Tom Efinger's The Dig It Post

We are thankful for Dig It's collaborations with the award-winning directors, producers and actors who have made all of our recent work as rewarding as it's been.—Tom Efinger

Set in the cut-throat world of modern agri-business, Ramin Bahrani's new film At Any Price—a movie that The Wall Street Journal celebrated as "A powerful film"—stars Dennis Quaid and Zac Efron, and is absolutely must-see.

Indeed, Roger Ebert noted, "A great film. Dennis Quaid gives the performance of a lifetime." 

Now playing in NYC & LA, At Any Price opens in SF, Chicago, Boston and surrounding cities and all across America throughout early May. You can click here for a city and release date near you.

Trust us, you'll dig it.—T.E.

You're running a Pro Tools session on an old hard drive and getting error messages. Now What?

Q: Recently I've been running a lot of big sessions off of my old external hard drive (it's a 5400rpm Seagate via firewire). Sometimes I encounter an error that says my "drive may be too slow or fragmented." Is there anything I can do to help this, assuming my drive isn't fragmented? I'm running a Macbook Pro with PT 10 and CPTK.

A: As a matter of fact, there is. There is a a new feature in Pro Tools 10 (HD or CPTK, not the standard version) called a 'Cache Size' option. This option allows you to cache a session's audio files from your hard drive into RAM to alleviate the stress of streaming from your hard drive. This function shines the most when using shared media storage, but it's been found to work wonders when dealing with an older/slower hard drive. Keep in mind though, this setting is really only beneficial to those that have machines with excessive amounts of RAM; otherwise, the 'normal' setting is best!—J.S.

In mixing sound for picture, we use many tools—but none so important as volume. To be sure, other tools play an important role; panning, equalization, dynamics control, and reverb come quickly to mind, but look at any mixing board and the first thing you see are faders for actively and easily controlling volume. Whether you are setting fixed levels, actively riding tracks in real time, or creating intricate custom shaped fades, there is no substitute for having your fingers on the faders. I hope to shine some light on key aspects of volume in day to day mixing for picture. I will discuss some tools of the trade, levels in creating a good mix, volume choices in different types of programs, like drama vs. documentary, creative choices, like nuance and transitions, overall volume levels and requirements, and common pitfalls with volume.

Metering
If you are going to get into some serious mixing, and you really want to be precise with volume, then good metering is key. I happen to like the Dorrough meters, but there are other meters just as good. I think it’s important to have meters that display both VU and digital peak. The Dorrough meters display both at the same time by having a solidly lit portion of the LED for VU and a higher floating single light for digital peak. I am mostly interested in watching the VU levels for all of the 5.1 channels of my main 5.1 out, but it’s also important to watch the peaks in louder moments in the mix. I am looking for dialog and mono FX to be driving the Center channel, ambience, stereo FX, and music to be filling out the front Left Right and surround Left Right channels (weighted towards the front), and bass heavy FX and music driving the LFE channel. I am watching the meters constantly to confirm that the levels are what I want. It is also a great way to see if there are problems like incorrect panning or inactive channels.

Calibration
Another key issue is speaker volume calibration. In order to be able to trust your monitoring environment, you will want to make sure that all of your speaker channels are calibrated. This can be quite involved, but the basic concept is to have pink noise feeding each channel of your 5.1 outs. The pink noise should be set to 0 VU on each channel's output meter. Then using a fixed position of your master volume knob and a hand held decibel meter, adjust each speaker to 85db (or 82db for smaller rooms). The point is that each speaker should be outputting equal volume when sent the same pink noise level for all channels.

Know Your Plan
I generally have a plan of attack when starting a mix. Since most of the sound for picture mixing in my world involves dialog, that’s where I start. I am looking for my average dialog levels to be hitting just below or just up to 0 VU on my center channel meter. I always mix with a set master volume level of 82db, so I know what that level feels like in my room, but the meters confirm it. If the program has VO/Narration I set that level just a slight bit higher than the average dialog level by 1 or 2 db. I then use the dialog as a reference for all the other elements in the mix. I set ambient levels next, then add the FX, Foley, and music. Once I have all the elements in the mix and have set a decent music level by riding the faders, I go back to the dialog and use the faders to make “pushes”. That’s what I call small boosts to the dialog where the music is causing the levels to be a bit too close for comfort. Dialog is king and needs to be clear in the mix.

Managing the Mix
Different types of programs call for different approaches. When I am mixing a drama I allow the dialog levels to span a big range in volume. I believe in dynamic mixes for dramas, meaning I want the dialog to be quiet when the characters are speaking quietly, and loud when the characters are being loud. I want it to feel very realistic in that way. Now, you will have to cheat a bit here or there. Sometimes too soft or too loud is just not going to work well, but I usually try and cheat the levels as little as possible in those cases. I take a different approach when mixing documentaries. I am much more consistent in the dialog level throughout the film. I also let the dialog average consistently hit 0 VU as opposed to being slightly under for a drama. So, the dialog mix is a lot less dynamic overall. There may still be times when I let things be a little quieter or a little louder depending on the content, but not by much. If the documentary is music heavy then I often mix the music in right after the dialog and then follow with FX. I do this because many documentaries are mostly dialog and music, so I want to set those levels first and get them to feel right. Once they are working well together, then I can see what the FX will need to be set at to be present but not overpowering. Trailers and commercials take the concept of loud and consistent dialog to a higher level. I generally mix the dialog for these types of programs a few db above 0 VU, and I use more aggressive compression and limiting to create a loud and consistent dialog volume across the spot. I then mix music, then FX, and I am usually pretty aggressive with all the elements. In most cases, there is not much room for dynamics in these types of programs.

Nuance
Creating great mixes is not just about the same level choices all the time. Using nuance and being very conscious of transitions between moments and scenes is key. I think a really good mix has an artful feel. Knowing when to feature certain FX, ambience, or music elements at specific times can make all the difference. I might make Foley very up front in one scene and very buried in the next, and the same goes for other elements. In the end, it’s about story telling and about creating a feeling for each moment. It’s equally important for the mix to flow from moment to moment. How scenes transition from one to another has a lot to do with how the mix will be perceived. I am a fan of mostly hard cuts between scenes, but there is a time when cross fades will create the effect that you want. I often make small volume adjustments to the elements at the end or beginning of a scene to create a smoother transition. If you are creating nuance then you are probably playing with subtlety. There is a fine line between subtle and inaudible. Remember the type of systems and environments that most people will be listening on. Mix things low if you want but be clear about what you are wiling to lose in the outside world and what you must hear. Last but not least, don’t be afraid to go big. When you want to have that maximum impact either in dialog, FX or Music, the top of the meter is fair game. For some types of mixes you may have to conform to a max level below 0 db, but until then hit the top when you need to. If you want a sound to be really big in volume you will probably want to build it in layers and have those layers hit all of your 5.1 channels.

Normalization
I just mentioned going big, but sometimes you will have to conform to either a max level and or hit a “Dialog Norm” number for your mix to meet delivery requirements. This is most common in the case of TV broadcast specifications that often vary between broadcasters. For a long time the accepted digital video level spec was simply test tone at -20 and Peak levels at -10. In this case 0 VU is set to equal -20db on a digital meter, and the dialog levels are supposed to exist around 0 VU but no peaks are to exceed -10db. There are still some people that ask for this spec, but what has become more recently the standard is to make your program meet a “Dialog Normalization” level number. This requires that you use a special LEQ meter that senses dialog levels and can indicate both a real time “Dial Norm” number as well as an “infinite” number. The infinite number is the one that you will need to conform to. The typical specification is for either -24 or -27db (+ or – 2db). Meaning that you should be at -24 or -27 on your infinite reading, but there is an acceptable tolerance of 2db up or down from that. This is generally a good system to control loudness. The problem with the older -20/-10 spec was that most experienced mixers would run the dialog well above the -20 mark and use hard limiting to keep it from going over the -10 peak. The result was that most mixes were running hot with limiting keeping them legal. The dialog was loud and the dynamic range was nearly nonexistent. The concept with Dial Norm is that since programs are measured by the dialog loudness as the main factor, then the mixes can be more dynamic since peak loudness is not the main determiner. The thing that really annoys me is when a broadcaster’s spec asks for a loud Dial Norm number like -24 and then also for peak level at -10. This completely defeats the purpose of having a Dial Norm level at all, because it makes for a very squashed mix with next to no dynamic range.

For the Web
There are a lot of web spots and Webisodes being produced these days, and it can be a good source of work for people in audio post. I generally mix web content with a relatively small dynamic range using a fair amount of compression and a moderate amount of peak limiting. The key is to “normalize” your mix after your done. This will make sure that your loudest parts go all the way up to digital zero. We usually do this in one of two ways. We either highlight the finished Left Right mix file and then select “Normalize” from the Audio Suite drop down menu and render the new boosted file, or we use the Waves L1 or L2 limiters threshold function to boost the program level until it hits the top at 0. Either will work to create a mix that is pushed right up to 0, and that works best for web audio.

The TV Mix
Another case for specific volume considerations is when making a “TV Mix” for a project that was first mixed in a more dynamic theatrical manner. The concept behind a TV Mix is simple: create a less dynamic mix that will not have people at home grabbing for their remotes to adjust volume while they watch. While in some cases, just doing a mastering pass to the mix with additional compression and limiting and also conforming to a broadcast level spec will give you something that can work on TV, the best method to create a TV Mix is more involved. First you will need to “stem out” your mix into separate groups of sound that can be individually controlled. To do this we create an elaborate bussing structure in our mix template. By creating several sub group aux faders and correlating record tracks for each one of those, we can simultaneously record several “Stems” at the same time. In addition to the Full 5.1 Mix and LTRT Full Mix we can also separately record Dia, VO/ADR, FX, Amb, Foley, and Music. If you then load all of these stems back into a ProTools session you can remix them to create the new TV mix. When we do the TV Mix, we monitor on small speakers (often through a TV) at a low volume. The process is to use all of the separate stems and to adjust volume up as well as down to create a mix with a more even listening level. This really is different from the compression and limiting approach as it allows you to boost low levels of specific elements like ambience, Foley, or Dialog—as well as pull down the volume of loud FX sequences or music. It is not uncommon to spend one or two full days in creating this mix.

The Pitfalls
There are a few overall volume pitfalls that you should try to avoid. Probably the biggest is to not mix elements too quietly. As mixers, we usually sit in an acoustically designed space, relatively noise free, surrounded by speakers that provide high quality full range sound. The rest of the world is not like that. Elements that sound soft and subtle in your mix room may not even be audible in some playback scenarios. The elements that are most commonly affected are ambience, Foley, and music. You can be subtle but be careful not to be too subtle. Another thing I try and do is to be careful not to make the opening sequence of the film too loud. The venues that will be screening your film will usually use the first few minutes of the program to set an overall listening level. If you start too big there is a pretty good chance that they will set your overall level too low. Also, look out for all forms of distortion, as distortion in a mix is not professional (unless it’s used in a special effect kind of way). Plug Ins will often overload and cause distortion. I am not overly afraid of a few red lights here and there, but trust your ears. If it sounds distorted find out what it is and fix it. Speaking of red lights, having them in your mix session on your plug Ins or individual tracks is one thing, but having red lights on your mix stems that you output is not a good thing. Even if you have no distortion accompanying those red lights, people downstream from you, who are laying back your files or doing Quality Control on a delivery, will probably be on your case. Using limiters on all of your subgroup and main mix outs with a setting of -1 db should keep the red lights off your final outputs.

As you can see, volume is a broad topic, but an incredibly important one in the scheme of sound mixing. I hope some of these insights will help in your mixing practices.—Tom Efinger

Is it time to upgrade your Pro Tools TDM system? We had several TDM systems at Dig It that were in need of an upgrade, and the big decision was whether to go with the HDX upgrade path or HD Native.

Since we are a professional facility, I was initially only considering the HDX upgrade. I won’t go into much detail about the various systems except to say that the HDX system replaces the older TDM DSP cards with new HDX cards that are significantly more powerful, whereas the HD Native system uses the host CPU’s internal processing power. Of course one of the big differences is cost. The HDX upgrade is a whole lot more money. The big question that we were having was can the HD Native system run our big movie mix sessions? This was the crux of the matter because the performance of the two systems is pretty much the same. The HDX system has lower latency and can be expanded to be a much more powerful system, running up to 768 tracks, but we were confident that the 256 tracks available in the HD Native would be sufficient for our needs.

Our biggest concern was could the HD Native system handle all of the Plug In DSP that we were going to need in our sessions? We usually run about 160 to 200 tracks for our big mix sessions and run up to 3 Plug Ins on a lot of those tracks. We also have an extensive sub group section in our mix template, and we run 5.1 Plug Ins on those sub group masters, which can really tax the DSP.

Since HD Native runs on the host CPU’s power, we knew that we would probably want a very powerful Mac Pro tower. After one of my freelance sound editors told me that he had loaded some large sessions onto an HD Native system and was amazed at how easily it had handled them, I decided to give it a go. We purchased a recently released Mac Pro 12 core on Apple’s refurb page for about $3200 and traded in one of our TDM systems towards the new HD Native system. We made the decision to get the HD Native Card to interface with our existing 192/IO and Sync IO interfaces. This is optional as you can use any Pro Tools compatible third party interface.

We hooked it all up and installed the Pro Tools HD10 software, and we were in business. The performance of the new system was astounding. We set the max CPU usage to 85%. I am currently mixing a horror film and, we just passed the 200 track mark when we imported all of the composer’s mix stems. Even with 200 tracks and countless plug Ins, all running in a very complex 5.1 session, we are using only about 32% of the CPU usage. So, we still have plenty of DSP power to spare.

All in all we are extremely pleased with HD Native running on a 12 core Mac Pro. From what we can tell it would run very well on any decent Quad core or 8 core machine. Make sure you have a good amount of RAM installed as that is important as well. It’s great to see that this kind of performance is available in such a reasonably priced system.—T.E.