The Recording process of our live ‘Hipnautical’ duet gig

The Recording process of our live ‘Hipnautical’ duet gig

Although my new camera is a ‘pro’ version with XLR audio inputs for balanced line input & Pro mics with phantom power, I’m just using the camera’s built in mics with it’s own nifty pre-amp with a handy compressor limiter; this works out real nice for that ‘audience perspective audio’…and to use as an audio reference which is by nature in sync with the video; this helps syncing the video up to the individually recorded audio tracks from our digital mixing board.

Since returning from years of sailing the world’s ocean on our sailboat Hipnautical we had to upgrade or replace quite a lot of the equipment that we sailed with for over 20,000 miles at sea.
Our new mixing board has class A mic pre-amps for audiophile studio quality, it also has very useful compressor, limiter, expander/Gates & EQ for each input to the board. Having these basic but critical sound tools to treat the sound
BEFORE it is recorded is a feature most mixing boards don’t come with built in. In the good ‘ol days when recording to tape, like the real high quality 2” 24 track machines that cost $40,000 to $100,000, the tape itself acted like a natural compressor when ‘hit hot & loud’. Good tape machines would capture the pure signal with all of it’s volume dynamics up to it’s fully rated level. When better tape became available the ‘signal to noise’ ratio was improved by being able to record hotter & louder signals before there was a diminishing return or ‘tape compression’. Drums were especially happy with this arrangement, by recording certain drums a bit too hot to tape gave them a bit of natural sounding compression, which allowed for even louder ‘in your face’ drum sounds.

The ‘in your face’ slammin’ drum sound concept it basically this, the over all average sound levels or how loud something generally sounds is of course an average of the loudest & quietest parts of the audio signal. Since the initial transient peak of a stick slamming down on a drum creates signal amplitude many magnitudes higher than the average perceived sound level, so if you could put a cap on how high a transient peak can go, you now have all that ‘head room’ to increase the perceived volume of the main body of sound, thus within the same peak volume levels you now have a big fat main body of sound that fills the dynamic range with a louder average volume.

When digital recording came around, ‘old school analog tape recording engineers quickly found that recording digitally removed that ‘fat in your face slammin’ drum sounds so soon there was a bit of a scramble to use compressors & limiters BEFORE the signal was to be recorded; since most audio engineers didn’t bother to use much compression to drums when recording to analog tape, the digital recorders revealed that you can only record as loud as the highest peak signal & once you hit that it was a glass ceiling & distortion was guaranteed. In order to get that ‘fat drum sound’, compression is used to gently hold back the main body of the signal while a limiter would catch the transient peak/spikes that would otherwise ruin the recording with a nasty hard line clip :-O

With the good ‘ol analog boards you had to use pesky cables to send & return a signal to outboard ‘sound tools’, which did the trick but this could quickly add miles of wiring & if you had use sub standard un-balanced connectors you would have to contend with noisy connections & loss of signal quality. Even lots of new ‘pro’ mixing boards lack these basic but vital sound tools & surprisingly enough most don’t have a way to
RECORD their digital audio if it ‘s ‘just a live mixing board’.

Luckily last year a clever audio company got the clue & put a great little live mixing board together that also sends all of the inputs to your favorite Digital Audio Workstation or DAW via FireWire. When most live mixers pro or otherwise don’t even have this option & those that do often just use a USB connection which is extremely limiting with respect to the amount of tracks you can record & the bit depth & sample rate.

The mic & instrument inputs are dialed in with all the sound tools and levels are brought up as high as possible without using too much compression, limiting & of course
NO PEAKING, dialing in these to match the characteristics of the particular instrument or voice but generally with less than 3db of compression limiting for normal range signals & less than 6db on the real loud stuff. Once these are good to go they are delivered painlessly via FireWire to the laptop & recorded in the DAW while the same signal is routed thru the mixing board where it can get panned, faded up or down, add reverb & delay or chorus... muted or soloed…none of these live sound operations effect the signal going to the DAW. Keep in mind that for the most part it is not necessary to strive for hitting the compression & limiter hard or even the less than 3db average & less than 6db peak range but rather to achieve a reasonable strong recording level.

So our camera captures the video & the reference audio from it’s internal mics & the DAW capturing each direct instrument & vocal mic signal to individual tracks we now have the elements for postproduction sessions. The first step is to import the video & it’s reference audio into our video editing software program, storing all those gigabytes on a nice big hard drive. Once the high res video has been transferred it needs to be rendered into a format the DAW program can use, I’m using the most popular ‘industry standard’ DAW which likes the QuickTime format. Somehow after converting about 25 gigs of 1920 x 1080 video & the reference audio to a much lower 720 res, it becomes about 32 gigs, but now I can import this video file into our DAW session.

Now with all the elements in place on our DAW, the first step is to sync up the tracks of audio recorded thru the digital mixing board to the reference audio recorded by the video camera, this is simply done by locating obvious peaks & dragging the clips to align up with these ‘obvious transients’. ‘Obvious transients’ for example are the 1,2,3 count in with a big loud sound on the ‘down beat’; for songs that sort of just meander in with no discernable peaks you’ll just have to find an ‘identifiable’ peak somewhere, anywhere within the song will be fine & the sync will stay ‘frame accurate’ the whole song.

Once the audio tracks are in sync with the video the individual audio tracks can now be mixed to picture. At this point my first objective is to take each track & use my best DAW sound tool plug-ins to make the track sound as good as possible. Typically starting with a good EQ, I seek out the most offending frequencies & pull them down ‘to taste’ & similarly I seek out the most pleasant sounding frequencies & boost them up ‘to taste’. On an already reasonably well recorded track these EQ’ing boosts or cuts are typically in the less than 3db range, or course the worse the quality of the recording the more drastic these EQ ranges will be.

If it sounds ‘befuddlingly’ simple like
when a master sculptor is asked how he creates a magnificent sculpture of a horse, he replies…” I just remove everything that doesn’t look like a horse”….Well, that’s a bit true & the master sculptor with a life time of obsessed practice can make it seem that simple & so will it be with the budding new audio-phyte with practice ;-)

While using these sound tools to improve the tone with EQ, control their levels with compression, limiting & Maximizing…compare these results not only with the track soloed but in context with the general mix; keeping in mind that as the song gets closer to the final mix the tone & levels of the other tracks may create a profound need to re-tone a track in order to allow it to be heard well & distinguished from the other sounds, this often is compensated by adding a bit more high mids & highs. Harmony vocals require more highs around 4 to 10K & to remove some lows below around 80hz. The highs give backing vocal tracks their clarity without excess volume, which is controlled with cleaver use of compression, limiting & Maximizing.

By only having two instruments in our duet we get a fair bit of clarity in distinguishing the two main sounds of the harp & the acoustic guitar so I can go for a more ‘well rounded tone’ from each without having to add more high end presents to a similar track if it was used in a full band mix.

A very handy tool is the multi-band compressor as compared with a standard single band compressor. A classic problem with a single band compressor is if used over the entire mix with a loud driving bass note or kick drum you can easily hear the highs pumping in volume each time the kick drum triggers the compressor, by using a multi-band compressor you can effectively compress only those frequencies you choose to & in this way you control the amount of compression the kick or bass notes get independently from the highs using just a 2 band compressor.

The large Celtic lever harps like Bobbie Jo’s 39 string Woldsong have a large tonal range so the same challenge is presented here. You don’t want the high ranges being compress by the low range triggering the compressor.

I tend to first set a starting EQ to get an accurate feel for what frequencies sound good or less than good…then I bypass the EQ I then setup a 4-band compressor & limiter. It’s important to place the compressor first in the insert chain, if the EQ is first before an active compressor, the more you boost the EQ the more you compress….
the EQ will just push into compression, like a dog chasing it’s tail.

There can be certain frequencies the multi-band compressor can handle well, for instance if a vocal tracks get to much low end breath by singing too loud & way up close to the mic, just create a low frequency range that deals with those unwanted occurrences so when the singer has gotten too close to the mic the compressor keeps those levels in check, other wise it does not effect the rest of the track.

Again if a vocal track has a bit of sibilance issues to deal with from HISSSSSING Ss… you can create a frequency band for that offending sibilance & apply compression pretty much only when the sibilance occurs. Luckily we don’t have too much sibilance issues, the few that do pop up I handle with volume automation during the mix down.

So back to the harp with it’s wide tonal range & the using a 4-band compressor. Each dominant frequency range can now be adjusted to gently be compressed independently which results in a very musical sounding tool which helps add gain & presents to the harp within a full mix. The multi-band compressor also has multi-band limiting which helps to gain control over the transient peaks that would otherwise limit the overall volume of a track and it does this in a very musical smooth manner as compared with a single band limiter.

….but use the 4-band compressor/limiter reasonably, ie less than 3db average & 6db peak gain reduction….Leave some room for a gain ‘Maximizer’ tool. This type of plug in uses advanced processing to analyze the signal & create a restructured gain mapping using Intelligent Release Control to dramatically lift the perceived audio level with the lowest psychoacoustically perceivable artifacts. There is a huge difference in the quality of musical audio or any audio for that matter when you compare using a single band compressor & limiter to attempt to achieve what a gain Maximizer does. A compressor or limiter does not intelligently release the gain reduction after each transient, it is set to a manually controlled ‘release time’. In real life, the dynamics of real time sound coming from instruments thru the air in a live room creates a sound that we perceive as vividly real, for lack of a better term. A compressor or limiter with it’s non intelligent release time pre set will render the hapless piece of audio less ‘dynamically life like’, while a Maximizer, set on it’s highest quality processing setting will result in surprisingly life like control of the transient peaks, along with the main body of sound giving you that bold ‘in your face’ sound that was born in part with the analog tape recorders.

There are a number stages in which the depth of processing is rendered, with each stage the processing becomes a very big burden to even a brand new MacBook Pro, so once the track is fully toned up, checked & rechecked with the rest of the mix these processing hog plug ins can be rendered to the track.

At this point I should give this advise:
The moment you start a session & have just imported all your production tracks, create a copy of these ‘un-molested’ tracks & then ‘hide & make them inactive’. If down the line you really screw up any of the tracks with bad edits or really bad processing, you can always restore the original; here again make another copy of this origanl track before you have
your way with it ;-)

Our shows have three sets of music or up to about 24 songs, minus the miss-takes or failed video camera batteries; still there are a heap of songs to mix. With applying all of these plug ins to each track & using the automation to blend the tracks nicely an easy 3 hours can go by in a big hurry. After a number of songs I realized that, for the most part, each track on each song generally used the same tonal processing; slight adjustments to the compressor & Maximizer gain controls but just slight. By the time I mixed our 2cd AV live recording I came up with even more ‘dialed in’ presets for each track and rendered the plug ins on each track across every song. In other words I dialed in the guitar & then applied that preset to the guitar track in all the songs in one ‘rendering’. This saves heaps of time compared with processing each track one song at a time, which only will work out if the tracks are consistently similar. If I changed types of guitars on different songs it would then be necessary to create individual processing to each type of guitar. I go for a real natural all organic acoustic sound, no chorus, flanging, delays & not like an electric that may be clean, overdriven, distorted, crunchy…So this is way I can come up with a nice compression, limiting, harmonic enhancement & EQ to use for the acoustic guitar though out the whole show’s mix. The same goes for the harp & for the most part our vocal tracks as well.

When we play an instrumental song I’ll point my vocal mic at the guitar to get more of the body tone to blend in with bridge pick up built into the guitar. My acoustic guitar has a built in pre-amp along with a processor that simulates how this guitar would sound in the studio with a great mic. They took this guitar into a great studio & used a few great mics to capture this signal into a computer, which then analyzes the characteristics & qualities of this signal. This data is used to create an algorithm program that will convert the bridge pickup signal to a signal that has the characteristics of the studio mic; it’s the guitars own sound, it’s just modified to capture the ‘air’ & ‘depth of field’ that only a good mic set an optimum distance from the guitar can produce. I use this ‘Virtual Imaging’ processor give my guitar the sound of a studio mic on the guitar even when just using the bridge pickup; adding a real mic only makes it better.

After creating custom processing for each track, similar to what is done when mastering an important CD for media release, but instead of mastering a two-track mix, I master each track with the forethought of how I’d like the track to sound when I make the final mix down to two tracks. This gives me even more control of the final mix. I’ve come across a lot of audio mixers that rely & depend the mastering engineer to do all these ‘miraculous’ processing on the all ready mixed two tracks. There they desperately attempt to add compression to the entire mix & often rely on their favorite ‘boutique’ single band compressor of iconic legendary repute like an all tube Fairchild, LA-2A…These are famously great compressors, but are only single band, you will get poor results like when the lows trigger the compressor, all the frequencies will pump at once. These ‘all star’ compressors truly are great, but they would work much better if used on the individual tracks & not the final two-track mix.

Yes of course I use compression, limiting & gain control maximization on the main left right buss, but since I’ve carefully processed each track in this manner, the multi-band compression & limiting is very subtle, there just to deliver my final mix at precisely the levels I’m shooting for with no over driven clipping. Along with mixing each instrument & mic track, I will determine if the ‘audience’ sounds from the camera’s built in mics are worth blending in with my mastered audio, sometimes the audience is not paying attention & keep yakking which is useless, but for the times they are ‘with the program’ by contributing to song set up or story that accompanies some of our sailing adventure songs or clapping enthusiastically I blend this stereo track in as well but fade the ‘sub standard quality’ track during the music portion.

Once the final audio mix is done I import the mastered audio mix into my HD video editing software. This is where I do all those photo shop like processes as needed to optimize the video image, usually compensating for poor lighting & contrast challenges. I’ll through in the obligatory video fade ins & outs & use a bit of digital zoom & panning…Yes…Good God Man this is not how a ‘Video-phile’ would do it, it’s much better to hire a camera operator to do optical zooms & real pans during the shoot so the picture resolution is not degraded by the digital zoom & pan process…but we can’t always get what we want….but if we try sometimes….
We get what we need… ;-)