Hacking a Bus Compressor When Making Stems

Nowadays it's pretty common to use a bus compressor over your whole mix. This can add those mythical qualities like 'glue' or 'warmth'. Or a bus compressor can make things louder, not that there's anything wrong with that. But what happens when you need to send someone stems of your mix? If you just print the individual elements through that magical bus compressor, your stems won't sound right when you combine them later. The guitars won't be reacting to the snare drum hits, the vocals won't duck the guitars, and the kick drum won't combine with the bass guitar to create that platinum record sound. So what can we do? If you have a compressor with a sidechain input, it's pretty easy. 

Print your whole mix, but don't print it through your buss compressor. Bypass that, and you'll have a 2-mix of your song minus any compression. Go ahead and assign that printed 2-track to the sidechain input of your buss compressor. Now when you go to print stems, the compressor will be responding just as it would have with everything combined, but it is now showing it's total effect on individual elements of the mix!

View fullsize

Click to enlarge

Compression Tips Using Clip Gain

With the introduction of Pro Tools 10, we had access to a new feature called Clip Gain. In essence, you could change the gain of a clip, pre-fader and pre-insert, right on the clip itself. This was cool for a number of reasons, but there's an aspect of clip gain that I hadn't used until very recently. Basically, you can insert a compressor on a track and use clip gain to drive that compressor harder during different parts of a song.

Adjusting Dynamics with Clip Gain

Adjusting Dynamics with Volume Automation

These two screenshots show the same guitar parts, with my Clip Gain adjustments highlighted in the top photo, and the automation moves that I made in the bottom photograph. On the surface, all I'm doing is adjusting the level in two different ways, but it's a lot more powerful than that. Both these tracks have a compressor on them, and I wanted to bring out more excitement and power in specific sections of the song, so I used Clip Gain to drive the input of the compressor harder, and the resulting tonal changes I could then ride using volume automation.

Try it out for yourself, and see if you can't use both Clip Gain and volume automation in your work.

Parallel Master Bus Compression

As an addendum to my Bus plugin video, here's a little video explaining how I use a multi-parallel signal chain to achieve bus compression with three different compressors. It's a kind of Michael Brauer technique, but mine is a bit simpler and set up differently. 

This doesn't work for every song, but if you're interested I go over how I set this up and the practical benefits.

My Bus Plugins

It's become a big part in how mix engineers get their sound, and so I thought I'd share a little video going over what plugins I use for my master bus, and how I set them up. Take a look and feel free to share what you use on your bus.

Featured plugins include:

Slate Digital's Virtual Channel Collection, Virtual Tape Machine and Virtual Bus Compresser

Massey's L2007 Limiter

Brainworx bx_hybrid and bx_meter

Maag EQ4

---

- Not long after this video was shot I picked up the Clariphonic DSP plugin from Kush Audio. I've only had it for a couple days, but it seems like an amazing tool for adding high end sheen to things. So far I've been using it in place of the Maag EQ. It's not a normal EQ, in that it works only on the upper mids and top, and it does so in parallel. It's almost scary how much high end you can add without it sounding weird. It sounds so natural that it's effect is really only noticeable when you bypass it and everything sounds dull. 

I guess it goes to show you how right I was when I said in the video that these plugins on my master bus can change next week. Always be open to changing things up!

Make Things Wider By Making Them Mono

Ever since stereo records came out in the 1960s people have been trying to get their mixes to sound as wide as possible. People think that having all their tracks recorded in stereo means that their mixes will be guaranteed to be wide, but that's usually the opposite of the way things turn out. If you're guitars are panned hard left and hard right, and so are your pianos and drums and everything else, then there's no contrast in your mix. Everything is in both speakers and the ear has trouble separating your mix.

Contrast can come from many different avenues, and taking mono sources and panning them around is one of the easiest and best ways of creating contrast in a mix. If you are given source files that are all stereo, make sure they're truly stereo, and not just a stereo track. If the source isn't truly stereo, consider throwing away one side and panning the remainder to help it sit in the mix or stick out of the mix. You could also use any number of plugins to reduce the stereo width of a track. Most plugins that aim to widen the stereo field can also reduce the stereo field, so don't think of them as one-trick ponies. Confining the stereo width of an instrument can help you place it in a specific area of the stereo spectrum.

Quickly Hide Tracks in Large Pro Tools Sessions

When working in very large sessions, it can be very easy to get lost in the sheer number of tracks. In a post-production mixing environment there are usually lots of tracks being routed to many different subgroups and output summing assignments. I have a template that I've set up for working in 5.1 surround sound, and you can see a glimpse of the mix window above. As you can see, there are 141 tracks in the session. While I may not use all of them, the template is set up with that number because it's not uncommon to end up somewhere near that many tracks. Daunting, yes, but with color coding and a clear naming scheme you can quickly get your orientation. Here's a little tip to quickly hide all the tracks you don't want to see, which is very useful when you've done all your editing.

Here's a look at some Dialogue tracks, and you can see that they're all summed through the Dia ∑ aux track that lives at the end of the mix window, along with summing auxes for all SFX, Music, and Atmosphere tracks. If you are using newer versions of Pro Tools you can right click on the output assignment, and click Show Only Assigments To... This will hide (but still enable) all tracks that aren't a part of the signal chain for that assignment.

This will allow you to see only Dialogue tracks, any Dialogue reverb tracks, and the Dialogue summing aux track. Because I like to keep all my summing auxes at the end of the session, I can have the best of both worlds: I can see how everything is coming together when I scroll to the end of the session, or using this tip I can focus on every aspect of the Dialogue all on the same part of the screen. Of course, this also works in the Edit Window, allowing me to dive deeper into editing whatever I feel is needing some work.

View fullsize

To get back to where you started, just right click on the output assignment again and select Restore Previously Shown Tracks.

Another aspect of this tip is that if you're using aux tracks to sum all of your audio, you can quickly focus on fine-tuning the final pieces at the summing stage while keeping an eye on the final output.

As you can see, I have all my final summing auxes next to my final output aux and my print track. These are at the end of my session, and are tracks 134 through 139. With two clicks I can go from that to seeing only those tracks, which are now tracks 1-5.

Mastering For Fun. Not Profit.

Right off the bat, I should tell you I'm not a mastering engineer. I know too much about the 'black art' of mastering to know that I have no place doing it. That said, in my mind the line between mixing and mastering is much more vague than ever before. I consider myself a mixing engineer, but I will craft my mixes with mastering in mind, and will try to coax things to a place where a mastering engineer won't have to do too much to fix my mixes. It's already competitive enough in this field, and I need to be able to make my clients feel like they're getting everything they pay for. Sadly, the pay usually stops at the mixing process, and I find myself doing quasi-mastering for clients, since a lot of them don't understand that mastering should be an additional step. C'est la mixing vie...

I recently tried my hand at doing some 'serious' mastering, in competition with a friend of mine. We both used the same 2-track mix, which neither of us mixed but were close to the project. We gave ourselves a couple hours, so I didn't get a chance to do all the requisite listening tests in various environments (listening through an open door in another room being a particularly revealing test). I figured I had the opportunity to try some new techniques that I haven't really used in a mastering sense. Here's a look at what I did:

First, a bit about the session itself. I had two tracks of the original mix, using one as a reference of where I started off. I also imported some other reference tracks to see how far I might have to go. There are three ways in which the audio flowed: the original audio track, which then had three tracks of parallel processing being sent to from sends, and which all flowed into another final aux track where more processing was applied.

I added a little Avid Trim plugin because I wanted to be able to have some control over how hard I was hitting the first plugin in the chain, the Slate Digital VCC, and I pulled the original mix back -1.5db, just because I felt I was starting off a bit too hot.

I then added the Slate Digital VCC, which I set to an API setting and increased the drive a little bit, just to add some harmonics and saturation. The API setting helps to tighten things up in the lower mids, but still adds plenty of harmonics, so it's not as clean as the SSL setting on the same VCC plugin. I originally set it to a Neve setting but found it too unruly and hairy in the lower mids, and I knew that this track needed those areas tightened up.

Next I added some very specific cuts using the brilliant Plugin Alliance bx_hybrid mastering eq. This eq has a great feature in which selecting any frequency knob shifts that band into a very tight Q boost and you only hear what lies inside that very narrow boost. It's great for finding problem areas and really zeroing in on them. I found a boom at 70Hz, again at 372Hz, and found some nasty sibilance/cymbal zing at 10KHz. At this stage of processing, I find that cutting out bad frequencies is much better than boosting good frequencies, as now things sounded more evened out, using the more transparent technique of cutting with EQ rather than boosting. 

Another nice aspect of the bx_hybrid plugin is that there is a very transparent stereo widening control, and I used it to add some width. This added more spread to the stereo field, but it also helped by making some more room up the middle by clearing some things further out towards the side. There was also a very deep 18db/octave lo-cut at 26Hz, just taking out any subsonic rumble that could be eating into my headroom. While we may not be able to hear it, that doesn't mean it's not there. Since this song had some quite aggressive synth bass parts, along with some deep kick samples, and since they were in stereo I used the Mono-Maker feature to make anything below 60Hz mono, and this seemed to do a little to help stabilize the real low end. That's a little technique I lifted from the days of transferring songs to vinyl, since having irregularities in very low frequencies could make the needle jump out of the groove.

So far I haven't done anything dramatic at all, merely adding some harmonics and tightening up the sound. It was only at this point I felt I could do some additive EQ. I turned to a very simple yet great sounding EQ, the Maag Audio EQ4.

I assume this is a vaguely Neve based plugin, but what matters is how it sounds, and this guy sounds great for what it can do. I added 1dB of 40Hz, took out 1dB of 2.5KHz, and added 5dB of 40KHz. Yup, 40KHz. Don't ask me why, nor how, but it can really add a great professional sounding sheen to things. It's called the Air Band for a reason.

The only other things I added on the original track was the Massey L2007 Limiter, not doing anything other than making sure I wasn't clipping before my next gain stage. I think I had the output set to -.1dB, and I honestly don't think it was doing anything, but it was more of an insurance policy. Last in line was the Brainworx bx_meter, showing me the levels of processing I was adding, and I kept it onscreen the whole time. I could have set the meter to show me many different kinds of metering, including the range of Katz K-Meters, but I kept it to dBVU since I was only concerned with the final output.

At this point I thought I'd do some experimenting. I had read about mixing engineer Michael Brauer's method of using many compressors in parallel, and sending different stems and instruments through them before finally summing them at the output stage. You can find more than that basic explanation here.

I ended setting up three aux tracks, each with different compressors designed to work on different aspects of the mix. First up was my go-to compressor, the FET Compressor from Softube. Obviously it's an 1176 clone, but there are some added controls that make it so versatile.

It's set up aggressively, with a fast attack and fast release, which on this plugin is very fast. Unlike most plugins that have a sidechain input (but not the original 1176 hardware unit) this one has a built-in high and low cut that only affects how this compressor reacts. Using a lot of cutting, I got the FET Compressor to really grab onto the midrange of the song, and really start pushing it into an in-your-face sound. Because it's on its own aux track, I slowly pushed the fader up until I started hearing it. The final fader setting was somewhere around -25dB, so it's only a very small part of the overall sound, but very effective in making the heart of the song sound more exciting.

I wanted to make the drums and transients pop a little more, so I turned to a great drum compressor, the Sonnox Inflator. I pushed up the Effect fader and especially the Curve fader, which makes things a bit brighter and 'pokier'. Again, since this was in parallel I only added enough to gain a punchier drum sound, and I think the aux track ended up somewhere around -20dB.

For my next trick, I felt that I was losing some of the immediacy of the female vocal track, so I looked for some way to bring that back, but EQ would also accentuate the snare and guitars. I tend not to use multi-band compressors a lot, but I knew there was an extreme preset on Waves' C4 Multi-Band Compressor that could work.

The original track and the three parallel tracks all summed to another aux track, on which I added some more processing. First up was the Slate Digital VCC Buss plugin, again set to an API setting, with the Drive pulled back just a little.

Next was another instance of the bx_meter mastering EQ, doing some more small cuts. You're probably thinking this is too many plugins, but I feel that it's the opposite of death by a thousand cuts, rather, since I'm only asking each plugin to do a little bit of work, no one plugin is doing any heavy lifting and it all adds flavor and spice and everything nice.

It's pretty extreme compression in the high end, but by adding just a little bit in parallel I found it acted like an Aural Exciter, just putting back some energy into making the vocals pop out with some more diction and immediacy. The parallel aux fader ended up around -30dB, so a little bit of a lot ended up sounding good.

Another non-invasive low-cut at 24Hz, just making sure the subsonic range stayed clean, then a 1.2dB cut at 75Hz, and some more cutting at 210Hz and again at 4.9KHz, taking out some edginess I was hearing. I felt I could add a little bit more width, so I added it at this stage.

Now for my final bus compressor, which in actuality is three different compressors, mangled together into the beast known as the Slate Digital VBC. I rearranged the order in which I hit the compressors, having it hit the FG-Red compressor first, just to tickle the transients and add some energy. A 2:1 ratio and medium attack time meant it wasn't too aggressive. I also set a pretty high high pass filter so that it wasn't triggered by the massive kick sound I already had going.

Moving on, I had the signal hit the FG-Grey compressor, which is a clone of the classic SSL Bus Compressor. Here's where I added some aggressiveness, with a 4:1 ratio and fast release. No high pass filter meant that I was getting some pumping from the kick drum, which is what I wanted. I liked the sound, but ended up backing off on the Mix control, to add more of the original signal back in.

The final stage of the compressor was Slate's version of the ultra-rare Fairchild 670 compressor. I usually look to a Fairchild to smooth things out and add heft and warmth, but I decided to try a different route. This plugin can do Mid/Side compression, where I could address things in the mono field separately from the stereo field. I wanted to make sure that the vocals, bass, kick and snare stood out, and since these were all mono sources panned up the middle I added some compression, maybe 2-3dB at the loudest parts of the song, with only a dB or two of make-up gain. All this did was make those mono elements pop with a little more excitement, while the stereo field of guitars, cymbals, keyboards and synths kept their same dynamics. The Side compression wasn't doing anything, since I only wanted to effect the Mid signal.

At this point I was very very close to being happy with the final mastering job I'd done, but I wanted to add one more flavor over the entire mix, and that was the Slate Digital VTM. I backed off the Input so it wasn't being hit too hard, but with normal bias, at 30ips and a 2" 16 track machine, I found it softened the high end just a little, added a little bass saturation, and just gave it a little more energy overall. 

My final plugins were another Massey L2007, not doing anything other than making sure no quick transients clipped the final output. Another bx_meter was kept onscreen to see how the different gain stages were interacting, and that's it. Lots of plugins, I know, but I think the end result was pretty tame considering how many plugins I ended up using. Since they weren't doing much, they all just added layers of flavor and energy.

Here is the full un-mastered song, as mixed by Ryan and Shawn.

Here's the full song after the mastering job I quickly did. It sounds very different than the un-mastered version, but if you've been playing the audio files that accompany each processing stage I think you'll see that I'm not asking any particular plugin to do too much, and I think that that philosophy gives pretty decent results. 

Again, I want to stress that I am not a mastering engineer, and that this was done in a couple hours on a dare from a friend, so I used some techniques that are a bit strange, just because I wanted to see how they would work. Thanks for reading this very long article.

Using Shared Reference Images Between Aperture and Capture One Pro

I've been using Aperture as my primary Digital Asset Manager for many years now, but I've recently started to branch out into using Capture One Pro for a lot of my RAW image conversions. Aperture's strengths are in areas like deep asset management/metadata, overall depth of adjustments, and the way its various GUI's are easily pulled up and sent away with the press of one button. Where I find Aperture to be lacking is in a couple of "finishing" steps, like output sharpening and exporting multiple sizes/resolutions simultaneously. This just happens to be some of the big strengths of Capture One. Here's how I've reconfigured my entire library to easily take advantage of using two different RAW convertors. 

My work computer is a Mac Pro, connected to a RAID 5 SAN, with a 4 drive RAID 0 internally set up for maximum speed. All my different bits of work reside on the internal RAID 0, which is regularly backed up to the RAID 5 SAN. Originally, when I was just using Aperture for my DAM (digital asset management) I was importing all my photos as managed files, meaning that Aperture created one large library that contained everything, my RAW photos, JPEG previews, metadata, adjustments etc... This is a great no-hassle way of doing things, and is how I operated for years. The problem with starting to use Capture One is that I wanted to be able to share all my original RAW files with both applications. This is where using Referenced Images comes into play.

Referenced Images refers to (no pun intended) setting up your own system of DAM and telling your software to 'refer' to those files instead of importing them directly. Things can get a little tricky, and due diligence is required to make sure nothing gets out of place or not backed up properly, but for what I'm trying to achieve, it's the only way to be.

As you can see, this is a birds-eye view of my folder setup for all my photography.  The most important thing is my RAW Archive. In here, I've arranged all my RAW photos by either project or year. Every client gets their own folder under "Client', and all my general shooting gets placed by year, month, and day. From this RAW archive, I can point both Aperture and Capture One to refer to the same photographs, therefore the only thing inside the Aperture folder is my Aperture library holding all the adjustments, previews and metadata. Inside the Capture One folder is my C1 libraries with all the adjustments I make living inside those C1 libraries.

Once I finish editing all my choice photographs, whether that be through Aperture or Capture One, I export the finished products to the Output folder, with sub-folders for large format printing, web uploading, etc... 

This is a pretty involved workflow, but once it's put in place, I find that the benefits are easy to reap, and it is a very solid system of managing all my photographic assets. Nerdy, I know, but a lot of photographers who don't have a system in place have trouble finding things and it becomes very easy to lose stuff amidst the thousands of photographs we might take in a year, or sometimes even in a month.