My band is working with a friend/enginer who is mixing a record for an outside mastering guy for the first time. I often read the board here and respect your opinions greatly. I'd be interested to hear what suggestions any of you have for the mix.

Here are a couple of mixes:
After the Show

Calypso

Dear Condi

Also, how "hot" should an unmastered mix be?

Your friend / engineer will probably know what to do, jusging by the recordings.

Sounds good.

You obviously want the recording to be fairly hot to use up as many bits as possible (assuming it's digital).

Ask the mastering guy what (s)he would like and (s)he may mention bit rates / bit depths, etc. Don't mix it down to 16bit if it was originally 24bit or likewise go from 96KHz to 44.1KHz, ie. maintain the quality of the original recording.

It's best processing at the highest bit rate and depth possible before going to CD.

Rodabod wrote:You obviously want the recording to be fairly hot to use up as many bits as possible (assuming it's digital).

If you are recording at 24 bits, which you should be, you do not need to use up as many bits as possible, in fact you shouldn't. The point of the extra headroom that 24 bits provide is that you can record at moderate levels (w/peaks at -6 dBfs) and still have plenty of headroom. It is much better to leave at least 6 db above 0dBfs so that the mastering engineer has "room" to make adjustments. You would have to record right arounf -40 dBfs @ 24 bits to achieve the same resolution as a 16 bit recording.

michaeltheangryrussian wrote:If you are recording at 24 bits, which you should be, you do not need to use up as many bits as possible, in fact you shouldn't.

It is much better to leave at least 6 db above 0dBfs so that the mastering engineer has "room" to make adjustments.

That is an interesting point you have there. Obviously 24bit gives a greater dynamic range for tracking, but I don't see why you wouldn't want to use the majority of the resolution for the final stereo tracks. The more bits you use, the less quantisation distortion you get and also digital processing works better with a greater number of bits to work on. Since software works internally at 32bit or 64bit, it will allways be bit-reduced anyway.

Also, what is the idea behind leaving "room" for the mastering engineer? I'm not sure I understand this - I can't imagine them processing a track to the point where it induced an increase of gain which would cause clipping. Also, if they were using software, this is no longer an issue.

Most importantly, I don't think you should make a specific effort to "use as many bits as possible" when recording at 24 bits because it implies specifically trying to keep your mix as hot as posisble without clipping, which inevitably either leads to wasted time trying to get it that hot without clipping (without any noticeable increase in sound quality), occasional clips since you are operating so close to the maximum threshold, or the most common outcome... limiting your mix to "squeeze" in as many bits as possible. The nice thing about 24 bit is that you don't have to worry about that. Having your mixes peak at -6 dB is not going to add any signifcant amount of quantization distortion (if any at all). This is common advice from professional mastering engineers. An excellent and very helpful resource on preparing mixes for mastering is "Mastering Audio" by Bob Katz. Very well written and easy to follow.

mike

I see what you mean.

I should have been more specific in referring to the volume of the final stereo mix rather than each specific track.

Obviously, setting the volume of the final stereo bounce is a simple choice - searching the peak value on the mix bus (computer) and then specifying the master fader gain to suit.

As you said, for tracking, worrying about using maximum bit depth is a waste of time and has risks.

thanks for your replies!