-Bit signal is truncated at the top, with everything above 0dBFS gone. Then why are my 'last stage' meters reading +6 dB?

Thanks,

mark4man
WaveLab 5