Several months after our initial attempt to record the first release for my Soundkeeper Recordings label (documented in the previous entry in this blog, Three Decisions (Part 3) and in Recording in Stereo (Part 2), Art Halperin and his band Work of Art joined me once again deep in the woods of upstate New York, at the 18th century church I’d come to call Large Green.

One of the things I’ve always enjoyed about Art’s music is its changing nature.  Our first project together had me mastering reggae, the second project, a blend of folk and rock.  The music Art brought to the previous session was primarily electric rock.  The new songs he composed in the interval between the sessions were a mix of folk, rock, bluegrass and a jazzy ballad.  (The recording containing electric rock would have to wait for a future project, which would yield the third Soundkeeper Recording.  My feeling was that having the artist follow his muse was by far the greater priority.)

Like his music, the constellation of players making up his ensemble also varies, depending on where his musical vision takes him.  Where the previous date at the church featured a quartet with two electric guitars, electric bass and drums, this time out the ensemble was primarily acoustic in nature.  Art still plugged in for a few of the songs, using the Fender Stratocaster given to him by his friend Eric Clapton, but on most of the tunes the instrumentation consisted of an acoustic guitar, a second acoustic guitar or a banjo, a mandolin, acoustic bass and drums (or percussion) to accompany Art’s vocals and the two harmony vocals.  In addition, some of the songs included a part for lute.  I smiled at the idea of having a lute and a banjo on the same record.

Unlike typical studio productions where each of the multiple microphone or instrument feeds is recorded to its own track on a multitrack recorder and combined later in the “mix” to 2-channel stereo, recording direct-to-stereo in the way I chose begins with no more than two channels.  Among other things, this means that rather than synthesize placement of the players on the stereo stage electronically, as is done in a typical mixdown, placement of all the instruments and voices upon the stereo stage is accomplished physically—in reality—and therefore, has to be decided upon ahead of time.  In other words, anything I want to hear from the left has to be placed on the left before we started recording.  And anything I want to hear from the background has to be placed in the background before we started recording.  Effectively, this means the mix must be done before the recording is made.

The multitrack recordings typically made in studios are usually comprised of multiple monaural captures of the instruments and voices, panned somewhere between left, center and right during the mix to simulate a stereo spread.  This isn’t at all the same as capturing the sounds in real stereo, utilizing the different types of cues our brains use to localize sounds in the real world.

In addition to the placement of voices and instruments on the stereo stage, the balances between individual voices and instruments need to be considered.  Further, the balances between each of the members of the ensemble and the room need to be considered.  Where a typical studio production would control this electronically, my approach to making the record—having the sound of a real room—means the balances, like the placement of the voices and instruments, must be achieved beforehand.  (I remember the day after one of my earliest stereo experiments, done while I was at Atlantic Studios.  The chief engineer was listening to a playback with me and after a short while, turned to me and asked “What did you use to get the room sound?”  He seemed surprised at my response, which was “The room.”)

There is no doubt that removing the convenience of the usual studio approach and the “safety” it provides places increased pressure on the players as well as on the engineer.  Musical performances and the sonic balance must all gel in specific ways for the recording to work.  If one musician makes a mistake, the whole ensemble must play again.  If the engineer doesn’t get the balance or some other aspect of the recording, the performance is missed.  With a pickup this sensitive, other variables come into play as well, such as the fly all the way at the back of the church, whose easily heard buzzing was not in the right key as the last chord of one song faded away.  That little fly made us do another take.  I came to refer to this method as “recording without a net”.

Certainly there are many ways to make a satisfying recording.  I moved toward this approach because it provides things I’ve not found with any other method.  First, the players must perform for real.  They must be at their best as individuals and as an ensemble.  They must listen intently to each other (as well as to how their sound interacts with the space in which they are playing) and they must do so in real time, as the music is occurring—just like a real performance.  The result is a certain “electricity” or frisson in the performance.  Foregoing the safety of the studio, namely being able to “punch in” to fix mistakes and being able to adjust balances later on, is considerably more demanding but provides considerably greater rewards.  In addition to the musical ones, the sound itself, when captured as one coherent stereo entity has an ease, a sense of focus and a sense of Life that I have never experienced from other recording techniques.

After a long day of music making and fun, it was a joy to get back to my studio and hear the first playback.  Finally, we had the makings of what would be the first release.  We still didn’t have it all though.  What we captured was wonderful but Art had more songs that we wanted to record, so we arranged another recording session.  While the equipment used at the recording session was the same as I mentioned in Three Decisions (Part 3), there were a few changes for the next session.  For those interested in such things, the gear was as follows:

Microphones:  Earthworks QTC-1 (aka QTC-40, matched pair)
Mic cables:  Mogami Neglex 2534
Interface:  Metric Halo MIO 2882 (serving as microphone preamps and A-D converters)
Laptop:  Apple PowerBook
Software:  Metric Halo Console X (Record Panel)
Power conditioner:  Monster Cable HTS-400
Vibration isolation:  Custom made base to support laptop and interface

Before the next recording session occurred, I tried a different set of microphone cables.  I knew from previous experience that cables could have a significant effect on system performance and with the microphone cables being the first cables the signals would encounter on the way to the listener, I was curious to hear what the wires from Nordost would do.  As I mentioned in Three Decisions (Part 1), “Of all the cables I listened to, several of the products from Nordost consistently stood out as I brought them back in to compare against others…  With the Nordost cables in the system, I felt confident I was able to hear past them, that they were merely passing the signal from one component to the next without editorializing.”  Once again, these cables impressed, to the point where I felt I was finding out that the Earthworks microphones were even better than I thought they were.  With the new mic cables, there was a sense of diminished grain and increased definition.  The sound was less “bright” but more open and extended, more coherent, solid and natural, more real.

The other piece of hardware that changed before the next recording session was the interface.  I knew Metric Halo was working on a new model and having found so much to like in all of the hardware and software I’d tried from them, I also knew I was going to want to get one.  In preparation, I’d sold the 2882 and as the new model was still off in the future, borrowed a different Metric Halo unit for the recording session.  This model, the ULN-2, utilizes different microphone preamplifiers than the 2882, offering a somewhat different sound.

So, for the next recording session, the following items replaced their counterparts mentioned above:

Mic cables:  Nordost Valkyrja
Interface:  Metric Halo ULN-2 (serving as microphone preamps and A-D converters)

Now that everyone was familiar with the space and with my approach to recording, that next session went very smoothly.  The rest of the songs were recorded and when Art and I listened to the results, from the opening strums of his Martin acoustic guitar to the fade of the church ambience after the last chord in the last song, we knew we had the album we wanted to release.  The feel of the music was there in spades and sonically there was a palpable three-dimensionality in the playback that really gave an uncanny sense of being in the room at Large Green, in the presence of the performers.  This was exactly what I set out to record.  After we deliberated and entertained dozens of possible candidates, Art came up with the name for the album: Lift.  I loved the simplicity and the optimism and thought it perfect for the first release on Soundkeeper Recordings.

From the beginning, it was my intention to offer Lift in several different formats.  Starting with the CD, prior to releasing the album, I spoke with folks at a few dozen different CD replicators.  My experience over the years, having sent CD masters to replication facilities all over the world, is that “pressings” from different plants all sound different from each other and none sounds indistinguishable from the master used to create it.  (This is at complete odds with the claims that it is all “just ones and zeros” and every copy is identical to every other copy.   My experience has been consistent and without exception, regardless of the playback hardware used, since the first CD I compared with its master back in early 1983.)  To be clear, the differences can range from subtle to not subtle at all.  Always, there is a loss of focus and fine detail when compared with the CD master.

At every plant but one, the personnel claimed their product sounds identical to the master.  I thanked each of them and contacted the next plant on my list.  Ultimately I found one where the contact, with no prompting from me whatsoever, said “They’ll never sound identical to the master.”  That plant got the job and in that instance and every other one I’ve sent them in the ensuing years, they have delivered finished CDs which, while still not indistinguishable from the master, are so close that I need a direct, synchronized comparison to discern the remaining losses.  (There is more to say on this subject, which will come a few paragraphs hence.)

Another thing I discovered early on is that a CD-R burned at slow speed directly from the CD master sounds closer to the master than even the finest CD pressings in my experience.  Here the difference is quite subtle but it is there nonetheless.  With this in mind, I decided to offer Soundkeeper customers the option of a burned-to-order CD-R as a slightly closer-to-the-master disc than the pressed CD.

Since Lift was recorded at 24/96 (24-bits and a 96 kHz sample rate), I wanted to offer a third option that would completely surpass the CD and provide the listener with the resolution of the original master.  At the time, the best way to do this was to take advantage of the fact that the DVD-V standard allows for 24/96 audio that is playable in any ordinary DVD machine that plays videos.  (This should not be confused with the subsequent—and now essentially defunct—DVD-Audio standard, which requires a special machine.)  These would also be custom burned-to-order on DVD-R discs.

A few years after Lift was released, computer audio started to take off among audio enthusiasts.  With computer audio, various software applications allow listeners to enjoy music played back from files rather than discs, directly from their computer rather than via a disc player.  At first, computer audio meant music distributed in the sonically compromised mp3 format (aka “eMPty3”) where the lion’s share of the data (along with the high fidelity) was removed in order to shrink file size.  Now, audio enthusiasts were starting to listen to full resolution (and high resolution) files via their computers.

Most interesting to me was that here at last was a no-compromise way to deliver to the listener the sound of the master itself.  Another benefit of computer audio, as I found, is that when different CD pressings are properly “ripped” (i.e., copied) to the computer’s drive, the sonic differences between them disappear.  Further, once played from the computer, the differences between these (as heard via a CD player or transport) and the master from which they were made also disappear.  Possibly the subject for a future entry in this blog but the short version is that while playback from disc, even with the finest CD players or transport/DAC combinations does not sound indistinguishable from the master used to create said disc, playback from the computer does.  At long last, the listener at home can have the sound of the master.

This prompted me to add two additional release formats.  I chose to stay with raw PCM formats such as .aif (my preferred format, in which I do all my engineering work) and .wav.  Many of the online services offering high resolution downloads have gone with so-called “lossless” formats such as .flac.  While these can result in audio which to my ears is very close to the original, there is a long, long way in my view, from “very close” to “indistinguishable” and one of my prime reasons for starting Soundkeeper was to release no-compromise recordings.  To this end, additional burn-to-order options in the form of 24/96 .aif files-on-disc and 24/96 .wav files-on-disc were added.

Each of the custom burned formats is shipped with the same printed artwork as the pressed CD.  I thought it would be a nice touch if the first several in each of the custom burned formats also was signed by the artist, so starting with Lift, that is what we have done with each new release.

The Lift page on the Soundkeeper Recordings Web site contains more information about the album, including samples from all the tracks, lyrics, quotes from reviews of the album, photos from the recording sessions and a link to an interview with Art Halperin.

The last part of the picture I had in mind for the new label involves the business end and the relationship with our artists.  First, where most recording contracts involve the label taking ownership of the publishing, I decided I wanted the composer of the music to retain 100% ownership of it.  Next, instead of getting pennies per sale, I wanted the artist to receive a very significant percentage of every sale.  Not only is the percentage considerably larger than what the biggest acts get from the major labels, the percentage increases with the number of sales.  I very much wanted Soundkeeper Recordings to treat its artists like the gems they are.  After all, they give us the gift of the music.

I am forever indebted to Art as well as all the Soundkeeper artists, who not only present us with the music but who have presented me with the opportunity to make the kind of records I’ve always wanted to make.  Art my dear friend, you helped me realize a dream.