I shoot for -6dBfs PEAKS at all times until the mastering phase - No individual track or mix will exceed -6dBfs. No limiters, no compression for the sake of volume.

I finally got a few of my "regulars" (studios that send mixes regularly) to start mixing like this and they've been thrilled with the results also.

To me the peak level you mix to is pretty irrelevant for the way I work as long as your output ceiling is -0.1dBfs at max, as I can easily attenuate the level I send to my process chains to where ever I need it to be. What really hurts the mastering process to me is if you have squared off the tops of your wav forms in your mix. This just makes additional processing sound grainier and in the case of transfer to vinyl master can sometimes break up when we're trying to go for max level.

I recommend never to use a peak limiter on your mix and only to use 2-bus compression if it's because you like the particular color that box gives you and not just to get level. If in doubt send a second mix in without any 2-bus processing on at all. If you like the level you've achieved with a limiter on a reference disc send that in too so that the ME can see where you'd like to push the average level to. But in general I always prefer to get an uncompressed unlimited mix.

Best regards,
Steve Berson

Too loud can mean too squashed and most likely does. The fact that a mix goes to "0" doesn't mean much, it's how it gets there. If you are limiting the 2-bus, Then that is what they mean. You are smashing the transients. You can go to the top, just do it nicely.

Another thing to consider is the bit depth of your mix. If you've recorded and mixed in 24 bits, then your 24-bit mixdown could peak as low as -48 dbFS and the ME could still deliver the same exact 16-bit results. The dynamic range of a 16-bit recording is about 91dB, compared to 24-bit's 139dB.

If you're sending a 16-bit file, though, then I'd check with the particular ME beforehand to find out what he prefers, or like the others said just make sure you're not limiting or clipping your peaks (or compressing for sheer volume).

HB

Most people say send 24-bit files, is it not okay or even better to send 32-bit files. I have the capability to do either.

Michael, help. I don't know what "squashed" means.

jblack wrote: Most people say send 24-bit files, is it not okay or even better to send 32-bit files. I have the capability to do either.

Depends on the studio. I can accept 32bit IEEE files with no problem as my DAW app allows me to open them without any format conversion. Main thing to realize is that a static (i.e. stored) 32bit IEEE file actually doesn't contain any more real audio data than a 24bit file does - it's just a 32bit container with 24bits data + 8 bits mantissa. For best compatibility with the largest cross section of mastering studios I recommend just sending mixes at 24bits.

And even if you recorded the individual tracks at 16bits you'll get better results mixing to 24bits - and I don't know of any reputable mastering house that can't deal with 24bit files at this point - so there is no reason to me to ever send less.

Best regards,
Steve Berson

jblack wrote: Michael, help. I don't know what "squashed" means.

The general understanding of the term "squashed" means highly compressed with a low dynamic range across the track.

Best regards,
Steve Berson

what Steve said.