Wednesday, April 21, 2010

Preparing Your Tracks for Mastering

Aside from poor mixing, one of the biggest problems mastering engineers encounter on a regular basis is receiving improperly prepared pre-masters. So today, I thought I'd share some tips on getting your music ready to be mastered with a minimum of hassles for the mastering engineer. Remember, the less time they've got to spend trying to undo problems caused on your end, the more time they'll have to focus on the music itself.

1. Communicate
It's vital to communicate exactly what you want to your mastering engineer before starting the process. Is it important to you to preserve the dynamics as much as possible, or are you looking to compress and limit your tracks back to the Stone Age? It might be helpful if you can point to another release in your style whose mastering you like to give the engineer a practical example of what you want. Likewise, listen to what your mastering person has to say. Rely on their expertise. After all, if you don't trust their knowledge and experience, why'd you hire them in the first place?

2. Leave Fade Outs to the Mastering Stage
Fade outs aren't as common as they used to be back in the day, but if you want some in your tracks, leave it to your mastering engineer to apply. This is partially because any fades you apply will be re-shaped by the dynamics processing the engineer does, but additionally, the mastering engineer probably has access to more precise tools to give you smoother, more musical fades. Again, be sure to communicate with your engineer to give him an idea of the length of fades you're envisioning, etc.

3. Bounce Your Pre-masters With Space Before and After the Song
As a precaution, it's always a good idea to leave about 3-5 seconds of dead space both before and after each song when you bounce them out. This helps avoid unwanted clicks at the start, possible "buffer burp" errors, or prematurely cut off reverb/delay tails at the end. The mastering engineer will appreciate the extra flexibility, too.

4. Leave Your Master Bus Alone
It's becoming increasingly and depressingly more common these days for bands to strap a limiter or compressor across the master bus of their DAW. There are many reasons this is a bad idea. The most obvious of these is that you're tying your mastering person's hands from the get-go. Dynamics processing is damn near impossible to undo, so if you've pre-squashed your audio, destroyed your transients, or clipped your recording, there's not much, if anything the engineer can do to fix it. Ask yourself again why you're hiring and paying for a mastering engineer if you don't trust them to get things right. Same goes for EQ. If your mix sounds unbalanced to the point that you have to apply an EQ to the entire track, it's better to take the time EQing the individual tracks until the balance is where it should be. Putting anything across your master bus is usually a pretty good indication that you haven't got the mix right yet.

5. Max Headroom
Be sure when you bounce your tracks out, that your meters aren't slamming into the red. Most DAWs these days will emulate the soft clipping you would get from an analog mixer, but again, you are squashing the dynamics before the mastering engineer even gets their hands on it. So be sure you have plenty of headroom on your bounces (within reason, obviously).

6. Garbage In, Garbage Out
One of the biggest misconceptions about mastering is that it can make a crappy mix sound great. Sure, there are some extremely talented mastering engineers out there, but they'd probably be the first to tell you that cases like this basically turn them into professional turd-polishers. Sure, it might be a slightly shinier turd, but at the end of the day, it's still a turd. So don't expect outright miracles from your mastering engineer. The final product is your baby, so take time to get it right.

7. Check and Double Check
A good number of problems that arise during a mastering session are from totally avoidable, stupid mistakes that weren't caught before they're sent out. So be sure to devote some time to quality control. I usually like to give the bounces a listen once on my monitors and once again with headphones. The headphone run-through makes it easier to pick up some of the more subtle problems, like, say, cut off audio on a track, or a cough you forgot to edit out of a vocal take. This can be a bit tedious, since you've probably listened to your track dozens, if not hundreds of time in the process of mixing it, but if it can save you from any hiccups and delays, I think it's well worth it.

Anyone have any other tips I might have missed? I'm sure we must have a mastering engineer or two here... feel free to chime in the contents to share more of your pet peeves or frequently encountered problems!


John said...

Great post Tom, thanks!

Q: I really struggle when mixing/constructing the bass end of the mix, balancing kick and bass synth etc. Is it better to roll it all off at around 30-40hz or leave it alone and let the Mastering engineer fix it?


btw, I'd really like to know how you get your low end, a song like Spark is a perfect example of what I'd like to achieve... Tight and punchy but also fat and squelchy too!

Tom@waveformless said...

John - It's always a good idea to highpass filter out unneeded low frequencies on a track by track basis as opposed to on a master bus. Bass frequency tracks can be rolled off around 30-40Hz or so, but non-bass tracks like synths, strings, pianos, etc. can often be rolled off much higher).

Will's said...

Great Post Tom:

To revist point#3: I have found that the extra space at the beginning and end has so many utilities in mastering that it is almost a crime when an engineer gives me a track with a headmute up to the one. Nowdays, with all of the restoration tools, ambient recovery, etc. there is so much you can do with the information that exists in the head and tail. Most often with acoustic recordings I will grab this info just to pull the wool off of a mix using either spectral cleaning or a denoiser of some sort. While this is great, the hidden benefit is the tiny bits of hearoom this will buy you sometimes as much as .2-.5dB which can go along way if you know how to use it right.



New Alliance East said...

Thanks for this post, having owned and personally run a mix/master studio for 10 years i couldn't agree more. no music day is as hard as one spent tip toeing around a damaged set of mixes ( trying to make more things better than worse! ) and nothing is more frustrating than people being late or showing up without the proper materials.

New Alliance East said...

Thanks for this post, having owned and personally run a mix/master studio for 10 years i couldn't agree more. no music day is as hard as one spent tip toeing around a damaged set of mixes ( trying to make more things better than worse! ) and nothing is more frustrating than people being late or showing up without the proper materials.

Toadie said...

"Ask and ye shall receive..." ;)

Thanks Tom! Wasn't expecting an answering post so soon. I took your advice btw and read the Sound on Sound articles available. They had some good tips, but I suppose I actually wanted some even more finite detail.

On point 5 (Max Headroom) can you give any sort of rough range of where you should mix to for this? I've heard some people say at min -2dB and others up to -5dB - is that a good target range? When I last worked with a mixing engineer who passed it on to a mastering engineer, he asked for tracks bounced with a limit of -2dB.

A lot of guys say during the mix process to compare your tracks to another similar acts and try achieve the same sort of balance in the mix... Are they talking specifically then EQ cos I find the mixes I produce are SIGNIFICANTLY quieter than mastered mixes and to do a proper comparison, I need to raise the overall volumes to do a valid comparison... Is this bad, or should I just focus on the frequencies and their relation to each other in a mix and forget the overall loudness of the mix?

As a result, all too often I end up with a mix that might be louder but that comes at the price of stripping out too much during EQ to maintain the volume without introducing things like a dulled bass end (and that's with compression on the bass drum for punch, rolled off low-end and EQ notches for mud freqs) and very thin sounding synths and vox because of EQing to stop frequency overlap volume spikes...

nulldevice said...

Heck, I've written an entire article similar to this one. It's up on under "about mastering." It goes on and on and on.

Regarding your above comment about lo-pass filtering: yes, yes and yes BUT. What most people fail to relaize is that most lo-pass filters are not hard-lowpass, but do have a slope. So some unwanted subharmonics will trickle through if you're not careful. It's always a good idea to have a good spectrum analyzer to see those things your speakers - or your ears - can't. There are some decent free ones like Voxengo SPAN that do a fine job of letting you see what's happening below about 80hz.

I'm of two minds on the bus compression thing. In about 90% of the cases I think it's a blight on the face of mankind. But that said, used correctly, properly, and more importantly, subtly, a good bus compressor can do some very nice things, especially when used in a parallel arrangement. It's usually best on a bus other than the master, but as is often the case in audio, hard and fast rules are rarely hard and fast.

Anonymous said...

I've just come back and re-read this article and its follow-up comments. Mixing is an important area for many of us for sure, that I'd like to see perhaps another blog post on sometime if possible.

I'd already independently come to similar conclusions about highpass filtering out unnecessary bass on a track-by-track basis, so nice to see at least I've been on the right track in that area. Initially unwanted bass still leaked through with this method but it seems to just be a matter of practice makes perfect.

The max headroom point is still a little unclear to me, in terms of exactly what dB peaks my project should be hitting on a good day... At this point I'm trying to keep them as low as possible in a vague sense. And disabling both normalization and overload protection in Logic when I bounce, which allows me to crank up the DAW master bus gain to generate a final master, since I am doing the mastering myself. This seems to be working ok.

My own advice is to not only use one set of monitors and headphones, but also listen to the final mix through at least 5-10 other popular consumer headphones of random quality, as well a modern hi-fi and car stereo while cycling though all the various preset EQ settings on them. Listening to your mix on all these different setups sometimes reveals weird issues or flaws that can be solved be changing the mix. And at least you can be happy with the knowledge that you know for sure, what your final track sounds like, under nearly all listening conditions. I guess for someone who is great at mixing this is probably not necessary, but for those of us who are not mixing wizards, this trial and error process works.

Ars Divina said...

Excellent! I have a new standard link to send to my clients! Thanks!

Steve Turnidge
UltraViolet Studios