DAVID KRONEMYER: I recently spent an afternoon with a colleague assisting with a complicated audio mix in Logic 8. A lot of quadruple-tracked guitars going off everywhere, that sort of thing. We decided early on that it would be necessary to establish a very strict timing reference for the beats, if the song was to cohere properly. This laudable objective, however, lead to some notable problems.
For starters, don’t rely on any of the timings in the pre-fab drum kits. I speak in particular of the “Acoustic Drum Kit” found in Apple Loops. Yes, I agree, it is astonishing that they (purport to) adjust their internal structure to the selected tempo. In this respect, they are a successor to the phenomenal – but-not-exactly-practical-to-use Roland Variphrase VP-9000 processor, or at least they use a similar class of algorithm. However, the internal beats are notably off, and it’s necessary to quantize them manually, which is a tedious task. Every single beat must be manually edited, due to subtle yet pervasive tempo drift. I have seen tighter timings out of the EMU modules from the early-2000s, circa that kind of weird transition phase when modules had digital outs, but before computer-assisted editing. Even then, although the EMU drums we ended up using were spot on, we decided to move them forward by 1/240th of a beat, they sounded better in the track “ever so slightly” ahead of the guitars. That’s more my issue than Logic’s, though, so I can’t cite it as a fault.
Logic 7 is a miracle, and Logic 8 even more so. It was not that long ago I was palpitating 1/4” tape with a single-edged razor blade. However, the miraculous status of these programs does not prevent me from having several quibbles with them (I will confine my remarks to Logic 8), which I now will set forth. I am not an “expert” with Logic 8, so if I’m mistaken as to any of these points, or I cannot manipulate the program adequately due to stupidity or lack of knowledge, I apologize in advance.
1. The faders don’t adjust to anything but pre-set values. The program makes it seem as though you can click on the controller, and sure enough, it comes up with a number, which you then can erase and type in your own value. However, when you navigate away from the controller, it simply pops back to the pre-set value, not the one you inserted.
2. Logic 8’s “multiple take” feature is a joke. There simply is no way to keep track of what’s going on. If you don’t put your takes on separate tracks, you’re screwed. I know you can unpack the take folder, but it’s so easy simply to propagate another track, that’s the way to do it from the get-go.
3. I started off believing it was a mistake to “overlap” audio regions, however, lately I have become more comfortable with this feature. The presentation of overlapped regions, however, is counter-intuitive. When you click on a region, it shows that region as the one doing the overlapping of its next-door neighbors – even though in fact it may be the subordinate region. As a result, it frequently looks as though audio regions are overlapped the wrong way. However, appearances are deceiving, and somehow they manage to figure out what to do.
4. This leads to another topic of interest, which is the “search zero crossings” feature. Evidently, Apple has devised some kind of algorithm that automatically selects (or at least purports to) the best spot to abut adjoining audio regions. It simply doesn’t work, and you shouldn’t rely on it doing so. Rather, it’s necessary to explode the view towards the furthest granular level possible, and edit the splice between audio regions manually. Furthermore, it’s a mistake to “search for zero crossings” to begin with. The “zero crossing” is the spot where the upward shape of the slope of the wave crosses “theoretical zero” and intersects with the downward slope of an immediately-adjacent wave. You don’t want this point, because it will result in a mis-match of audio files. Rather, you want a space where there is an absolute zero, that is, a flat line where no waves are crossing. This invariably is the best spot for a click-free, context-appropriate splice.
5. Then there is the curious feature called “strip silence.” I understand the objective – primarily, to eliminate breath and mouth noises from vocal takes. We decided to play around with it, to see if it worked. Unfortunately, we conducted our first experiment after we painstakingly had edited a track to accomplish the same result. Much to our surprise, “strip silence” restored a lot of what we had deleted, and even deleted some of what we had retained. We adjusted various db settings, but there was poor fit, in that none of them eliminated the undesirable noises, without also deleting necessary consonant lead-ins, trail-offs, and what not. Once again, the only way to do this reliably is by inspecting individual audio regions and editing them appropriately.
6. When you select a track in Logic 8, it also “automatically” selects all of the audio regions comprising that track, even if you want nothing to do with them. As a result, you may blithely proceed to edit the particular audio region that is of interest, inadvertently altering the internal structure of all of the other audio regions in that track, as well. This is a ludicrous feature of Logic 8 (I believe it also was part of Logic 7) that has no business being there. When you want to edit a single audio region, you don’t want to edit all of the other ones in the track, as well. Whoever thought of this should have their head examined, and Apple should and must change it immediately.
The only thing to do by way of remedy, insofar as I can discern, is to undertake the following three steps, in quick but precise order. First, after selecting the track, immediately convert all audio regions into “new” audio regions. This will prevent them from thinking they are clones of each other. Second, “deselect” everything. This will eliminate the “group edit” problem. Third, save the entire file (on principle). Then, proceed apace to edit the specific region of interest.
I know there’s a way to change this in “Preferences,” but this should be the way it starts, not the other way around.
7. Another unwanted feature is the way Logic 8 handles track automation. One merrily edits track automation for an audio region. What’s the first thing you want to do afterwards? Play it back, either solo or in the context of the track. Astonishingly, though, Logic 8 kicks the track out of automation mode – even though the “automation” button is in “read” mode (a) on the “disclosure” strip; (b) on the track strip; and (c) in the mixer window. You then click on the automation window – which illuminates as though the track is in “read” mode – only to find that it mysteriously has reverted to “off.” As a result, you have to address one or some combination of these three views manually, in order to get the track to automate, which is where you wanted it, to begin with.
We played around with selecting the entire track in order to get the automation to read, which sometimes works – though then you activate another undesirable feature (see above), which is that you inadvertently end up editing all of the audio regions comprising the track. I don’t know if this simply is a quirk of the way track automation operates, or was thought by someone to be a desirable feature for the program. It isn’t. It shouldn’t be so hard. We were, however, reasonably successful in hand-drawing various automation adjustments.
8. It’s impossible to discern how the automatic cross-fade feature works. In order to accomplish the desired result, we finally figured out how to use cross-fade overlap. It requires considerable experimentation to get a track to cross-fade properly, because the way it displays is counter-intuitive. The shaded cross-fade area does not properly overlap the new part of the track that you’re cross-fading into. Rather, it appears much earlier – while still in that part of the track you’re trying to cross-fade from. It actually is simpler to cut the overlapping region, as far in as the audio you wish to cross-fade from. Who would have thought it was this tedious. Even then, the cross-fades always are shorter (temporally) than you might have thought, so after you listen to the track, you’re probably going to want to adjust them to be longer.
9. Thanks to the forward march of progress, there now are some extraordinary software instrument plug-ins available. They sound like nothing else in this world, and I mean that as a complement, because the biggest mistake you can make is thinking they sound like some analog device, which they’re supposed to be emulating. They do not. They sound like themselves, not like what they might purport to represent. To me, the whole “substitution phenomenon” is akin to thinking that “video” somehow can be tweaked to substitute for “film,” or drum machines somehow can be fiddled with so they sound like “real” drums, rather than drum machines. This is a fool’s errand, and it’s best just to accept everything for what it is, rather than pretend it’s something else, or like something else.
Having said this, we experimented with several plug-ins, including: Hartmann Neuron VS; Waldorf PPG Wave 2.V; Camel Audio CA5000; G-Force String Machine; and East-West “Fab Four.” Here are the results.
a. We had Logic 7 on a MacBook Pro. Hartmann Neuron and Waldorf PPG loaded perfectly and had no problems. Neither of them loaded on Logic 8. You can run the plug-in “validation” program until you are blue in the face, and futz around all day moving files between folders, but it just doesn’t happen. Reciprocally, the “Fab Four” would not load at all in Logic 7, but readily revealed itself in all of its 1960s glory in Logic 8. Camel Audio and G-Force, on the other hand, worked like champs in both Logic 7 Logic 8; kudos to both of them. By the way, I looked in the various user groups, and nobody seems to have a work-around for these issues. In fact, nobody even has mentioned them, insofar as I can discern.
b. Hartmann Neuron VS is such a peculiar (but amazing) program that we definitely wanted to use some of its sounds. So this prompted an odd work-around. We had Logic 8 on a MacBook Pro. “No problem,” I said, “we’ll just use an interface to get the audio out of the PowerBook G4 into the MacBook Pro.” Easier said than done! I recently saw an ad for a teeny-tiny little box made by Henry Engineering Co. that takes audio out of a computer via USB and then converts the data stream into either AES/EBU or SPDIF. “Perfect,” I thought, and, furthermore, available through several on-line retailers. Unfortunately, it costs $519, which is a ludicrous price for such a device. For that, you can get a used Apogee “Mini-Me,” which does the same thing with more features and better, too. We looked in the “box that holds all of the used computer equipment that we’re not quite sure what it does anymore,” and found an Edirol UA-1D. “Perfect,” I thought (again). It’s a USB audio interface that takes the digital out of the computer and converts it to either SPDIF or optical. Unfortunately, the SPDIF feature was non compos mentis, if you know what I mean. The optical worked fine, however, its output was very distorted. I know all about drivers and gain structures, so I’m reasonably confident it wasn’t me. Of course, what do you expect for around $40. Money that was wasted, as far as I’m concerned.
c. So we ditched the Edirol and just took the analog out from the headphone jack on the PowerBook G4, then ran it into an A-D converter (an Apogee Ensemble), to get it into the MacBook Pro. This sounded bee-u-tee-full, even though it was ontologically inefficient, seeing as how it involved an extra A-D conversion when, properly understood, this was superfluous. Of course there is another little problem with the PowerBook G4, which is it no longer recognizes its internal speakers – a phenomenon that seems to stump the “Mac Genius” folks, if on-line chatter is any clue.
d. After these various band-aids and half-assed compromises, we were able to output the audio from the PowerBook G4, using Logic 7, into the MacBook Pro, running Logic 8. One could not, under the circumstances, ask for a more perspicuous result. Everything sounded beautiful. I would argue, however, that this is a paradoxical outcome.
In conclusion, I offer the following thoughts on the discipline of audio engineering. A lot of it (producing) isn’t knowing what sounds “good,” rather, it’s knowing what sounds “bad.” A lot of things will sound OK in the context of a song, but when you really sit down and listen to them, they sound pretty atrocious. You need to be able to identify those instances, and then be able to take proactive steps to correct them. I guess this power of discernment is what separates the wheat from the chaff, at least in that world. In a way, I pity a lot of the people who want to get into the audio engineering business. If you think you’re going to be working with rock stars, arranging mikes, dialing in compressors and equalizers, etc., forget it. Mainly what you’re gonna be doing all day is sitting in front of a computer screen, moving audio regions around, like any other geek.
And, as long as I’m on the subject of sound recording tips, I have two others to offer.
First, if you play the 12-string guitar, get used to tuning it without fail between takes using a strobe tuner. If you deviate from this procedure, you will be out of tune, without fail. Even if you follow it, you probably will be out of tune. In fact, you may have to tune to specific chords, and then comp the takes.
The reason why is because, as an instrument, they are incredibly difficult to intonate properly, as any good luthier will tell you. Unfortunately, this particularly is true of the industry-standard Rickenbacker 12, which may be one of the reasons why Roger McGuinn always played slightly distorted, and with a flangy- or chorusy- effect. To conceal the intonation issues. Not his fault, he’s the gold standard of the genre – but more of the instrument’s. McGuinn is a musical hero to me – the cultural equivalent of Davy Crockett when I was growing up. He set the style and defined what was “cool.” I can remember going the junior high school carrying a 45 rpm of “Turn, Turn, Turn” tucked into one of my books (this was before the era of back-packs).
McGuinn used finger-picks, but at the same time he somehow managed to hold a regular flat-pick. I have tried this, and find it almost impossible to pull off. I play much better with finger-picks and then also a thumb-pick. Heavy gauge, all around – the lighter-gauge ones tend to curl over. I remember Fred Walecki at Westwood Music was amazed when I walked in one day and ordered 100 of each.
McGuinn has incredible manual dexterity. It takes a lot of skill to pluck all of those notes, fret them correctly, and then do it on the beat – keep in mind there are four 16th notes in each quarter note, which is what you typically pick in 4/4 time. With Logic, if you want to take the time, you can edit each 1/16th note so the transitions between notes are just about perfect. This is helpful for second-rate players like me, who can’t hit McGuinn’s stride.
McGuinn also applied just the right kind of finger pressure to the notes on the neck. This too is difficult to accomplish, because the Rickenbacker neck is so thin. For me, at least, some of the industrial-strength 12-strings, like the Shergold, are much easier to play. They have a wider neck, which makes them easier to fret. They have different pick-up sounds, which makes them sonically interesting. They intonate better. When I used to play in clubs, I always had several 12’s with me. I’d change instruments each song, and have somebody re-tune them backstage. If I’m not mistaken, most of the guitars on records by The Church are Shergolds. I also am partial to some of the mammoth Vox 12-strings, which sound terrific in juxtaposition to one of their 6-strings – 12-string panned hard left, 6-string panned hard right. Schecter Guitar Research once made a peculiar 10-string instrument, which omitted the second string on the E and A strings. I could wax rhapsodically about the nuances of different 12-strings all day, so please excuse the digression.
Another ingredient to McGuinn’s sound is compression. Compression makes finger-picking come alive, because it modifies each note’s envelope, boosts frequencies, smoothes them out, and adds ring, chime and sustain. Over my playing career, I have investigated just about every compressor known to man. Studio compressors don’t sound that great with guitars. Guitar compressors sound best. Two I particularly like are the “Janglebox” and the Dan Armstrong Orange Crusher, which is a remake of the original Dan Armstrong Orange Squeezer. Early MXR compressors sound great. There is a peculiar device called the Sustainiac. It’s not a particularly good compressor, but adds extraordinary color and tonality between the compressor and the amp, i.e., it goes in the middle of the circuit. All you need for a good 12-string sound, though, is a properly intonated instrument; the right kind of strings; the right kind of picks; a good compressor; a clear amp; and the technique of Roger McGuinn. I guess that’s kind of a lot to ask for.
Second, it’s really hard to sing in tune. I recently heard Neko Case singing “Hold On, Hold On,” which is a brilliant song, and she sings it beautifully, with near-perfect intonation. Unfortunately the production sucks, which may or may not be her issue. It’s too reverby and boomy. It reminded me a little of the Tegan & Sara song “Walking with a Ghost” – both most excellent songs. They’re better singers than I’ll ever be.