Cracking the Code Impulse Response Pack
Record the speaker cabinet sounds of Cracking the Code with the Cracking the Code IR Pack!
Includes all the pine cabinet H30 impulses — plus the Marshall 1960A impulses we use for rhythm tracks, and a couple ultra-rare Cornford 2×12 impulses featuring the Vintage 30 speaker recorded via the Shure SM57.
Amid the seemingly illimitable interest in amplifiers, with their mysterious tone crafting abilities, and their steam-punk innards so attractively aglow with filaments and glass, it’s easy to overlook the humble wooden box at the end of the chain. But if your 10-second Chevelle is only as fast as the slicks that hook it to the tarmac, then your supercharged big block of an amp only sounds as good as the device actually emitting those pressure waves into the room: the speaker cabinet.
Indeed, the speaker and cabinet assembly is a good part of any classic tone. That sparkle on the first Van Halen album? Reportedly the work of the glittery JBL D-120 speaker. But with a nearly endless variety of speaker cabinets available in all shapes, sizes, and power handling capabilities, why build your own? As is often the case in the world of extreme guitar sports, the answer is simple: because we can.
The plan was simple enough. I liked the sound of my Cornford 2×12, which has plenty of sparkle. I also own some exceptionally well made birch plywood cabinets, fortress-like in both their construction and bass-rich sonics. Could some combination of the two attributes — both sonic and physical — produce a more balanced all-rounder?
The Cornford cabinet is actually pine — unusual, but not unprecedented. Most of Fender’s classic combo amps were pine, and very much a part of the chain responsible for the classic Fender chime. Pine is a soft wood, prone to denting, and possibly not as roadworthy a choice if your guitars often find themselves packed in flight cases. But this cabinet will be a studio denizen. And built of 1″ stock and braced internally, pine can be plenty solid. The result of this experiment, as you can see in the video above, is a slightly oversized 2×12 cabinet, at 29 inches wide, to fit most head boxes, but a full two feet tall and 14″ deep. This is a shade deeper than the Cornford, three inches taller — and almost double the depth of its beefy birch comrade.
Speaker of the House
Once I completed the cabinet build, I immediately turned my attention to speakers. I had already acquired a small menagerie of greatest hits from Celestion — the Marshall of speaker manufacturers, in terms of sheer ubiquity — in anticipation of undertaking some type of exhaustive Consumer Reports-like roundup of their various permutations in all my cabinets. But making these acquisitions was an education unto itself.
To begin with, the nomenclature is impenetrable: G12H-30? G12M? G12T-75? Whatever happened to Corvette and Mustang? Oh right, Celestion is European. XK-E or 300SL, anyone? Anyway, early rock guitar players must have had a tough time with this as well, since various nicknames quickly arose, such as “Greenbacks”, to describe the green plastic surrounds on the Celestion G12M speakers that shipped in Marshall speaker cabinets.
After a few rounds soldering and de-soldering in the sisyphean crusade of auditioning every possible combination of speaker and cabinet in my possession, I quickly came to understand a couple things. Firstly, I’d be old and gray[er] before I finished.
And second: Man, guitar speakers sound weird.
What’s the Frequency?
As anyone who has ever attempted to rig up a guitar cabinet as a home stereo quickly discovers, guitar speakers are not full-range speakers. A whole chunk of the upper part of the audio spectrum — most dramatically above about 5000hz — is simply gone. What this does to the sound of your Billy Joel records when they are played through it during parties, you crazy teenagers, is not euphonious.
But it gets weirder. The frequencies that do exist below the continental shelf are no by means uniformly represented. You can see this very clearly with that most useful of studio tools: a clean amplifier. I use a Tech 21 Landmark 300 bass amp for this purpose, whose power amp is simply a straight-line amplifier hitched to Tech 21’s famous tube-emulation SansAmp circuit. In other words, jacking in directly to the power amp input, and bypassing the Landmark’s bass amp emulation circuit entirely, gets you a perfectly good clean power amp suitable for simply making things louder with little extra distortion or other sonic degradation.
Meet Your Match
Piping some white noise out of Logic into the Tech 21, connected in turn to the cabinet of your choosing, and recording the result through your usual recording chain, will yield a kind of sonic topographical map of your speaker and mic combination. To invert this sonic Cynthia-Plaster-Caster-esque creation, and see what your recording chain’s, er, capabilities actually look like, you can employ an indispensable diagnostic feature of Logic: Match EQ. Using Logic’s Match EQ plugin to compare the pure white noise with the copy of itself filtered through the tortuous echo chamber of your speaker cabinet, produces the graph below.
That craggy iceberg of a curve is what a Celestion G12H-30 speaker installed in the newly minted pine box, and recorded with a Shure SM57 microphone, actually looks like. In other words, the yellow areas of the graph represent the part of the white noise input signal that the speaker and cabinet could not effectively reproduce. Chipping away at this big block of hiss, almost like Michelangelo attacking his marble substrate, the speaker and cabinet produced the very distinctly guitaristic curve we see here.
Now consider for a moment that this particular speaker, dubbed the “H30” in guitarist parlance, is actually considered somewhat flatter than the usual guitar speaker in the upper mids — the jagged coastline on the graph between 1khz and 5khz — and you’ll begin to appreciate how strange this relationship of guitar, amp, and speaker has come to be.
Mic vs. Cab
As a point of comparison, take, for example, the Beyerdynamic M160 microphone. Slung from the rafters in the foyer at Headley Grange, the M160 captured those cavernous drums on Led Zeppelin’s “When The Levee Breaks”, and came to typify the Bonham sound. The mic is a ribbon mic, known for both is subtle rolloff of high frequencies, and also for its upper mid boost — that same craggy area on the H30 graph — that lends an aggressive edge to brighter instruments like cymbals.
When I was first setting up my studio, knowing essentially nothing of audio engineering, but knowing well the legend of Bonham’s Levee sound, I picked up two of these mics with the intent of installing them as overheads — the two microphones charged, in Bonzo fashion, with capturing the stereo image of your drum kit from above. This is a role they still occupy today, and many hundreds of recordings later, I’ve become quite familiar with their sound.
Taking a look at the actual frequency response chart provided by Beyerdynamic reveals precisely the two M160 sonic fingerprints: the slight plateau from 2000hz to 8000hz, and the gradual decline thereabove. But what’s instructive is to read the Y axis of that chart. That storied plateau? About 2db. That high-frequency rolloff? About 4 db, give or take. You’ll also note the general smoothness of the curve itself. That plateau rises and falls over a span of about 10,000hz. Even the comparatively aggressive high-end rolloff so typical of ribbon microphones slopes over a good swath of the spectrum. Gentle curves like this are typical of most studio equipment.
Now compare this to the absolutely ludicrous, meth-addicted scrawl that is the H30’s frequency response. It swings wildly by 12db to create deep notches smack in the middle of the spectrum. It drops calamitously above 5khz by 30db or more. Compared to the Beyer microphone, which is a moderately colored microphone, and itself not ruler-flat in its response, you can see how positively strange the frequency response of an otherwise typical guitar speaker really is. And it’s not just the H30. Most guitar speakers, broadly speaking, exhibit the same stalagmite shape, producing significant response only between about 100hz and 5000hz, and packed with bizarre seismographic peaks and valleys in-between.
But the tallest peaks and deepest valleys in this are actually the work of the cabinet. By enclosing the speakers in a tiny reflective chamber, phase interference occurs as sound waves intersect inside the box — reinforcing some frequencies, and canceling out others — and enhancing the irregularity of the speaker’s response. And comparing the Match EQ plot to the official frequency response graph for the H30 provided by Celestion themselves, we see that some of the extreme notchiness of the Match EQ plot is indeed gone. The region from 100hz until about 1000hz is much smoother on the Celestion chart. And this is because the speaker was likely tested outside a cabinet, in an anechoic, or echo-free chamber. But the area above 1000hz is still a frequency response freakshow of the sort that would be positively alien in a microphone, headphone, or studio monitor speaker. So it’s really a complex dance between speaker and cabinet, as we’d intuited all along by our interest in building one.
All of which is why playing back your top 40 hits through a guitar speaker cabinet won’t sound at all the same as listening through a set of full-range stereo speakers or headphones. It’s also why, after testing a half dozen of these bizarrely named and strangely notchy creatures in a handful of my cabinets, including the new one, I quickly came to the conclusion that I had no idea what I was listening for. While notable differences had begun to emerge between the handful I’d selected — the H30s had upper mids, the 75s did not — it’s not clear what the point of those observations was supposed to be.
Every speaker you can imagine probably sounds good for some musical purpose, and testing them without any particular song or mix in mind is kind of like evaluating paint without a painting. Do you like blue? How about indigo? Uh, yes.
Taking the Lead
Absent any immediate recording objective, and not subjectively moved to pick one or the other, I grabbed the two speakers I happened to buy in a pair — the H30s — and installed them. I’d constructed the rear panel with bolts instead of screws, in anticipation of a bustling docket of audio engineering experiments requiring frequent speaker removal, but for now I was happy enough to get back to playing that I sealed it up like Tutankhamun’s tomb and put away the drill.
As it happens, this last minute double-fisting was fortuitous. The combination of the pine box’s natural liveliness and the H30’s natural upper midrange boost gave lead tones extra presence, while the cabinet’s oversize dimensions added bass response that kept things from getting too harsh. Soon I was getting lead tones, like this one, that I was pretty happy with.
Recording rhythm guitars was a different matter. Riffs were attractively clear, but through the H30s, they also telegraphed like lead parts. This was fine in compositional situations when the riffs themselves were intended be the centerpiece. But as I began writing and recording songs for Cracking the Code, I spent a lot of time fussing with EQ and levels to get rhythm parts to lay back out of the way of lead lines and voiceover. Scooping out chunks of midrange to make way for other musical parts is often what mixing entails, particularly when you’re given someone else’s tracks to mix where the future interlocking of various parts hadn’t been fully considered at the time of recording. But when you’re recording your own material, and the cabinet is sitting right next to you, I began to think that this kind of surgery shouldn’t be compulsory.
Smoke and Mirrors
In considering these mixing challenges while recording the Cracking the Code theme song, Arcade, I remembered an old Marshall 1960A cabinet we had babysat in the studio for a while. The cabinet was now gone, but through the miracle of recording technology it lived on inside my computer in spectral form. And this bears some explaining.
Since Les Paul produced the first stunning experiments in multi-track recording, with multiple copies of him and his partner, Mary Ford, improbably harmonizing themselves on songs like “How High the Moon”, the evolution of studio technology has been marked by a dialectic of innovation and abuse. Most of the comparatively recent developments, like Vocalign, for synchronizing the timing of multiple vocal tracks, and the canned samples used to steroidally enhance drum recordings, aren’t well known outside the recording industry.
But some, like Auto-Tune for correcting vocal pitch, have broken through in pop culture argot, like Jello and Band-Aids, as a metonymy for the increasing commoditization of musical creativity. Infamously employed by the aptly named hip-hop artist T-Pain, the intentional overuse of Auto-Tune to produce mechanized, synth-like vocal textures in pop recordings has shone a necessary light on the increasingly manicured portrayal of human music abilities in pop recordings.
Of these innovations, however, there is one advance that even to the most cynical reactionary offers nearly unvarnished upside: impulse responses. Through a mathematical process known as convolution, sonic fingerprints of real ambient spaces can be captured and applied independently to studio recordings. The cavernous reverb of St. John the Divine? The legendary clarity of Avatar’s Studio A? All available at the press of button provided the proper impulse response.
The process is straightforward: First, a test tone is recorded in the ambient space — typically a sine wave sweep across the entire audible range, from single-digit herz up to around 20khz. Then, through a mathematical process known as convolution, that recording is converted into a type of sonic fingerprint, or impulse response, readable by any number of popularly available reverb plugins. By simply loading the impulse response into the plugin, a non-reverberant, or dry studio recording can instantly assume the character of recordings in the original impulse environment.
But there’s more. Impulse responses aren’t simply useful for recreating the time domain characteristics of reverberant spaces, but also the frequency response of the entire recording chain utilized to create the impulse response. An impulse captured in a dark-sounding environment will impart that rolled-off high end to the target recording. Similarly, an impulse recorded with a dark-sounding microphone will impart the frequency response of that microphone, in concert with that of the room, to the target track.
It didn’t take long for enterprising guitarists to put two and two together: By recording impulses through particular guitar cabinet and microphone combinations, those impulses could then be applied to neutrally-recorded guitar recordings after the fact. These directly injected, or so-called DI guitar tracks, recorded by taking a direct feed from the amplifier with no speaker cabinet or microphone in the chain, could then be fed through plugins running the speaker impulses. Except this time, instead of recreating the lush reverb of a gothic cathedral or rarefied concert hall, the plugin would simulate the effect of the given speaker and mic combination — long after the original guitar recording had been completed.
The advantages of this space-age superpower are numerous. Imagine loading up an impulse response of the exact dual-mic (if legend is to be believed) setup on Van Halen’s first album: one mic against a vintage greenback, another against the shimmering JBL D-120. Blend to taste. Better yet, have Don Landee, who recorded that album with Ted Templeman, dial in the exact blend they used and bake that into the impulse itself. A one-impulse ticket to Brown Sound. Just add Plexi and Variac.
Other possibilities are completely pedestrian but no less convenient. Loud band in the studio next door? It won’t matter, since you’ll be recording DI, with no microphone at all.
And perhaps most importantly, microphone positioning, part of the black art of the recording engineer, is one of the most fungible components of the entire recording process. Even knowing the exact microphone, cabinet, and amp settings used, it can often be impossible to recreate the exact sound of a particular take at a later date. A position difference of even half an inch left or right of the speaker’s centerline can create a notable difference in the treble frequency response of the recorded signal — to say nothing of the myriad possible variations when a microphone is placed on an angle, or off-axis, to the speaker cabinet.
On the other hand, recording an impulse response of the mic position for a given chain effectively freezes that combination in amber. Future recordings, even those performed in entirely different studios, can sound nearly identical to the original tracks when performed via DI and run through the same impulse responses.
Sound and Soundalike
And therein lies the bottom line: all the whiz-bang innovation in impulse response recording would mean nothing if the results weren’t convincing. The good news, is that, as much as probably can be in the subjective world of audio recording, they really are. As an example, the following riff was recorded through the Cornford Hellcat, through the pine 2×12 cabinet, stocked with H30 speakers, and miked with a ’57 about 1.75″ inches left of the speaker’s center. Tone settings on the Cat were 6, 6, 4. Here’s the live microphone track:
No eq or compression were added to the above take — just a touch of reverb. And then, without touching mic, I impulsed this exact setup. I already have impulses of this combination of cab, speaker, mic, and mic position, but for the purposes of demonstration, and to elimination any possible variation in mic position, I ran the impulse of this exact configuration without moving a thing. Here’s the impulse track:
Accounting for playing differences — this was not a reamp — I honestly cannot tell the difference between these two takes. While there are certainly articulation differences that would differentiate the performances — it’s almost impossible to play muted bass notes identically between takes, for example — I can hear no practical distinction between the frequency profiles of the two recordings.
To be fair, after a couple decades of loud stereos, guitar amps, and drums, my hearing above 10khz is not what it once was. My cousin once amply demonstrated this by — unbeknownst to me — playing a 16khz tone via his iPhone at shrieking volume only a couple inches from my ear. (I didn’t flinch. Oh well. There goes my mixing career.) As a sanity check against my encroaching aural frailty, let’s look at what Match EQ has to say about the two recordings.
Even with smoothing turned completely off — which normally results in an EKG-like plot of spikes — these two signals are remarkably flat. There is perhaps a tiny increase of a db or so, on average, around 10khz in the impulse track, as well as perhaps a db more bass around 200hz. But in the world of Match EQ these deviations are tiny. Not only can I not hear them, but more importantly, when played independently, I would never peg one of these recordings as fake. While the impulse may diverge to some minuscule degree from the live mic in this instance, a subtle change in mic position between to live recordings would diverge much more. For all practical matters, these two takes are functionally indistinguishable.
But to hear just how different these two recordings are in reality, here’s the DI track with no impulse, and no additional reverb:
Wow. It’s fifty years post the rock and roll revolution, and all manner of amplifier designs are now available, both modern and retro — including digital simulations thereof which rock pioneers could not possibly have imagined. And yet guitar amplifiers and speaker cabinets exist in a kind of primordial, Discovery Channel-like symbiosis. Like the deep-sea anglerfish, the male of which actually embeds itself within the female until the two fuse together into one organism, the modern guitar amp and the nearly jurassic guitar speaker are wholly unsuited to any other partner.
Marshalling the Forces
Back, then, to the Marshall, and to the Arcade mixing process. I understood the basic sonic fingerprint of the Marshall’s factory speaker, the Celestion G12T-75, from playing this cabinet on occasion in the studio, and from my initial tests with the new one I had bought for the pine cabinet. It was a fingerprint that was pretty easy to identify: the scoop. Big round bass, crisp highs, and almost nothing in between. It’s like a time machine right back to the heydey of thrash.
This hyper midrange scoop is almost like grabbing the knobs on my old Boss HM-2 distortion pedal and inverting them — bass all the way up, treble all the way down — and it’s enhanced by the layout of the cabinet itself. Marshall’s iconic 4×12 speaker cabinet, which quickly became the industry standard design for such cabinets, was designed simply, as legend has it, to produce the biggest speaker cabinet that Marshall could build from one sheet of plywood. Even better, the angle panel, and its two upward-canted speakers, turned out to be an effective means of sound dispersion on-stage. But it was actually chosen purely for cosmetic reasons: Jim Marshall thought it looked cool perched atop the company’s square 1960B “bottom” cabinet.
The phase cancellation that happens in a Marshall cabinet, with four speakers crammed into a relatively shallow space, almost like an A-frame house, gives the cabinet its unique brand of notchiness. It’s a fun sound, but I always considered the pronounced midrange scoop something of a specialty sound for nostalgic metal riffs. It’s also a problematic sound for many guitar amps because the mids are where the meat of the tone lives. Scooping heavily in that area will exacerbate, relatively speaking, any problematic fizziness in an amp’s upper-end response by removing frequencies below it which would tend to disguise it. Put simply, scooped tones may sound fun to play, they’re not very often fun to mix, since most of the guts have been eviscerated.
Make Way for the Mids
To survive a trip through the 1960 without turning into Metallica’s black album, an amp needs to have strong midrange presence. And this probably explains why these cabinets are a good match for Marshall’s classic Super Lead and JCM800 amplifiers, given as they are to strong midrange emphasis.
Thankfully, the Cornford Hellcat that I used on that track shares this mid-centric profile. And thankfully at the time, even though I had no immediate use for the 1960A cabinet, I understood its placement in the pantheon of great guitar gear that I thought to impulse it with a ’57 in a few microphone positions.
And now, a few years later, when I dropped the impulses into the rhythm tracks, it all clicked into place: The rhythm guitars instantly ducked behind the synths, but not so far that they became an indistinguishable mush of static and bass. When the leads arrived, still playing through the H30, they were instantly audible with no further eq. This allowed me to crank the rhythm guitar even louder since there was no longer any audible competition. And here’s what that sounds like:
This combination of scooped rhythm tracks and mid-forward lead tracks has become something of a formula in guitar-driven rock mixing, including the chorus the Greatest American Hero cover in the accompanying cabinet construction montage. And the secret, again, is carving enough of a hole in the rhythm tracks that those guitars can in fact be turned up even louder without stepping on the lead lines. Throw in a serving of thunderous bass, and this results in the classic, and paradoxical “wall of sound”-style rock guitar mix in which everything is loud, and yet, somehow, individually intelligible:
Alongside the discovery of tonal recipes that worked, I was also the developing a sense of musical context that made subjective decisions more meaningful. Having composed and engineered dozens of musical cues for Cracking the Code, it was becoming clearer what sorts of sounds might work in the kinds of mixes I typically encountered, and by extension, easier to evaluate what kinds of tones might ultimately be useful, even before needing them in a mix.
Although the pine cabinet had evolved to become, essentially, a great lead playing cabinet, I now wanted to see what the T75 speakers sounded like again in that box in a rhythm application. With that in mind, I ran a test using some of my favorite Cracking the Code rhythm tracks, from the song “Epilogue”, at the end of episode three. For reference, here’s that track again:
And here are the guitars isolated, again running as usual through the 1960 impulses. I dig the riff, and the full mix turned out big, full, and powerful enough to support it.
Installing the 75s in the pine cabinet and re-recording the riff was a bit of a letdown:
I immediately missed the bass response, and found the top end too crispy. This was particularly apparent in the isolated guitar tracks:
Some of this may be due to speaker break-in, as this particular 75 has spent most of its life in a cardboard box rather than a wooden one. But the relatively dramatic difference in the resulting sound — more so than I would have expected — is stark enough that it may be that this combination of speaker and cabinet simply isn’t going to provide the girth of the bigger box. At least not for this type of heavy rock mix.
The 75 may not have worked in the pine cabinet as a rhythm speaker for this project, but I can hardly complain. The Marshall rhythm / Pine H30 lead combination has been a dependable one for heavy rock production in Cracking the Code, and really, how many sounds do you need? If you ask the purists, Eddie had precisely one.
In addition, at the breakneck pace we typically move to produce episodes of Cracking the Code, the maturity of impulse response technology has been a productivity boon. Not miking up an amp means near-zero setup time, and the unpredictability of the studio environs means that recording direct provides some indemnity against the sudden arrival of next door’s Spinal Tap tribute band just when you least expect it.
The Cracking the Code Impulse Pack
That being said, there’s no telling for what future use this particular speaker combination may be destined. Having already benefited once so far from the forethought to image the Marshall cabinet, I took the time to impulse the 75s in a several mic positions with a ’57. Down the line I’ll probably drop in a few more of the speakers currently sitting on the sidelines — like the G12M-65 — and see what they sound like in our mixes. I recently updated the wiring in the cabinet to use spade connectors, so no more soldering iron required, and speaker swaps are now quick and painless.
If you’d like to experiment with these sounds on your own tracks, and you’re handy with a DI box, I’ve bundled up a collection of the most commonly utilized impulses on Cracking the Code tracks, and I’m making them available as a downloadable bundle.
To run them, you can use a copy of Logic, which includes Space Designer, a convolution reverb plugin that can load impulse responses. Alternatively, you can utilize any of a number of freeware impulse response plugins available on both Windows and Mac. The web site of Redwirez, a provider of very high quality, professionally recorded impulse responses of unique and hard-to-find vintage cabinets, has a useful tutorial page on the subject.
Also, while hooking up your DI box, just remember to keep your speaker cabinet plugged in. You won’t be miking it, but the resistance provided by the speakers pushing back against the amplifier is critical to achieving a realistic DI recording. Although a variety of so-called “load boxes”, like the Palmer Speaker Simulator, which I own, exist for the purpose of enabling completely silent amplifier recording sans speaker cabinet, none that I’ve tested provide accurate results. Just remember to keep your speaker connected in addition to the DI box, and you’ll be fine.
Included in the impulse pack are all the pine cabinet H30 impulses — which, in addition to the go-to Shure SM57, also include the Sennheiser MD421 in multiple positions, and the deluxe Royer 121 ribbon mic in several positions. The Royer exhibits the classic ribbon response of rolled-off highs, and has been a favorite of recording engineers for guitar amp recording practically since its introduction.
In addition, I’ll throw in the Marshall impulses we use for rhythm tracks, and a couple Cornford 2×12 impulses to boot.
Three bucks for all that? Now that’s an impulse purchase.