Great progress has been made in recent years in modeling technology, allowing guitar players to emulate the sounds of a vast array of different amplifiers and effects in software. But what about the speakers we play the models through, do they make a difference too? Let’s find out.
If you are not familiar with the concept of modeling, here’s the quick lowdown. Modeling companies use sophisticated computerized tools to analyze the sound and behavior of the parts of the guitarists signal chain including amplifiers, effects, microphones and speaker cabinets. They then use the data to create software ‘models’ that allow musicians to emulate the sounds on various devices. There are many different implementations to suit most needs and budgets. Here are a few:
Purpose built modeling amplifier. Specialized computing devices designed specifically as guitar processors. These come in rack mount, cabinet top or pedalboard form factors, and include much of the required additional hardware including pre-amplifier, display screen, knobs, switches and connectivity. Examples include the Kemper Profiler, Avid Eleven Rack, Fractal Audio Axe-Fx, and Line 6 Helix.
Audio Workstation plugin. These run as plugins to a Digital Audio Workstation. They require that you have the other hardware and software pieces such as a Mac or Windows computer, audio workstation software, sound card, guitar interface etc. Take a look at Pete Thorn building a guitar tone from scratch using Apple Logic Pro and plugins from UAD to see how this works.
Standalone software. Similar to the above, standalone applications run on a computer but don’t require that you already have audio workstation software. Scuffham S-Gear is a standalone application that’s also available as a plugin. Software like Apple Garageband is a kind of cross between plugin and standalone since it is an audio workstation of sorts with some existing guitar emulations that comes for free as part of OS X.
Mobile software. A big growth area is in modeling software for portable devices like smart phones and tablets. Check out Jam Up Pro.
Included in most if not all of these, along with the models of various vintage tube amps and boutique effects, are cabinet simulators (cab sims). Cab sims are models of guitar cabinets and speakers.
In the un-modeled world of the traditional tube amp, a guitar speaker cabinet is typically designed to work best in conjunction with a guitar amplifier. The speaker works within a limited frequency range, and the cabinet and speaker themselves contribute significantly to the overall tone. The roar of a Marshall stack, or the chime of a vintage Vox AC30, come partly from the amps themselves, but also in a large part from the speaker drivers and cabinets used with them.
Over the years, designers have learned how to build cabinets to accentuate specific amp characteristics. Musicians have learned how to pair amps and cabs together to create their desired tone. Compare this with studio monitors or even hi-fi speakers. In this case the designers are pretty much aiming for the reverse. Their job is not to add add anything at all, and to represent the recorded signal as accurately as possible.
Check out these two speaker frequency response graphs:
If you are not familiar with a frequency response graph, the X axis displays frequency in Hz, and the Y axis displays sound pressure level (SPL) in dB. The black line shows relative loudness of specific frequencies.
The graphs are from two different speaker drivers from Eminence. Eminence manufacture an excellent range of guitar speakers as well as pro audio drivers for use in monitors, PA systems etc. The top graph is from their Legend 1258 speaker which is intended as a rock guitar speaker. The bottom graph is of the Alpha 6A which is a mid range driver for pro audio applications.
Notice how the graph at the bottom is relatively flat between 120Hz and about 5kHz? If we were to add a woofer to handle below 120Hz, and a tweeter for the 5kHz – 20kHz range, we would have the basis for a pretty decent monitor or hi-fi speaker without any frequencies being hugely boosted or attenuated by the speaker. By contrast look at the top graph. Notice the big peak at around 2.5kHz and how quickly it drops off after about 4kHz? 2.5K is where we get bite from an electric guitar. Boosting around this frequency helps cut through the mix with an aggressive rock tone. Above 2.5kHz and into 5kHz or so is where we can get harsh and brittle, so attenuating those is just fine for this application. The speaker driver in the bottom graph has been designed to be as flat as possible for a monitor application, whereas the driver in the top graph intentionally exhibits certain characteristics to assist in generating a particular guitar tone.
This presents a bit of a challenge for modeling systems. If I turn on a cab simulator and then run the signal through a physical guitar cab that adds additional frequency characteristics of it’s own, then all the careful work of the modeling designer is undone. My resulting signal is nothing like the actual cabinet I’m trying to sound like, and has the potential to be seriously altered if my virtual and actual cabinets end up doubly accentuating or canceling out particular frequencies.
So what can be done? Well, firstly I could just turn the cab sim software off altogether. Most modelers will let you do this. This will work up to a point because now my tone only has one cabinet element; that of my physical guitar cab. I avoid the situation of the virtual and physical cabs fighting each other. The trade off is that now I have lost the features of the cabinet simulator. My tone will always have that element of my physical guitar cab. I can’t take advantage of all the work the modeling engineers did to make me sound like I’m standing in front of a wall of Marshall 4 x 12 stacks, dang it.
What about using my studio monitors, hi fi speakers, or headphones? This can work with a couple of caveats. It’s fine in the studio or practicing at home, but I for one am not taking my expensive home theater speakers to the local bar to get beer spilled over them. Even if I did, they are just not going to be loud enough in most cases. There’s also another issue I’ll get to in a minute. For practice and recording I can just use headphones, but I normally end up tangled in wires.
Not using guitar speakers at all is an option, and in fact is the way most people use these today. Use headphones at home (cable tangle notwithstanding). For live shows connect the modeler direct to the desk of the PA system and use in ear monitors. For this to work though, the venue has to have a PA in the first place, and facilities for in ear monitors, which you have to be comfortable using, not to mention be able to afford, etc etc. This is not really a solution for playing open mic night at the coffee shop, or rehearsing in moms garage.
Using a PA though, poses an interesting question. PA speakers are designed for full range reproduction like a studio monitor or hi fi speaker. They usually have multiple drivers to reproduce a wider frequency range in as neutral a way as possible as in the bottom frequency response graph above. What If we used a power amp and a small PA speaker couldn’t we create a suitable system? We could, and there are a few purpose built small PA systems and powered speakers available for the job such as these from Line 6 and these from Bose. We could also use a rack mount power amp and a monitor wedge to build our own version.
The newest class of systems integrate the technology of the full range speaker into a traditional guitar style speaker cabinet. These systems allow you to get the full benefit of using cab simulators in a format that looks more like a regular tube combo or stack.
OK, so there are plenty of choices there that should cover the needs of most, of us. However, there is one last item to cover, and here it gets complicated, not least because it is subjective. Many players find that listening to their guitar tone through a full range system, even with a cabinet simulator turned on, at the very least sounds different, and at worst sounds harsh, grating and unpleasant. So why is this? As I said, it’s complex and different for everyone, so there is no one answer; what works for my signal chain and hearing, may not work for yours. You will have to do some tuning to suit your particular environment, but hey, that’s where a lot of the fun is, right? Here are a few possibilities.
If you are used to standing in front of a 4 x 12 cab and have replaced it with a pair of 6″ full range drivers, then part of the difference may just be in the volume of air in motion. Even if the smaller drivers can reproduce the same frequencies, they are never going to move the sheer volume of air that 4 x 12 can, and this can result in a different perception: ‘There’s no substitute for cubic inches’, as someone once said. If this is a problem for you, then look for monitor systems with larger drivers. Remember, you are going to have to carry them around though.
Since the monitor systems are capable of operating at higher frequencies, they can reproduce harmonics that the guitar speaker did not. Even with a cab sim in place, these harmonics can be present and reproduced by the speaker. You can use some EQ to roll off the higher frequency harmonics if you don’t like them.
Thirdly, get used to it. The fact that the type of modern guitar systems we are discussing here are capable of reproducing tonal characteristics way beyond those of traditional systems is a great thing. It opens up all sorts of possibilities for new guitar sounds that go way beyond the limitations we had before. Try something different. None of the explanations here are rules. If you find that running two different cab sims in series through a broken phone speaker sounds good, then do it. Remember that many of the great guitar tones we are trying to emulate now, were first made by someone doing something they weren’t supposed to.