Studio SOS: Mixes That Don't Travel

Another reader's studio gets the benefit of expert SOS staff attention. This month, it's the turn of Tim Way, whose mixes sound fine in his own studio, but don't travel well.

Dorset-based Tim Way got in touch with SOS recently to ask for help with the sound of mixes produced in his studio. The problem was that they lacked sparkle and seemed slightly woolly at the bass end when played back over other people's systems, even though they seemed OK in his studio. Technical Editor Hugh Robjohns and I loaded up my car with a pair of Mackie HR624 monitors I'd had in for review, some test equipment, and a selection of recordings that we felt we knew the sound of, and set off down to Tim's place to sort out his monitoring problems.

Tim Way in his home studio. The acoustic tiles you can see behind him were left from a previous studio setup, where the sitting position had been facing that wall. However, with the newer equipment configuration these tiles were causing an imbalance in the stereo image.The studio, which Tim uses with his songwriting partner Steve Shepherd, is located in a converted bedroom, where a significant part of the room has been walled off to create a large vocal/instrument booth. The resulting L shaped space is a little cramped, and after initially setting up the room working lengthwise (the speakers on one of the narrow walls), Tim decided to move things around. The main impetus was that he felt the stereo image to be too narrow and uninspiring when he was composing and mixing.

Turning the room around allowed the monitors to be set up either side of a large keyboard/outboard stand, more or less in the corners of the long wall. This worked out better than expected, despite the proximity of the vocal booth's glass window directly behind the mixing position, possibly because the speakers had to be spaced further apart than ideal. Tim decided to mount the speakers horizontally on their stands, and because they were so far apart, he had to angle them in more steeply than usual. This, combined with the narrower treble dispersion when speakers are placed horizontally, seems to have resulted in the glass window causing less of a reflection problem than expected. Unfortunately, he'd also used foam acoustic tiles covering most of the original end wall, so now these were on his left with a more or less bare reflective wall to his right — not good for symmetry.

Though this didn't seem to be causing any obvious problems, an impromptu experiment with a duvet and two wooden battens to deaden the live wall indicated that the stereo imaging would be improved a little if half of the original tiles were removed from the left-hand wall and relocated to the right, to restore some acoustic symmetry. As the glass control room window wasn't causing any serious problems, we suggested that Tim might think about fitting a wooden slatted blind which could be left half open when mixing, to help diffuse any reflections.

Because of the wide equipment rack at which Tim worked, his speaker stands were spaced further apart than would be ideal.The first main thing to do was listen to some of Tim's mixes and check these against some of our test recordings. Our reference material quickly revealed that in this particular room, and with Tim's choice of hi-fi amp, the Spirit Absolute passive monitors were producing very little true bass (despite being in corner positions) and they also sounded quite aggressive in the upper mid-range. Because of this, Tim was tending to mix with more bass than necessary to compensate and he was also backing off on the bright sounds, which is why his mixes sounded dull when played elsewhere. Furthermore, because the monitoring wasn't adequately accurate, Tim was using quite a lot of EQ to try to get sounds to fit, and in some cases this turned out to be counter-productive.

As an experiment, we set up the Mackie HR624 active monitors we'd brought with us — I have HR824s in my own studio, so I'm familiar with the general character of these speakers. We expected the sound to be different in some respects, but we were surprised at exactly how much difference there was. Suddenly there seemed to be about another octave of bass, and all the upper harshness disappeared. Considering the unplanned acoustics of the room and its currently asymmetrical setup, what we were getting back from the speakers sounded surprisingly accurate on our known recordings. That isn't to direct too much blame at the Spirits, because the performance of any passive monitor can vary enormously depending on the type and rating of amplifier used to drive it. The Mackie speakers also cost considerably more than the Spirits, and monitoring is generally a case of 'you get what you pay for'. My own feeling is that Tim's hi-fi amp wasn't really powerful enough for the job, in which case a lack of bass and high frequency roughness is what you'd expect. At a later stage we tried the Spirits the right way up and found the sweet spot widened significantly (as expected, because of the greater horizontal dispersion), but the tonal problems we'd experienced originally were still there.

Into The Mix

Following the visit, Tim decided to upgrade his monitoring, and bought himself a pair of Mackie HR624s.For the rest of the session, we reverted to the Mackie HR624s. Now that we had a monitoring system we could trust to be reasonably accurate in Tim's room, we set about analysing a mix he was working on and came up with a few suggestions. The first step was to strip out the EQ on his tracks to see what the source material was like. Even with the new monitoring, the bass end of his mix wasn't working too well, so we concentrated mainly on the kick drum sounds and the bass instruments, in this case an electric bass guitar and a bass synth arpeggio. Everything was recorded into Pro Tools LE, so we were able to use plug-ins for testing out our ideas on EQ and dynamics. Any MIDI instruments not recorded as audio were also driven from Pro Tools LE, which meant we also had the opportunity to experiment with different sounds as necessary.

First off, we felt that the kick drum wasn't working. It was a sample that Tim liked because its attack had an acoustic quality, but it didn't have much punch and, as the track had a very dance-style rhythm section, he'd tried to address the sonic deficiency with EQ. The problem was that piling on enough bass lift to bolster the LF energy had made the sound quite woolly, so we suggested he might be better off using a combination of two samples — one with the attack quality he liked, and another with a very deep sound, such as a TR909 kick.

We trawled his keyboards and sample CDs for a suitable kick, but surprisingly few of them were solid enough. In the end, we found one that also included (unwanted) vinyl noise that we felt would work well enough just to prove a point. Layering this with the original kick Tim had chosen produced a much deeper, more contemporary sound, and instead of having to use a lot of EQ, we found we could create a wide range of tonality simply by adjusting the balance of the two samples. I messed around with a gate trying to reduce the amount of vinyl noise following the sample, but this wouldn't have been necessary if a more appropriate sound had been found in the first place.

Moving to the bass instruments, we tackled these one at a time. working on the bass guitar first, as this seemed to be getting lost in the mix once everything was playing. As is so often the case, the audible part of the bass sound was somewhere in the lower mid-range, so peaking up the EQ at 367Hz gave us a sound we could hear as well as feel. Some very deep bass was rolled off, as the DI'd bass track had too much sub-70Hz energy, and then, instead of using normal compression to even up the sound, we used the Waves L1 Ultramaximizer plug-in to limit the peaks quite heavily, with between 5 and 10dB of gain reduction showing up. This dramatically improved the audibility and presence of the bass guitar, allowing it to carry the track while still sounding tight and controlled.

In order to make a more symmetrical stereo image, Hugh and Tim cleared some space to the right of the workstation and hung up a large duvet to reduce reflections from the untreated wall.When we brought in the bass arpeggio, it conflicted with the bass guitar rather too much, so we applied 18dB of low cut using a Waves Renaissance EQ plug-in (high-pass mode, shelving frequency 980Hz, Q 0.71). This stripped all the low end out of the sound but still retained the musicality and the rhythmic element of the arpeggio, and these aspects of the part seemed to work nicely alongside the bass guitar and rhythm track.

To add a little more sparkle, without cluttering the track, we brought up the level of the hi-hat loop (a rather processed sound from a sample library) and again limited this, using another L1 plug-in, to firm it up a little. To help clean up the mid-range, we then filtered some low end out of a string pad so that it wouldn't muddle the mid range. Tim had already identified an irritating edge in the string pad, which he'd notched out using EQ, so the low cut was all we really needed to do. The string pad sounded extremely wide, and when Hugh checked for mono compatibility we discovered a great deal of the high sparkle disappeared from the strings in mono. Reducing the stereo width slightly helped to create a more consistent sound, although the effect could not be removed completely, as it was an integral element of the effect processing of the string pad. If you are hoping for radio play it is important to check mono compatibility, as there are still a very large number of mono radios out there!

A little gentle 'air EQ' was applied to the vocal track, which was quite heavily processed already. This entailed using a Waves Q4 equaliser to add around 1.5dB of presence boost at 3.5kHz and 7dB of 'air' boost at 13kHz. These settings definitely provided better definition and clarity for the vocal part, but we had to take care not to process it so much as to exaggerate the natural sibilance of the voice. The vocal was already compressed, so we didn't need to change that.

Final Touches

After getting a rough mix together, we made a test CD-R and played it in the car stereo, to get a feel for the overall tonal balance, and found that we'd erred slightly on the side of being too generous with the bass. However, bass level problems are common in small rooms, so this came as no surprise. A little further experimentation with levels redressed the problem, which is why it's essential always to get used to your monitoring system by playing known commercial recordings over it whenever possible.

So, as always, the real problem did not have a single overriding cause, but was due to a combination of smaller factors. Certainly the chosen monitoring system wasn't producing accurate results in that particular room, but the problem was compounded by the fact that Tim often used large amounts of EQ to try to bend sounds into shape, rather than finding sounds that worked properly in the first place. In other respects, the songs were very well written and arranged, and it was all I could do to stop Hugh singing the track we'd been working on during the long drive home!