Questions often arise from mix engineers who are preparing a mix for mastering.
These are some questions I'm often asked:
-
What is the desired format and resolution I should deliver for mastering?
-
How much headroom should I leave in my mix for mastering?
-
What average or RMS volume level should I aim for in my mix?
-
If I give my client a loud mix reference, should I send that to mastering as well as the regular mix?
-
Should I avoid stereo buss compression in the mix I send to mastering?
-
How about EQ on the master buss? Is that OK?
Each mastering engineer will have their own preference, but here are my answers to these questions, based on my opinion and experience.
What is the desired format and resolution I should deliver for mastering?
For digital recordings, the mix delivery format should be at least 24-bit stereo (interleaved or split mono) at the same sampling rate at which the track was recorded or mixed. Thirty-two bit is becoming common in mix delivery these days, which is fine to send as well. It is best to leave the sampling rate and bit depth conversion until the mastering stage, so there is no need to do it on your end prior to mastering.
I do not have a preference as to what sampling rate should be used for recording and mixing because I've heard both great and not so great mixes done at every available sampling frequency. Because of this, I’ve come to believe that a good mix done by a good mixer will translate regardless of sampling rate choice.
I’ve come to believe that a good mix done by a good mixer will translate regardless of sampling rate choice.
The advantage, however, of recording and mixing at higher sampling rates is the ability to deliver the recording to a wider range of release options, and also to be forward compatible with future Hi-Res formats. Recordings made at higher sampling rates are better able to meet the technical qualifications required to be released on HD tracks and through other audiophile type distributors.
I also feel that higher sampling rates translate better when going to vinyl due to the finer resolution and increased detail gained from it. Though some of this detail is lost when transferred to the vinyl, doing the transfer from a higher resolution source is still a better starting point.
There is no advantage to up-sampling your recording to a higher sampling rate, because all of the higher frequency information has been lost due to the filtering used at the lower sampling rate. Up-sampling does not recreate this lost frequency content.
Most mastering studios are well equipped to handle up to 96kHz recordings very easily. Going beyond this, up to 176.4kHz, 192 kHz, 384 kHz, or DSD can also be done in many mastering studios, but you should have a conversation with your ME to make sure. It is sometimes the case that the engineer's beloved ‘magic box' (ie. A-D, D-A, outboard digital gear, or plug-in) doesn't operate beyond 96kHz, so it would be best not to surprise your ME on the day of your session. Letting them know in advance would give them time to seek an alternative tool to use on your project.
I have found that a little dynamic management done at all stages of the recording, mixing and mastering processes tends to yield the best sounding and most cohesive and enjoyable result.
Before I delve into the remaining questions, I want to give an overall opinion with regard to them, since these questions are somewhat interrelated to one another. My feeling is that when preparing a mix for mastering, a client should always deliver the best sounding mix they possibly can, given their experience, their recording/mixing environment, and the tools available to them.
In the best case scenario, the mastering will simply enhance what is already there, bringing more ‘music' out of the recording, making it a more moving and emotional experience, whatever that emotion may be. However, sometimes it is necessary for the mastering engineer to make radical changes, or ‘turn it inside out' as I tend to say, to accomplish this. I don't recommend trying to anticipate what your ME will do and thereby ‘leave the details to the mastering'. I don't recommend mixing in a way that assumes what you think your ME will do, and that they will automatically know what you are thinking and ‘fix’ your mix.
For instance, sometimes folks will opt not use buss compression because they read somewhere that it should be left for the mastering. I don't agree with this at all. I have found that a little dynamic management done at all stages of the recording, mixing and mastering processes tends to yield the best sounding and most cohesive and enjoyable result. More on buss compression later.
Now into the details of the remaining questions….
How much headroom should I leave in my mix for the mastering?
I'm asked this question all the time. I used to give a stock answer of 3-6db of peak headroom, but now I generally leave it up to the client. I am comfortable working with peak levels up to 0dBFS. It’s nice though when the peaks are not super low, like -20dBFS, but even that I can deal with.
When preparing a mix for mastering, what average or RMS volume level should I aim for in my mix?
This will depend on the style of music of course, but I tend to master somewhere between -12 and -8dBFS, with peaks that reach 0dBFS (or sometimes a ceiling such as -0.2dBFS). For a modern pop or rock project, I will aim toward the higher end of this spectrum. For an EDM recording I will sometimes aim a little higher if it’s possible, and if it goes along with the desire of the client. For a jazz or classical recording, I will aim toward the lower end of this spectrum. Overall, the level tends to be dictated by the genre as well as the clients' expectations.
What this means for your mix is that it would be good to be within 3-4dB or less of this average level if possible. Again I’m speaking here of the relationship between the average volume level and peak volume levels, so you’ll have to do some math to figure out where your mix sits in this relationship.
This is just a guideline, of course. Sometimes I find folks who are so worried about their mixes being competitive in average level that they overcompensate, resulting in a recording that feels over compressed, unmusical, and unnatural. I tend to believe there is a middle ground where it can be musical, but also have the dynamics well managed.
Using a pair of real, mechanical VU meters is an invaluable tool for managing your dynamics. They will give you a visual representation of the relative average volume, so I highly recommend using them if you don’t already.
If I give my client a loud mix reference, should I send that to mastering as well as the regular mix?
Yes, absolutely! Otherwise your ME has no idea what sound you and your client have gotten used to hearing in the recording. It’s tough to compete with a louder recording and no ME wants their work to come in quieter than the mixer’s reference, even if it is more musical sounding. It can be done, but it’s a tough sell. Sending the loud mix references to your ME will help point them toward the minimum of what you are expecting back from the mastering in terms of volume. It can also open dialog about loudness in general. Usually the ME will send back something at a similar loudness or more often louder, but if they decide to master it quieter, they will at least have the opportunity to talk it over with the client, letting them know why they have made that choice.
I recommend the loud mix reference to be created by adding a limiter to your printed mix, to simply add volume to it. Ideally, the loud mix reference will sound very similar to your regular mix, just a louder, more limited version of it. It should not be radically different than your regular mix.
It is also good to include documentation in the files you send to mastering stating what limiter you used, and even what settings were used. A simple screenshot or text file will be adequate for this.
Should I avoid buss compression in the mix I send to mastering?
This is an area where I feel there is great misunderstanding. Often I will receive a note from a client saying that they printed the mix with ‘no master buss compression’ and my first reaction is ‘why not’?
Even worse, sometimes a client will mix while listening through some buss compression, but then will take it off to print the mix they send to mastering, thinking that the ME will have better compression options than they have.
This latter occurrence, however, is not at all advisable because if your volume, tonal balance, and automation decisions have been made with the buss compression on, it can, and most likely will, sound completely different when the buss compression is removed.
This ends up becoming a completely moving target for your ME. This will require your ME to guess what type of compressor you were using (like VCA, optical, Variable-Mu, feedforward, feedback, etc.), what settings you were using (ratio, attack and release times), and how much compression you were applying, in order to recreate your mix as you intended it to be. Even if you have given your ME your loud reference mix, it will be a difficult task for them to recreate your ‘real’ and intended mix.
How about EQ on the master buss? Is that OK?
I generally don’t recommend sending a mix to mastering that has had EQ applied to the master buss. To me, it is better to leave the overall EQ shaping to the mastering engineer. Chances are, your ME will have better sounding tools at their disposal, or at least will have a more accurate listening environment to be able to make better decisions with regard to the EQ of the recording.
If you must apply EQ on the master buss, I recommend doing it only in small amounts. If it is used on the loud reference for the client, I recommend leaving it in also for the mix you send to mastering as well.
In answering some of these more common questions I encounter from mix engineers and clients, I hope I’ve been able to shed light on the details surrounding mix file delivery as well as on some common misconceptions I have encountered.
Lastly I would like to say that, when preparing a mix for mastering, when in doubt ask your ME their opinion. A little communication goes a long way toward achieving a great sounding recording with which everyone is happy. Now stop reading and get on with that mix!