This month, the SOS team tame a couple of square rooms and help a guitarist to master his Logic DAW setup.
Mark Archer is a talented guitarist but also has a good ear for recording and mixing. His studio is situated at the bottom of his garden in a wooden outbuilding that he and his son assembled from a kit, and it's used for teaching guitar, rehearsals and recording. So far, Mark has recorded a few bands (including his own), solo artists, and even a barbershop quartet, as well as preparing backing tracks to support his guitar‑tuition business. The day of our visit was at the height of the December snows, but luckily for us, Mark's home was only around an hour's drive away!
Internally, the wooden building is divided into an entrance space, a control room and a live room. As luck would have it, both the control room and live room are very close to being square, approximately three metres a side. From previous experience, we know that square rooms are not conducive to even‑sounding bass, and where the room is fairly small, as these ones are, the mixing seat tends to end up being placed close to the centre of the room, where low‑frequency cancellation can be a particular problem.
The position of the doors and the equipment Mark already had in the studio didn't leave much room for adjustment, as, in addition to his Logic Pro-based Mac laptop setup, he also had three old Alesis ADAT machines hooked up to a large Behringer Eurodesk, plus some additional outboard gear in a tall rack, including a Focusrite preamp and a two‑channel graphic equaliser.
Prior to our visit, Mark had asked about basic internal treatment and I'd advised him on a DIY approach to backing the control room doors with Rockwool (mineral wool) slab absorbers, as well as building a Rockwool‑faced 'cloud' that could be hung over the mixing position to reduce ceiling reflections. These are simple affairs in which the Rockwool, around 50mm thick, is fitted into a frame and then covered in acoustically transparent fabric to keep any stray fibres inside. Mark had built these by the time I arrived, but he'd not got around to covering them, as he wasn't sure what type of fabric would be best. I assured him that pretty much any 'breathable' fabric, where you're able to blow through it, would be OK, so he decided to use some dark material he already had at home. Anything made of cotton will work well, though if it's in a place of work remember that you must pay attention to fire regulations. Camira's Cara fabric seems a popular choice.
Another problem he'd experienced before we arrived was a low‑level digital buzz from his monitors. This turned out to be emanating from his Focusrite Saffire Firewire audio interface, which he'd been running from his Apple MacBook using the Firewire bus to power it. I suggested using the Saffire's mains power supply to see if that helped, and indeed it did: the buzz disappeared. That left some issues with the monitoring environment, a few questions as to how best to use the limited space in the live room and some Logic‑related queries.
Mark appreciated that monitoring in such a small space could affect the low frequencies, so he'd bought a test CD with various test tones and then used his two‑channel, 31‑band graphic equaliser to try to compensate for this. His Wharfedale active monitors are fairly small, with bass drivers of around six inches in diameter, and he had them lying on their sides on his monitor shelf, with just a thin piece of packing foam below each one. I suggested that we play some commercial mixes through the system, using songs he had stored on iTunes to see how well his EQ was working, but Mark said that iTunes always played back through the laptop's internal speakers. This was easily fixed by resetting the Mac's sound preferences so that the Saffire was the default audio output device.
Playing a few tracks revealed the usual dip in bass frequencies in the centre of the room, but it wasn't as bad as expected — presumably because the wooden structure and Mark's suspended cloud provided some natural bass trapping. The bass was, however, not very well defined and the general stereo imaging was pretty unfocused. The overall sound also seemed lacking in definition.
Bypassing the graphic equalisers improved the sound quite significantly, so we set about optimising the speaker positions before resorting to EQ. Fortunately, Silent Peaks had kindly given us a pair of their vibration‑absorbing speaker‑mounting pads, which comprise a dense foam platform topped by a steel plate and a thick layer of hard neoprene. These models were uniform in thickness, which meant that we had no option to tilt the speakers, but by standing the Wharfedales upside down so their tweeters were at the bottom, we managed to get the tweeters at ear height. As usual, these were also angled inwards towards the listener.
This change in orientation, combined with the decoupling effect of the speaker pads, improved the tightness of the low end to a very noticeable degree, and as long as Mark sat close to the mixer, rather than leaning back in his chair, the low end stayed surprisingly consistent.
The Wharfedale speakers had controls for level and a bass‑cut option, but no means of tweaking the high end. After listening to a range of material, we both decided that the speakers were a little aggressive‑sounding around the 3.5kHz region, so as an experiment, I reset the graphic equalisers to be flat other than for a very gentle dip centred around 3.5kHz. The result was subtle, but did help smooth out the sound from the speakers, so Mark decided to try working that way for a while to see how he got on.
To improve the stereo imaging, we needed to put some foam absorbers at the mirror points, which to the right meant moving a marker board, and to the left, hanging a panel over the live‑room window in such a way that it could be removed and redeployed in the live room when tracking. In fact, we decided to make all the panels movable, so we glued old CDs to the rear upper centre of our Auralex 2 x 4 foot foam absorbers, to allow them to be hung up on simple nails or screws.
The final arrangement we arrived at was to have one Auralex panel each side of the mixing position and two more at each end of the back wall. The centre of the back wall was occupied by shelves with a heavy rug hung over them, so that section of the wall was already pretty dead. I suggested that Mark experiment by hanging some heavy, rubber‑backed carpet from the ceiling just in front of the shelves, as this should act like economy barrier mat and absorb some low‑frequency energy.
Two more panels were hung on the plain wooden back wall of the live room, where they could be augmented by the one we hung over the window for mixing. This would be useful when tracking vocals or anything else that might benefit from a drier acoustic.
Mark had already bought himself an SE Reflexion Filter, and achieved some good results using it in conjunction with a duvet behind the singer — but with the three foam sheets intercepting reflections from the wall behind the singer, he could expect a similar result when using his SE 2000A microphone, but without losing as much space. The floor of the live room was carpeted and the ceiling angled, so we anticipated no serious reflection problems from those directions.
Mark then commented that his acoustic guitar recordings sounded better when the instrument was miked near a reflective wooden wall. As with many acoustic instruments, the acoustic guitar does sound more lively in a slightly reflective space, so I suggested he try the trick of placing a hard, reflective board on the floor beneath the guitar and mic, to generate a few floor reflections. A piece of hardboard or MDF works fine for this and can be stored out of the way when not in use.
For electric guitar recording, Mark had been using his Line 6 Pod and Logic's Guitar Amp Pro plug‑in, but having recently bought an Egnater Rebel 30 valve combo, he was keen to try miking that. He played me some of his works in progress, the quality of which was generally excellent.
However, he said that he hadn't quite got to grips with Logic's Space Designer reverb, and so he'd been using some of the less sophisticated algorithmic reverbs that come with the program. It was a simple matter to save a copy of one of his songs, set up a post‑fade send bus to feed a Space Designer and then to tailor one of the plate reverb settings to suit his mix. I usually roll‑off some low end from the reverb (below 200Hz) and also add between 60 and 90ms of pre‑delay.
More problematic was Logic's tendency to report a CPU overload problem every few bars, even though the mix in question had relatively few tracks. It turned out that the heavy CPU usage was due to the use of multiple incidences of Amp Designer Pro, as well as one or two other hungry plug‑ins. I suggested using Logic's Freeze Track function, which renders any designated track as a temporary audio file with all the effects 'burned' on, to free up CPU resources. Most DAWs now include such a feature, and this is particularly useful on less powerful laptops. If you need to make changes, you simply click the Freeze button again, make the changes, then re‑freeze the track. It's a lot cheaper than buying a new machine!
Mark didn't know where to find this feature, as the track Freeze buttons are only visible in Logic if ticked in the View/Configure Track Header menu. Ideally, this should be set up in the default song to save having to do it every time you start a new project. For the same reason, it's also worth setting up a post‑fade send feeding a reverb, as the chances are you'll need at least one reverb, whatever kind of music you're mixing.
Although Mark had a separate monitor screen that he tended to use to show Logic's Mixer page, I also suggested including screen sets in the default song. These would be particularly useful when working only on the laptop, as it would be easy to switch between the Arrange and Mixer views using the number keys. While Logic Pro 9 allows everything to be displayed on one page, it does get rather crowded on a small screen, so switching between traditional full‑screen views is often more user‑friendly.
Mark also confided that he'd used some of the preset channel‑strip settings, which essentially load up a channel with plug‑ins with appropriate settings for a certain style of treatment. These can work OK, but often the EQ nas to be tweaked to suit the actual source — and of course compressor thresholds always need to be adjusted to get the desired amount of gain reduction. Basically any preset involving plug‑ins that have threshold settings, such as compressors, expanders, limiters and gates, will need adjusting to match the level and dynamics of the source material.
Some of Mark's mix channels were also running slightly hot, causing clipping on peaks, and while you can bring down the level of the whole mix by click/dragging a box around the required mixer channels, then pulling one fader down by a few dBs (which makes all the others follow), this doesn't work if any of the channels involve level automation, as that simply overrides the manual fader settings.
Logic does allow the automation on a particular selected track to be scaled back by Command‑dragging the yellow meter bar that shows in the track header when automation is active, but they still haven't come up with a way that lets you scale back all level automation in one go. If anyone at Apple is reading this, maybe Command‑dragging a fader when several channels are selected in the Mixer page could be made to automatically scale back the automation values as well as the manual faders?
Finally, Mark asked about the best way to de‑ess vocals, so we went through Logic's de‑esser plug‑in, which has several useful monitoring modes to make setting up easier. For example, you can listen to the side‑chain signal post‑filtering, to help you tune the filter to the sibilant part of the vocal range, then you can listen to the 'sensitivity' control's effect, so that you hear only the sibilant sounds as they are detected above the threshold. After that, you need only set the amount by which the detected sibilants should be reduced, and in most cases the ess‑filter frequency can be left set to be the same as the detector frequency. Moreover, if the sibilance only becomes obvious once you add reverb, it is possible to de‑ess only the reverb send and leave the dry part of the vocal untreated.
With the snow starting to fall again, we decided that we'd covered enough ground, and that if I was to get home before spring, now would be a good time to start!
Mark Archer: "The changes made to the speaker setup and studio acoustics have given me a clearer stereo image and a far more accurate frequency response, so I can feel confident that what I am mixing is going to sound comparable on other systems. I also picked up a great number of tips from Paul on some of the finer points of using Logic, especially how to remove those unwanted vocal 'ess' sounds using Logic's de‑esser. All in all it was a great day, and I cannot thank Paul and the SOS team enough for what they did.”