1/99 - updated 10/8/07
|back to tuck's corner|
Patti uses B&K microphones (now manufactured by DPA), both live and in the studio. She has done extensive comparisons and always comes back to the B&Ks because of their extreme transparency.
In the studio, since 1985, she has used a matched pair of B&K 4003s, small-capsule omnidirectional mics requiring 130V power. Originally we used the B&K 130V power supply without a mic preamp, but in 1994 (Learning How To Fly) we began using a Millennia preamp, which has 130V power built in. Since Learning How To Fly she has sung in a small, double-walled, double-floored, double-ceilinged, audiologist's booth custom made by Industrial Acoustics, isolated from its concrete foundation using rubber donuts. Its walls and ceiling are relatively standard: Panels four inches in thickness consisting of fiberglas bats sandwiched between solid steel plates except for the innermost faces, which are perforated steel, making the absorption closer to that of four inches of fiberglas except in the highest frequencies; the inner and outer panels are decoupled and separated by four inch airspaces. We put up a lot of additional, burlap-covered fiberglas bats to absorb more, moving them around and listening until we all thought it sounded best. It sounds almost anechoic, and is very soundproof. It was easier for us to figure out how to accomplish this than to figure out how to make and soundproof a great sounding room, so we committed to that course years ago, starting out with a single-walled booth in our previous studio. She sings very close to the mics, which have coincident heads and are angled apart about 60 degrees. She also has the most amazing mic control and vocal technique I've ever witnessed. In 2000 she switched to the Stedman ProScreen, which is much more transparent than the pop filters she had used before. You'll hear the difference on our album, Taking The Long Way Home.
Exceptions: Tears of Joy was recorded before we had a vocal booth, so we soundproofed our rented duplex as well as we could, Patti sang into the corner of an L-shaped baffle in the living room with 4 inches of fiberglas from floor to ceiling and quilts over the top, and we recorded between 1 am and 5 am when there was no traffic, discarding takes where the occasional car rumbled by or one of our cats purred or meowed. Dream was recorded similarly at Different Fur, using our same equipment, when we had moved from our duplex but not built another studio.
Live Patti uses a B&K 4011, which she bought as soon as B&K introduced it, their first cardiod mic (probably in 1990). She had been waiting for years for B&K to come out with a directional mic suitable for live use. Often a hypercardiod would be the first choice for more gain before feedback, but the off-axis frequency response of the 4011 is so smooth that we have fewer problems than we ever had with hypercardiod mics from other manufacturers.
My recording guitar is a 1953 Gibson L-5 CES. My road guitar is a 1949 Gibson L-5 CES, stuffed with foam rubber and socks to reduce feedback. Each has a single Bartolini neck pickup (1CTA on road guitar, 1HCX1 on recording guitar, although prior to Learning How To Fly it was also a 1CTA). Neither has a bridge pickup; what looks like a pickup is an empty shell. On the road guitar we built a preamp into that shell; on the recording guitar the preamp is attached to the inner back of the guitar using Velcro. The pickup feeds the onboard discrete (no ICs) buffer preamp, which was removed from a Carrotron foot pedal made in the 70s (long since unavailable). We modified it, increasing the power supply from 9V to 18V (two 9V batteries in series) to give it plenty of headroom without distorting. Its input impedance is 1,000,000 ohms, and its output impedance is 100 ohms. One of the pots on the guitar adjusts the preamp gain between unity and +20 dB. The preamp output feeds another, passive volume pot (always turned all the way up), which is wired to the output jack of the guitar. All other pots and switches are disconnected. Mogami cable was used throughout. Even though they are vintage Gibson L-5s with humbucking pickups, they sound much more like Strats with single coil pickups due to the pickup/preamp combination. To improve shielding, we completely wrapped the pickup of the road guitar with copper foil, attached to the guitar's ground, but with a switch hidden under the pickguard to unground the foil when needed. Then we put black gaffer tape around the foil for cosmetic purposes.
The buffer preamp is amazing. Guitars are invariably impedance mismatched relative to whatever follows them, with a substantial but unnecessary loss of clarity before even going through any circuitry. We all grew up hearing and therefore loving that sound, but I have come to prefer the sound of the more theoretically correct approach. I have not fully tested it yet, but Bill Bartolini designed an onboard buffer preamp using my Carrotron as a reference then built it into a pickup (the TXE), and reports that he surpassed it in all respects, which I don't doubt. This is very unscientific, but I would estimate that about half of my sound is the pickup/preamp, and the other half the extreme EQ we do (discounting the role of the guitar itself and the fingers). I haven't used a guitar amp in almost 30 years, and given that I basically just use one tone with the goal of a crystalline orchestral sound, that has turned out to be a good move.
I always play through a volume pedal. For many years I used Ernie Balls, then they switched to pots whose tapers did not work for me; they acted more like switches than pots. I switched to Boss, then the same thing happened. Between 2002 and 2004, as I nursed my remaining handful of old Boss pedals along, sometimes dismantling and cleaning several each night, I bought and tested every volume pedal on the market. Out of all of them, the only one that came close to working for me was the Goodrich LDR pedal, which happily was almost ideal, and I started using it in late 2004. But much earlier we had already begun a complex design project (now almost finished) for a custom VCA with microprocessor-programmable taper controlled by a Hilton pedal modified to adjust a digital encoder. This will finally solve all the problems I have experienced with pedals: Taper is not smooth, usually to an extreme degree. Taper is subject to change without notice depending on manufacturing run of pots. Taper varies widely depending on the impedance of the guitar and what follows the pedal. Pots are low-quality, become scratchy and cannot be cleaned very successfully. Most pedals make mechanical noise. Active pedals, where the audio signal does not actually pass through a pot, can solve many of those problems, but at the expense of coloring the sound more.
From that point the recording and live signal paths diverge. We have gone through three recording setups and two live sound setups:
Vocal: Currently the Millennia mic preamp feeds both channels of a Lavry Gold A/D converter. We record unprocessed raw tracks into a Sequoia DAW at 24 bit/96k. Prior to 2007 (I Remember You), we used GML A/D converters, recording into a Sonic Solutions DAW at 20 bit/48k. Prior to 1994 (Learning How To Fly) the mics fed the B&K 130V power supply, which fed a DBX-700 digital recorder (an early, wonderful-sounding, proprietary, non-PCM, delta-sigma format which stored two digital tracks on consumer videotape).
Guitar: Currently the volume pedal output goes through an Aphex 124A -10/+4 interface for amplification to line level, then into a second Lavry Gold A/D into Sequoia. Prior to 2007 the guitar went directly into the Aphex 124A, then through a passive volume pedal, into a second GML A/D into Sonic Solutions. Prior to 1994, while using the DBX-700 digital tracks for the vocal mics, we used one Beta Hi-Fi audio track with DBX noise reduction for the guitar, despite both Sony and DBX telling us it would not work (Aphex interface was unneeded). For Reckless Precision and Hymns, Carols and Songs About Snow the guitar was recorded on one digital track of the DBX-700.
In 2006-2007, coinciding with our longtime engineer and patient teacher Howard Johnston selling Different Fur, where we always mixed, we decided to set ourselves up to mix with his help in our own studio, which we originally designed only for recording raw tracks. What follows is my original description (edited later) of how we did it until 2006, followed by an update for 2007 and beyond.
We mixed at Different Fur in San Francisco with Howard Johnston from our Sonic Solutions workstation onto his Sonic Solutions workstation at 20 bit/44.1k (48k on Learning How To Fly, which necessitated a sample rate conversion at mastering), again using our Apogee D/A and GML A/D converters. Prior to Learning How To Fly, we mixed tracks recorded on our DBX-700 to 16 bit/48k (originally DAT; later Sonic Solutions) using Apogee A/D converters.
The vocal tracks went through an Avalon EQ (GML EQ before Learning How To Fly) into the SSL G Series console direct inputs. We've never used compression, only reverb; Howard combined 10-12 different reverb units, from Lexicon 480s and EMT plates on down.
Beginning with Learning How To Fly the raw guitar track has gone through an Avalon compressor, set to virtually never compress; we just like its slight coloration. Then we EQ using multiple channels of Avalon EQ in series (GML EQ and no compressor before Learning How To Fly), with left and right adjusted slightly differently to give a little stereo spread. We use two channels (8 bands) on the mono signal, then split the output to two more channels (4 more bands each) which go to left and right. We use a lot of high end boost (as much as 30 dB or more at 20 khz!) to get the air and harmonics. The curve looks sort of like the Fletcher-Munson curve, only drastically exaggerated in the top two octaves.
The EQ outputs fed the SSL console direct inputs. We've always mixed in some EQ'd Dolby A, encoded but not decoded; we used a lot on Reckless Precision. It adds something we couldn't achieve using EQ, but we've never found words to describe the effect. We've never used compression or other effects, only reverb.
The EQ and reverb variations from song to song are very minor. The changes from album to album are more substantial; we always spend a day or two up front trying to improve on the sonics of the previous album, and I believe we have succeeded on every album. On ballads we wage, and sometimes barely win, an ongoing EQ war when boosting high end (tone vs. hiss coming from the EQ'd buffer preamp noise floor, given the relatively soft attack and sustained chords without fresh attacks to mask guitar hiss).
Exceptions: For Love Warriors Different Fur was booked, so Howard mixed at Savage Studios in San Francisco. We do not have detailed records of signal path for those sessions.
The path on "All This Love" and "Dance With Me" from Paradise Found was completely different: Neumann M-49 into undocumented preamp and Austin Hatchett (my miniature road backup guitar with same electronics as road L-5) through API EQ, recorded analog, mixed at Seedy Underbelly in Minneapolis by Ricky Peterson and Tom Tucker using our Avalon compressor and EQs but API console and different effects. The "All This Love" remix was mixed at Different Fur from ADAT safety copies.
2006 and later: What changed when we started mixing in our studio:
The raw tracks, recorded in Sequoia, are converted back to analog using two Lavry Blue D/As. They go through our Avalon compressor and EQs as before, then are reconverted to digital using two Lavry Gold A/Ds, then digitally into a Yamaha DM-1000 digital console at 24 bit/96k, going analog again only to feed Dolby A and two external reverbs: A Quantec Yardstick (with GML A/D and Apogee D/A) and a Lexicon 960L configured as four independent stereo reverbs. By using such transparent external converters for all critical signals we effectively bypass the sonic disadvantages of mixing digitally and reconverting to analog to go through the Avalon EQs. The mix is recorded back into Sequoia. We wired the studio with Mogami digital cables for audio and digital signals. Dan Lavry generously advised us on the best way to clock everything in our particular setup, saving us from resorting to an external master clock.
For I Remember You, the first CD we mixed at home, I of course had to measure the latency of each signal path and get everything properly aligned. In the process, I accidentally found a set of delay relationships that improved the effect of the Dolby A that had always been the secret ingredient in the guitar sound. Relative to the non-Dolby guitar, the Dolby returns arrive later by: Left: 1.396 ms (134 samples at 96k). Right: 0.458 ms (44 samples at 96k). This was for us the magic combination, as it happened by mistake, then we later tried every combination of all possibilities for both numbers and still liked this pair best. The result was to widen the guitar image substantially, making it feel more like an orchestra surrounding Patti in the middle.
Howard mixed I Remember You while we were still working our way through the learning and debugging curves, including starting from scratch with all new console, routing and reverbs, so we spent much longer than ever before mixing, but it is our best-sounding CD to date, so it all paid off. Ultimately, for just one voice and one guitar we ended up using 45 of the 48 console channels and all of the buses and aux sends!
Since 1991 Meyer HD-1 nearfield monitors have been our trusted reference. When we built our studio in 1994, Mark Johnson at Meyer suggested that we have acoustician Bob Hodas optimize our control room. Bob spent a day with Meyer's SIMM system in our room advising us about placement of speakers, listening point and absorption and adjusting our Meyer CP-10 parametric EQ. The results were unbelievable; it went from unintelligible to a monitoring room that we trusted so much that we quickly stopped feeling any need to listen through a variety of systems. We always took our HD-1s to Different Fur for mixing, then brought them home to check the mixes.
When we upgraded our studio in 2006 for mixing, we had Meyer restore our HD-1s to new operating condition and asked Bob Hodas to refine and retune our room, since it would now become our critical primary monitoring environment, without Different Fur as an additional reference (although spread around the house and wired to the studio we also have KEF 107s, Dynaudio BM5As with sub and a second pair of HD-1s). We also realized that putting additional equipment in the room had changed the acoustics. In a series of visits, Bob subtlely but substantially improved it further, specifying an additional, moveable wall, adding a Bag End Infrasub-12 PRO Subwoofer, crossed over at 30 hz by a Marchand XM6 Crossover, repositioning the speakers and readjusting the CP-10 EQ. The results were once again amazing, perhaps more so this time, given how good it had already sounded. The low end, previously very flat to 32 Hz, was now razor flat to below 15 hz, enabling us to accurately experience and judge subsonics, and everything else became even more flat and coherent than before, with improved image, to the point that it had become a great mixing room. We were all thrilled when we took it to Bernie Grundman for mastering, and found that it sounded just the same in his room as in our studio. In a lifetime of do-it-yourself experimentation we could not have approached what Bob Hodas achieved in a matter of hours. To us there is no question that any room designed for serious listening should be tuned by an expert using the best tool (Meyer SIMM), and Bob would be the first person on that list.
The signal path feeding the monitors is as clean and minimal as possible: Sequoia's mix outputs are routed back into the DM-1000 console (I have proven that this is a bit-accurate process). The console's digital outputs feed a third Lavry Blue D/A converter, and its analog outputs feed the monitoring path using passive attenuation.
We mix both our Future Sonics in-ear monitors and Front of House (FOH) from our onstage equipment. In early 2005 we miniaturized it all into an oversized briefcase (my carryon when we fly, although a little heavy at 46 pounds) that opens flat to become my pedalboard. It contains: BSS Soundweb (a black box with 8 inputs and 8 outputs that you program with a PC; I have configured it to do all the mixing, routing and graphic/parametric EQ we need, as well as to serve as a leveller on my monitor mix to control dynamic range, since I can't get my fingers on a volume control while playing). Lexicon MPX-550 reverb. Aphex Compellor (set primarily for levelling, rather than compressing or limiting, it transparently reduces the dynamic range of our FOH mix by 16 dB and is the reason we sound so present live). Studio Technologies headphone amps modified for extra headroom. Goodrich LDR volume pedal. Boss TU-12H Tuner. Custom power supplies for everything. Box with stereo stepped attenuator for FOH output, signal ground lift switches and selector between FOH mix, TV/radio mix and test tone. Patti's mic and cable. Audio Technica AT849 stereo PZM mic (for us to hear the audience).
Soundweb has eight control voltage inputs; you can connect a pot or switch to one and program it to control anything. So we built a small box onto Patti's mic cable with four miniature remote controls: Headphone volume, audience mic volume in our monitor mix, privacy switch (mutes her in FOH and reverb but not in our monitor mix) and vocal reverb mute switch (for talking). Audio does not pass through any of these; they are just remote controls for Soundweb. In addition there is a headphone jack and an inline headphone mute switch. Her mic cable is four-pair Mogami snake to accomodate the mic signal, headphone signal and all the remote control voltages.
The guitar cable is also four-pair snake, and I have a box attached to my guitar strap with controls that duplicate Patti's so her and my backups could serve for the other person if necessary, but mine is only programmed to control my headphone volume. One cute feature is the Hypertronics miniature 12-pin connectors that lock but can be gracefully connected and disconnected with one hand, which I have to do whenever we go on or off stage.
We could walk into a venue with the pedalboard and my guitar, hook up to the PA and do a show, but we also bring a backup guitar and a big suitcase full of backups, tools, supplies and lots of audio adaptors and problem solvers just in case.
Until early 2005 we accomplished the same functions with a heavy rack whose lid opened up to be my pedalboard. Life is easier now! For this we thank Peter Miller at CAE in San Mateo, CA, a compassionate genius who has done all our tech work for many years and manages to turn my wild-eyed designs into actual things that work reliably. (We seriously considered mixing on a laptop instead, but the possibility of having to reboot during a show ruled it out; our whole system is ready to go five seconds after powering on.)
The in-ear monitors are consistent from night to night; all we have to do is make sure they are working. Likewise the FOH mix, so all of our soundcheck is spent adjusting the FOH sound system, as if we were playing a CD, only it's live. Patti and I alternate, one making sound while the other listens in the house, with both of us instructing (and learning from) the engineer. We serve as good checks and balances on each other's ears and room analyses. We have not traveled with an engineer since 2000, and not that much before then, so most of our experience has been working directly with the house engineer.
A Typical Soundcheck
After first setting the console's input trim using a test tone from Soundweb, we isolate and deal with silent or bad speakers, listen for phase coherency, listen to any existing system EQ curve and decide whether to modify it (sometimes) or flatten it and start over (usually), reposition speakers if needed, restructure gain and/or grounding scheme if noise is audible, readjust crossovers/processors, amp levels and balance among various system zones (left vs. right, stacks vs. center cluster vs. front fills vs. other fills) if needed to get the system sounding as good as possible without EQ, then EQ the system using the appropriate combination of individual system 1/3 octave EQs on each zone and global 1/3 octave EQs inserted into our console channels. (Sometimes, in very complex systems with lots of zones and known to be set up well, at festivals with several groups or in situations where we are not the headliner, we accept the existing system setup in advance, committing to doing all our work on the inserted 1/3 octave EQs.)
Fortunately we seldom have to worry about feedback, because we do not have speakers on stage and Patti's B&K mic is so flat (and the guitar is stuffed). This means that almost all engineering decisions can be made solely on the basis of what makes the music sound best. After almost 30 years of this, we can usually optimize a system quickly and consistently in a soundcheck, making mediocre systems sound good and good systems sound great. Even though we're unusually good at this for musicians, we always try to form a team with the house engineer, taking advantage of their knowledge of the system and the room. By the end of soundcheck we have demonstrated to the house engineer the sound we are after so they can fine-tune the system during the first songs once the audience comes in and changes the acoustics.© 2007 Tuck Andress