Always consult the official Owners Manuals first!
Pages A-C are under review. Pages D-Z have been checked and are up-to-date.

Cab block

From Fractal Audio Wiki
Jump to navigation Jump to search

H cabs.png

The Cab block

Available on which products

  • Axe-Fx III: 2 blocks (4 IRs per block)
  • FM9: 2 blocks (2 IRs per block)
  • FM3: 1 block (2 IRs)
  • Axe-Fx II: 2 blocks
  • AX8: 1 block
  • FX8: no

Channels or X/Y switching

  • Axe-Fx III and FM3 and FM9: 4 channels
  • Axe-Fx II: X/Y
  • AX8: X/Y

About cab modeling

Cab block.png

Introduction to cabinet modeling

Read this: Amp and Cab modeling for beginners

About Impulse Responses (IRs)


The Cab block relies on Impulse Responses (IRs) to reproduce the sound of speaker cabs.

Read this: Impulse responses (IR)

"If you want "nonlinear IRs" you need to use something like Volterra kernels. I've experimented with this and, in fact, the Speaker Drive and Speaker Thump parameters essentially create higher order Volterra kernels based on various amp parameters. The harsh reality is that speakers are pretty darn linear. Things like cone cry are design/manufacturing defects and shouldn't be modeled IMO. There is some low frequency distortion and we model that but a classic, linear IR is the gold standard for replicating the sound of a speaker. Armchair pundits like to pontificate about "IRs are the weak link, blah, blah" but they're typical internet loudmouths who have little to no expertise and would fail a double-blind test every time. Now, thermal compression is definitely a factor and we model that quite extensively. As you play the voice coil heats up and its resistance increases. The radiation resistance does NOT change, however, so the resulting acoustic output decreases. There's a lot more involved when simulating a tube amp though and I can't disclose that stuff because it's proprietary. Another factor is the actual speaker impedance. It's a function of displacement. Again, we model this extensively. As the speaker moves in/out of the magnet the Bl product changes and therefore the inductance and various other electrical parameters change." [1]

Cab processing on the AX8, FM3 and FM9

The FM3 has the same Cab block functionality as the Axe-Fx III, but provides a single Cab block instead of two, supports two IRs instead of 4, and some functionality has been left out, including smoothing.

The FM9 has two Cab blocks like the Axe-Fx III, but supports two IRs instead of four per block, and some functionality may have been left out (such as smoothing).

On the AX8, FM9 and FM3, cabinet modeling runs in an CPU accelerator instead of the core DSPs, which saves CPU. On the FM9, the CAB 2 block doesn't run in the CPU accelerator so it takes up more CPU than CAB 1. [2]

(AX8) "They are done in an a separate accelerator so they have minimal CPU loading. An accelerator is a DSP unit dedicated to performing a defined task. In this case it performs convolution. The difference is that a "DSP" as we commonly call them is really just a microcomputer optimized for Digital Signal Processing. The term DSP most correctly refers to any device that does signal processing using numerical computations. That device can be an FPGA, ASIC or fixed hardware unit. In the AX8 it is a convolution processor." [3]

(FM3) "The Cabinet block uses an FIR accelerator to do the IR processing. This FIR accelerator offloads processing from the CPU and, as such, doesn't reflect in CPU usage." [4]

”FM3/9 only support 1K samples for normal IRs since the IR processing is handled by a coprocessor whose max. length is 1K samples.” [5]


Cab processing does not increase latency.

"Cab blocks do not add any latency." [6]

Cab modeling developments


Axe-Fx III product information, with firmware Ares:

"The redesigned Cabinet block features a 4-channel mixer based on our popular Cab-Lab software, providing the capability to mix and remix IRs on-the-fly as you would with real mics on a speaker cabinet. Factory content includes selections from the best of today’s IR producers and artists, including Fractal Audio, AustinBuddy, Celestion, ML Sound Lab, Ownhammer, Chris Broderick, John Petrucci, Chris Traynor & James Santiago, Valhallir, York Audio, Dr. Bonkers, and more. An additional 2,048 “User Cab” memories allow you to load Cab Packs (including any of those compatible with the Axe-Fx II) or 3rd-party IRs, and a built-in utility allows you to capture and save your own speaker tones (now with 16 “Scratch Pad” locations!) Our celebrated Tone Matching block is also improved, now with the impressive ability to clone the tone of an amp or recording in UltraRes."

  • The Cab Picker makes it easier to select cabs, employing filters. On access the Cab Picker on the hardware, press ENTER in the Cab Number field.
  • Muting an IR in the Cab block decreases CPU usage.
  • The SPEAKER SIZE parameter and separate microphone modeling are not supported.
  • The Cab block normalizes its output level when using multiple IRs. The LEVEL parameter of each IR doesn't simply set the output level of that IR, but it determines the relative levels between the IRs when more than one is loaded.
  • When set to Stereo Input, two of the IR slots are fed by the left channel, and two are fed by the right channel (Axe-Fx III).
  • When powering on, the unit reads all user cabs. This happens in the background.
  • IRs can be visually aligned.
  • Room ambience has been improved and includes floor reflections.
  • Micro Delay has been changed to MIC DISTANCE.
  • IR LENGTH lets the user adjust the length of an IR.
  • LOW CUT and HIGH CUT and SLOPE can be set per IR.
  • Smoothing, Proximity and Delay can be set per IR.
  • Much steeper filter slopes are possible.
  • A separate IR Player block also processes IRs, but has less functionality and therefore requires less CPU than the Cab block.

More information in the Owners Manuals

"Another cool thing is the Cab block. You can mix up to four IRs each with independent Pan, Distance, Proximity, Smoothing (De-phase). And it has four channels so you can switch between four completely different mixes, instantly." [7]

(about changing IR levels in the Cab block) "The volume stays constant." [8]

(about normalization in the Cab block) "Yes. This keeps the volume consistent regardless of the number of IRs and their mix levels. For example if you were to use two IRs and set each at -3 dB the volume would be half as loud. Behind the scenes it figures this out and compensates." [9]

"Adjusting IR levels is not possible. Cab-Lab automatically normalizes IRs for "unity energy". 99.9% of the time this results in IRs that are the same volume but every now and then an IR will have energy outside the normal range of hearing which confuses the normalization routine. It's superior to the usual amplitude normalization but not without its faults." [10]

Firmware 17 and later for the Axe-Fx III feature FullRes IRs. Read more

Firmware 22 and later for the Axe-Fx III feature "DynaCab" cabinet modeling, see below.


Firmware 22 and later for the Axe-Fx III feature "DynaCab" cabinet modeling. This allows freely positioning of the microphone. Set the Cab block to "DynaCab" mode to be able to select a mic type and set its position and distance. Behind the scenes this will load the appropriate IR. A graph visualizes the parameters.

List of DynaCabs

The available microphone types are: Dynamic 1 (Shure SM57), Dynamic 2 (Shure SM7), Ribbon (Royer 121?), Condenser (Soyuz 023).

Additional DynaCab packs with more mic choices can be bought in [p Fractal Audio's webshop].

"You can mix up to four DynaCabs. You can mix-and-match any combination of cabs and mics. Position sets the radial distance of the microphone from the center of the speaker. Distance sets the distance from the grill cloth. As you move the mic away from the grill the bass typically rolls off and more room is heard." [11]

"We actually used six mics during the captures, two of each type. I picked one from each type that I thought sounded best. The dynamic is an SM57 (naturally). I actually prefer the SM7 captures we made but given how iconic the 57 is on guitar cabs I went with that instead." [12]

"While the SM57 is harsh on its own that extra brightness is useful when blending mics. Yes, I prefer the SM7 on its own but the SM57 is ubiquitous in studios when blending two or more mics." [13]

(Soyuz) "Best sounding condenser mic I've ever used.” [14]

"I found it's worth exploring each cab to find the sweet spot. Some of the cabs I like the mic near the edge whereas others I like it closer to the cap." [15]

"Dyna-Cabs IRs were captured with a Neve preamp. Legacy IRs were shot with an API. API preamps have more mids. I prefer the sound of Neve." [16]

"The closer you are to other speakers the more interference you get from them. It's even more noticeable as you pull the mic back from the speaker (again, not surprising). We actually captured the entire speaker for this Dyna-Cab stuff. I then auditioned radially and found that I preferred the IRs that moved away from the other speakers. I.e. for a 4x12 I preferred a radial line from the center to the nearest corner of the cabinet. Even single speaker cabs I heard a difference in some cases. For example, our Tweed 1x12 the speaker is off-center on the baffle board so it doesn't sound the same moving left vs. right." [17]

"0.0 represents as close as possible to the grill cloth without touching it." [18]

Onboard factory DynaCab IRs are 2048 samples on the Axe-Fx III. The resolution can be decreased to Standard (1024 samples) to save CPU. On the FM9 and FM3, onboard factory DynaCabs are 1024 samples. Fractal Audio's DynaCab packs are based on longer (Ultra-Res) IRs.

They are time-aligned without destroying phase information, allowing them to be mix-and-matched.

"The files are initially aligned to the same reference point -- a universal standard across all DynaCapture IRs. In other words, IRs for mics that are farther away have been time compensated to be aligned with IRs for mics that are closer. You can use the controls on the Align page in the usual way, however, to add time to any individual IR." [19]

DynaCabs have no Proximity parameters.

DynaCabs support smoothing.

Enable "Auto DynaCab Impedance" in the Amp block to make automatically select the correct Speaker Impedance Curve in the Amp block when selecting a DynaCab in the 1st IR slot of the Cab block. If the preset contains only the CAB 1 block, Auto Impedance controls both AMP 1 and AMP 2 blocks. If the preset contains CAB 1 and CAB 2 blocks, CAB 1 controls AMP 1, and CAB 2 controls AMP 2.

"The Amp block now features “Auto Dyna-Cab Impedance”. When set to ON the speaker impedance curve of the Amp block will follow the Cabinet Type in the first mixer slot of the associated Cabinet block. I.e., if the Cab Type in the first mixer slot of Cabinet 1 is, say, 4x12 5153 and the Mode is Dyna- Cab then Amp 1’s speaker impedance will automatically be set to 4x12 5153."

"Cabinet 2 Slot 1 controls the SIC of Amp 2. Cabinet 1 Slot 1 controls the SIC of Amp 1." [20]

"If only Cab 1 is in the grid then that controls both Amp 1 and Amp 2. Otherwise Cab 2 controls Amp 2." [21]

"All Dyna-Cabs have corresponding impedance curves that were taken from the actual cabs." [22]

Leon Todd:

FX8 and speaker simulation


The FX8 does not provide cabinet modeling or another way to process IRs. But you can use its tools to get close.

To simulate a speaker, use the approach described in this post. This lets you use the FX8 as a speaker simulator, for example when your amplifier is connected to a load box, or when using a preamp-only pedal.

The link above also shows how to simulate a basic amp, without an Amp block.

Combined, these tricks let you simulate both an amplifier and speaker cabinet, without Amp and Cab blocks.

Cabinet modeling: on or off

If you never use cabinet modeling, turn it off in the Global Settings menu. This will decrease CPU usage. You can also leave out the Cab block per preset.

You can bypass the Cab block in a preset to disable cabinet modeling, but this will not decrease CPU usage.

Selecting a cab

How to choose

It’s a matter of personal preference which cab(s) you want to use with a specific amp model. You can choose a traditional combination, or think out of the box. The differences between cabs can be huge. The cab has at least 50% impact on the sound of an amp+cab combo, more than adjusting amp controls.

When comparing cabs, don't judge too quickly. Each time you select another cab, your ears have to adapt. Also, you need to adjust the amp settings to suit the selected cab.

Traditional combinations of amps and cabs are listed in the wiki list of Amp models and in Yeks Guide to the Fractal Audio Amp models.

Leon Todd's demonstrates his method to select an IR

Background information on guitar speakers and microphones

Factory cabs

The amp modelers come with many built-in factory cabinets. These are also referred to as "stock cabs" or internal cabs. These can be selected with the Cab block in Legacy mode.

Number of factory cabs:

Axe-Fx III + FM3 + FM9 – 2048 "new" factory cabs (2 banks of 1024), and 189 "legacy" cabs (same ones as in the Axe-Fx II XL+ and AX8)

"When I was capturing IRs I specifically chose to obfuscate the names to force people to use their ears." [23]

Axe-Fx II XL + XL+ – 189

Axe-Fx Mk I + II – 132

AX8 – 189

All stock cabs are time-aligned (Minimum Phase Transform), which means that you can mix them without phasing issues.

Detailed list of all stock cabs

(stock cabs in the Axe-Fx II and AX8) "The factory IRs were hand-selected by me after auditioning thousands of OH and RW and other IRs. Some of the IRs are custom mixes of mine. My rule-of-thumb was to select as neutral sounding IRs as possible. However, what I like may be much different than what others like. Some people complain the Axe-Fx sounds too bright. Others say it's not bright enough. It's a no-win situation. This is why I've been harping on capturing IRs. It's personal preference. Producers probably spend more time perfecting mic placement than anything else when getting guitar tones to tape. An IR is the same thing, it's capturing the mic and placement." [24]

To generate a list of the cabs used in presets and scenes, use FracTool.

Note: stock cabs cabs only be mixed with other cabs on the hardware itself. Cab-Lab can't mix stock cabs with other cabs.

Factory cabs can be captured and turned into external SYX and IR files with Cab-Lab, as explained here.

Alternatively, use Tone Matching.

User cabs

If you want to look beyond the stock cabs, try external impulse responses. The amp modelers provide "user cab slots" (in the Cab block's Legacy mode) which can be filled with "external cabs" (impulse responses), using Fractal-Bot, Cab-Lab, the Editors or a MIDI librarian.

External IRs for Fractal Audio processors are files with a "SYX" filename extension, sampled at 48kHz, 24 bit.

Files with an "IR" filename extension are also IRs for Fractal Audio processors, but restricted for use in Cab-Lab. It's a proprietary Fractal Audio format. You can't load these directly into the hardware.

IRs in WAVE format can be imported directly into the Axe-Fx III and FM3 with the Cab Manager tool in the editor. Simply drag-and-drop any number of .wav files into the Browser pane to allow Axe-Manage to convert them. Cab-Lab can convert a batch of WAVE files in one go.

Number of user cab slots:

Axe-Fx III Mark I – 2048 (2 banks of 1024)

Axe-Fx III Mark II – 2048 (2 banks of 1024), plus a bank for 64 FullRes IRs

FM3 – 1048

FM9 – 1048

Axe-Fx II XL and XL+ – 1024

Axe-Fx II Mark I and II – 100

AX8 – 512

The amp modelers and software editors display the names of the external cabs in the user cab slots. The name is contained in the sysex data of the file. Impulse responses can be renamed using the editor or Cab-Lab. The name is shown in italics or a different color when it's an UltraRes impulse response.

To empty a user cab slot on the hardware, use the software editor or Cab-Lab. Some modelers (including Axe-Fx III and FM3) allow you to delete ALL user cabs at once through the Utilities menu. Be careful!

Tutorial: loading Cab Packs

Inventory of commercial and free IRs



Read this.


Scratch-Pads (which are the very last user cab slots) are reserved "dummy" locations, meant for temporary loading of impulse responses. This allows auditioning impulse responses without overwriting any of the user slots. The number of Scratch-Pads depends on the hardware. Scratch-Pads are erased when powering off the unit.

Firmware 17 for the Axe-Fx III implements a FullRes Scratchpad for auditioning IRs during the capture process.

Position of the Cab block on the grid

Iii grid.jpg

In the "real" analog world, it makes a difference where you put effects: before or after the speaker cabinet. It's different with digital processors.

(Javajunkie, Axe-Fx Standard/Ultra) "You can place the effects loop anywhere in the chain (just add the fx loop block). Unless you are running a stereo cab or 2 mono cabs panned hard L/R, you may want to place stereo effects after the cab. The cab is a linear time invariant effect (unless you add drive) so effects like delay and reverb will sound the same before or after it. As Cliff and others have stated on numerous occasions LTI effects can be placed before and after each other and they will sound the same. Only when placed before or after non-LTI effects (drive, amps, et. al) it really matters. The one caveat there is that some effects are mono, placing effects before and after that makes a difference." [25]

"The difference in having the cabinet before or after the effects is usually subtle. It depends on how non-linear or time-variant the effect is. For effects like EQ, which are linear and time-invariant, it doesn't matter at all. For slightly time-variant effects like chorus and flanger the difference isn't very pronounced. For highly time-variant effects, like pitch shifting, the difference can be marked."

"Linear means that the output is related to the input by a straight line: y = mx + b. Filters are example of linear systems. A cabinet IR is a filter. Distortion is an example of a nonlinear system. Linear systems are associative and commutative. Associative means that a * (b * c) = (a * b) * c. Commutative means that a + b = b + a or a * b = b * a. Therefore you can do cab -> eq (a * b) or eq -> cab (b * a). The cab block is "completely" linear if motor drive is non-zero but it is "wide sense stationary" so you can treat it as linear." [26]

"Since a cabinet is linear (or mostly linear) the order is unimportant as linear systems are commutative (a+b = b+a). However if the cab block is mono your effects will collapse to mono if placed before." [27]

"Since linear systems are commutative EQ > Delay is the same as Delay > EQ. Now, if the system isn't linear (i.e. there's some distortion) then the sound will change. Also if the system is time-variant the sound will change but most modulation is wide-sense-stationary enough to not create a noticeable change if the order is reversed. The most likely explanation is stereo delay into a mono cabinet which will destroy the stereo image." [28]

"The cab block is level-dependent if the Motor Drive is non-zero. So if you turn up/down the level out of the amp block you may need to compensate by doing the opposite with the Motor Drive." [29]

"You gain nothing putting it before the cab and risk collapsing the stereo image if the cab is mono." [30]

"Given that most post effects are linear or "wide sense stationary" the order of effects after the amp doesn't matter. Reverb -> Cab is theoretically equivalent to Cab -> Reverb because linear systems are commutative (i.e. a * b = b * a). However... if the cab block is mono then you'll collapse any stereo effects to mono. Or if the cabs aren't panned fully L/R you'll lose your stereo imaging. Pitch effects are not linear so putting them before the cab block will sound different than after. Anything that causes distortion is not linear so the order matters. If the distortion is subtle then the order is less important." [31]

(Bakerman, cabinet blocks in parallel rows sound louder than a single Cabinet block) "It depends on how you're panning. Assuming a mono signal sent to cabs: Stereo cab w/ Pan L and Pan R fully left & right will be the same output level as 2 mono cabs w/ balance L & R. If pans/balances are centered the 2 mono cabs will be 6 dB louder. Balance elsewhere would be between 0 and 6 dB louder, and balance doesn't correspond 1:1 to pan L/R for the same placement. Balances will need to be further toward -50 or 50." [32]

Cab block: mono or stereo

The Mono / Stereo mode of the Cab block is important.

The Cab block will sum the incoming signal to mono when using a single IR. It will process and maintain an incoming stereo signal when set to stereo or when using two (panned) Cab blocks to handle the left and right channels.

If a stereo Cab block is followed by a mono effect, such as Drive, the resulting signal will be summed to mono.

Axe-Fx III — If the Cab block is set to Stereo Input, two of the four IR slots are fed by the left channel, and two are fed by the right channel.

"In Mono left and right are summed and fed to all four cab slots. In Stereo left goes to slots 1 and 3, right goes to 2 and 4. Stereo: left goes to slots 1 and 3, right goes to 2 and 4. Left: Left goes to all slots. Right: Right goes to all slots. Sum L+R: Sum of L+R goes to all slots." [33]

More about mono and stereo signals

Cab block and CPU usage

CPU usage of a Cab block depends on its configuration and the unit.

AX8 + Axe-Fx II:

  • a mono Cab block uses less CPU than a stereo one
  • an UltraRes IR uses more CPU than a Mono or Stereo Normal block.

Axe-Fx III + FM3 + FM9:

  • mute an IR in the Cab block to decrease CPU usage
  • use the IR Player block (Axe-Fx only) instead of the Cab block for less CPU usage.

FullRes IRs use more CPU than Ultra-Res IRs.

When importing presets created on the Axe-Fx III or FM9 into the FM3, IRs are set to Standard resolution to save CPU.

See which cabs are used in presets


FracTool can poll the device and show which cabs are being used in presets and scenes.

It also shows which user cabs are not being used in presets, so you can decide to delete these.


Cabinet modeling usually reproduces the sound from a close-mic'd speaker. The player hears the sound through headphones, FR amplification, IEM or studio monitors. This differs from the so-called "amp/cab in the room" sound, where the player hears the guitar sound coming from a traditional guitar rig.

Read this to get the "FRFR sound" closer to the "amp/cab in the room" tone

FullRes IRs are Fractal Audio's approach to adding room ambience to the sound through headphones and IEM, and recordings.

Sharing a preset with an external IR

When you use an external IR in a preset and you want to share the preset/sound, you need to share the preset as well as the impulse response. There are two ways around this:

  • integrate the impulse response in the preset by replacing the Cab block with a Tone Match block, after having captured the tone of the Cab block.
  • create a Preset-Cab bundle, if the firmware and editor support it

It's NOT permitted to share commercial IRs (license violation)!

Preset-Cab bundle

Read this: Preset-Cab bundle

Recording 4 different cabinet signals

This G66 tutorial demonstrates how to create 4 separate cabinet signals in the Axe-Fx II, to be mixed at will. It comes down to using 2 stereo Cab blocks, with one of these connected to an FXL block to feed Output 2. In both Cab blocks the impulse responses are panned hard left and right. The stereo outputs 1 and 2 are connected to 4 separate channels on the mixers.

The Axe-Fx III makes this much easier to accomplish, because it supports four IRs per channel.

Cab block alternatives

IR player

The IR Player block on the Axe-Fx III can process a single IR, offers less features than the Cab block and therefore requires less CPU.

The FM3 and FM9 do not provide this block.


A Filter block can be used to create an "amp+cab-in-the room" sound. Here's a demonstration. Cliff's walk-through.

Importing IRs in the editor

In Settings > Preferences you can instruct the editor how to import external IRs (.syx or .wav).

You can choose between MPT and auto-trim, and you create UltraRes or non-UltraRes cabs.

Bad IR

"IRs are stored in FLASH. FLASH memory can become corrupted though it is rare. If you've ever plugged a USB memory stick into your computer and some of your files were damaged that's the same thing. Reloading the IR is the solution." [34]


"The Cab and IR Player blocks automatically "phase correct" IRs so that the peak signal is positive." [35]

Tutorials and more information


The Owner's Manual explains all parameters.

Input Select

This parameter lets you select the source signal that enters the Cab block. For example, if you wish to run two panned Cab blocks in an Axe-Fx preset, you can use this parameter to force one side of the signal to go into one Cab block, and the other side into the other Cab block, for stereo separation.

Room ambience

Read this: Room ambience

Low Cut, High Cut, Filter Slope

Most IRs have been captured "close-mic'd", and produce a lot of highs and lows.

HIGH CUT and LOW CUT in the Cab block (= low-pass and high-pass) allow you to EQ the material, preventing boomy bass and harsh sounds. Equivalent to using EQ controls on a mixing board to position the sound of the guitar in the mix. These are VERY important parameters to fine-tune the tone.

Important: while these parameters appear on the Preamp page in firmware Ares and later, they are still operational when the preamp simulation in the Cab block is turned off.

Common settings are 80-200 Hz for LOW CUT (cuts bass), and 5-10 kHz for HIGH CUT (cuts treble). Of course: YMMV, as demonstrated by Justin York's (Paramore) approach:

"I think it all depends on what you’re going for. I always start with 20-20k and dial in the amp. An IR “hears” exactly what the mic picked up so start wide open if you want the sound of a real mic’ed cab. It sounds better in a mix, but for solo playing, cutting the highs may sound more pleasant to you. For live sounds you may want to high cut, but I find that cutting a little 2k-4k 1-2dB gets rid of the fatiguing frequencies when cranked. Cutting a little top will slightly tame your tone, while cutting too much will take away note separation and you’ll lose yourself in the mix. For recording, leave it wide open on the top and cut around 80 Hz on the low end." [36]

Low Cut and High Cut are also available per individual IRs in the Cab block, including selectable filter slopes.

FILTER SLOPE in firmware Ares and later selects between 1st order (6 dB/octave), 2nd order (12 dB/octave), 3rd order (18 dB/octave) and 4th order (24 dB/octave) filters for LOW cut and HIGH CUT. Firmware Ares and later lets you use different slopes for LOW CUT resp. HIGH CUT. The "pop" when switching between the values is normal.

More about filter slopes

"Using Low Cut in the Cab block is akin to what you would do in the studio to carve out room for the bass player." [37]

"LOWCUT FREQ" in the cab block sets sets the -3dB point of a highpass filter at the output of the cab block." [38]

"If at the min/max the filters are off." [39]

"People often talk about applying low cuts and high cuts. This is because the cabinet models used in modelers are almost always (with a couple exceptions) based on near-field samples of guitar cabinets. IOW, the mic is pushed up against the grill cloth. This just happens to be the way that record producers/engineers mic a cabinet in the studio and the way guitar cabs are mic'd on stage. This is done primarily for isolation reasons. The downside of this approach is that the resulting tone will have a lot more lows and highs than when listening to the amp+cab "in the room". What the mic "hears" when pushed up against the grill cloth is not the same thing that we hear standing 10 feet away. The most common technique to deal with this is to simply cut out the lows and highs using blocking filters, e.g. highpass and lowpass filters. Producers routinely do this when mixing as excessive amounts of lows and highs will cause the guitar tracks to get "lost in the mix". Live sound engineers often do the same thing. The Cabinet block has blocking filters built in for just this very reason. You can also use a couple dedicated filter blocks or a parametric EQ block. For now let's use the Cabinet block. My personal settings are Low Cut around 80 Hz and High Cut around 7500 Hz and Filter Slope set to 12 dB/octave but these are just a starting point. Far-field IRs are available but they are rare due to the difficulty in obtaining them. They require a large facility and special techniques making the process impractical in most cases. So, until an abundant source of far-field IRs are available we need to think like a producer/engineer who is dealing with the mic pushed up against the grill cloth. This means shaping the tone with EQ to remove unwanted frequencies." [40]

"The slopes are all maximally flat (Butterworth)." [41]

(Cooper Carter) "Touched on one of my favorite topics here, so I did want to chime in about cuts. Something I discussed a good bit in my AxeFest clinic. Nothing wrong with sometimes very aggressive cuts in the cab block, especially with close mic IRs, as Cliff said. Close mic IRs bring in so much of that amazing depth and complexity we love in the upper mid-range of a tone but often introduce too many other frequencies for a balanced mix. In my experience it comes as a surprise to many how narrow the frequency band of guitar tones is much of the time in pro mixes, whether live at front of house or at the board in a studio. What sounds rockin' to you solo isn't always very relevant to what sits well in a mix. Cutting down to 4-5k and up to the mid or even high 100s not only isn't unheard of but would more be described as common practice. There are a lot of instruments in a band. Concentrate on letting the guitar speak where it speaks best! Just one of the inumerable strengths of the Axe-Fx is that we have the ability to sculpt with this kind of detail with the click of a mouse. Something standard amps can only dream of!" [42]

"I don't think there's a rule. Sometimes I'll drop it as low as 6K. Sometimes it's wide open. Depends on the cab. Another way to tame the high end is the studio trick of placing the mics at different distances. Use the Align page to do this. As you separate the IRs in time it will put a notch in the high frequencies." [43]

Demonstration of slope:


Demonstration video by Cooper Carter

IR Length

(Firmware Ares and later) Applies to individual IRs in the Cabinet block. Shortening the length can remove room reflections and/or decrease CPU usage.

More about the length of IRs


This controls a sophisticated process that removes the “phasiness” from impulse responses by reducing the prevalence of peaks and valleys in the IR. This yields a more “amp/cab in the room” experience. This is especially helpful when using multiple impulse responses.

Smoothing is identical to "De-phase" in the Axe-Fx II.

Smoothing is supported for DynaCabs.

Cab-Lab can apply this process when mixing impulse responses together to produce an IR with the effect built-in. This way even the AX8 and FM3 and FM9, which do not support De-Phase / Smoothing in the hardware, can benefit from this feature.

The processing required is extreme and the control can have some lag. No extra CPU usage or audio latency, however, is incurred.

"Close-mic'd speakers can sound "phasey" because you are in the near field. When sampling the near field of any source the frequency response and beam pattern is rough. This occurs due to multiple spherical waves arriving at various phase angles. These multiple waves come from the various modes of the speaker, internal cabinet reflections and from other speakers in the cabinet. In the far field the response is more uniform because the wavefronts get flatter and the phase angles converge. The De-Phase parameter removes some of the phasiness due to multiple wave arrival using a complex FFT technique." [44]

"The higher the setting the more "character" you remove. De-Phase removes some of the character but that's precisely what you want to do as a cab has less character in the far field." [45]

(Why is De-Phase necessary?) "You don't listen to a guitar speaker with your ear against the grill cloth." [46]

"It's so simple that even experts in the field don't realize why it works." [47]

"It smooths the IR in the frequency domain." [48]

"Cabinet smoothing does not increase CPU usage. Must be something else you changed." [49]

Motor Drive

This parameter was part of the Amp block before Quantum 9 (Axe-Fx II and AX8). In Quantum 9 for the Axe-Fx II it was replaced with Speaker Compression in the Amp block. Motor Drive is still present in the Cab block in the Axe-Fx II (not on the AX8). It models the effect of high power levels on the speaker.

"Accurately models the compression of guitar loudspeakers by factoring in the reactive aspects of the compression."

"The Motor Drive simulation is available in both the Amp block and Cab block now. It is recommended to use the simulation in the Amp block when using an FRFR configuration as the Amp block simulation uses the speaker resonance information in the calculations whereas the Cabinet block uses fixed values. When using a conventional guitar cab, or a hybrid configuration with monitoring via a conventional guitar cab and speaker emulation to FOH, the Motor Drive in the Cabinet block can be used instead. The simulation in the Amp block also has the advantage of being independent of the block’s output Level control."

"Gain monitoring of the Motor Drive is available on the MIX page of the Cabinet Block and the PWR DYN page of the Amp block. In the case of the Amp block the monitoring is available when the Motor Drive parameter is selected. Note that typical guitar speakers have around 3-6 dB of compression when driven hard with American speakers being on the low end of that range and British speakers being on the high end. Some speakers can exhibit even more compression than this with compression amounts of 8 dB or more depending upon the magnetic materials used and the construction of the speaker motor."

"The thermal time constant of the virtual voice coil is adjustable using the “Motor Time Const” parameter. Typical guitar speakers are anywhere from 0.05 to 1.0 seconds depending upon the mass of the voice coil and the materials used."

"Set it to 4.5 and rip the knob off." [50]

When using two UltraRes cabs in a preset, don't use Motor Drive on just one of these, because this will introduce comb filering (phase cancellation).

"Motor drive isn't EQ. It models efficiency reduction due to thermal effects." [51] And: "What I have found is that thermal compression is somewhat noticeable and measurable. This is modeled by the Motor Drive parameter." [52]

"Motor Drive will cause compression if not set to zero (as it models driver compression). Otherwise the cab block is completely linear and will not cause any compression." [53]

"Motor Drive simulates power compression due to voice coil heating." [54]

"Guitar loudspeakers are intentionally designed to compress. FRFR speakers do compress a bit but not nearly to the extent that guitar speakers do." [55]

"Makes edge-of-breakup tone stupid easy." [56]

"Speaker Drive models the magnetic compression (which is actually distortion) that occurs due to the nonlinear speaker excursion vs. applied voltage. Motor Drive models the change in power transfer due to heating of the voice coil. When the voice coil heats up the speaker sensitivity decreases, in some cases quite dramatically." [57]

"The thermal time constant of a typical guitar speaker is about 0.52 seconds. Magnetic time constants are zero." [58]

"So what I've done for the final release is put Motor Drive in BOTH the Amp block and the Cab block. If you're strictly FRFR then you can use the Amp block. If you are using a conventional guitar cab or a hybrid configuration (convention cab for monitoring and direct to FOH) then you can use the Cab block. Doing it in the Amp block also has the advantage that the speaker resonance information in the Amp block is used to calculate the frequency dependent heating whereas the Cab block uses a fixed set of data that is representative of a typical speaker. Finally I've made the time constant adjustable. I did some more calculations and measurements and found that a typical guitar speaker is actually lower than what I had previously calculated because thinner wire is used than I was assuming. Regardless you can now set the thermal time constant to get whatever response rate feels best. When using the Motor Drive in the Amp block it's before the output Level control so you don't have to worry about the behavior changing when you adjust the Level knob." [59]

"The actual value for a particular speaker is all over the map. The time constant is proportional to the mass and the thermal resistance of the voice coil. Both these values can vary widely. 200 ms is based on a typical theta of 1 degree C/W and a mass of 10g." [60]

"The formula is tau = M * C * theta where M = mass, C = specific heat of the voice coil material (typically copper) and theta = thermal resistance between the voice coil and the magnet gap." [61]


Mixes some of the "direct" signal entering the Cab block with the "Cab-processed" signal leaving the Cab block. This adds some "air" to the sound, which some users find to add realism to the tone.

It's not supported on the AX8.

AIR FREQUENCY lets you adjust the cutoff frequency of the mixed signal. Increase the Frequency to its maximum value for a straight mix.

Tip: if you want to listen to just the "air" part of the signal, select an empty user cab, and turn up Air.

Warning: adding AIR can cause phasing issues when using multiple IRs which are not aligned.

"Air is just clean signal mixed in. It WILL cause phase issues if the IRs are not minimum-phase or delayed." [62]

"There was a change to the Air stuff. I've been tempted to remove Air or change it to a shelving filter because it causes problems like this, especially with non-minimum phase IRs." [63]


This parameter sets a micro delay for stereo applications. When running a Cab block in Stereo mode, or when using two panned Cab blocks in parallel, delaying one side relative to the other can achieve interesting comb filter effects. A common practice in studio recording is to use multiple mics on a speaker at different distances to intentionally introduce comb filtering.

"My secret to realistic cab sounds is Delay. Use two IRs in stereo or two cab blocks and put a small amount of delay on one (using the Delay parameter in the Cab block). I like around 0.06 ms. You may like more or less. Producers experiment with placing mics at different distances to enhance the recorded guitar tones. This is the same as using a small amount of delay. Adding a bit of delay introduces some comb filtering which creates notches and peaks in the response which, in turn, adds a sense of "space" to the tone. Try it." And: "If you have any cab packs try mixing the "Back" IR with one of the regular IRs. I use more delay when doing this, 0.1 ms or more. I lower the level on the back IR by a couple dB. This gives a nice "in the room" open-backed cab sound."

Firmware Ares and later uses millimeters with the MIC DISTANCE parameter. Before that, milliseconds were used. To convert, multiply the old value in milliseconds by 343.

Firmware 17 for the Axe-Fx III and later increase Mic Distance to 3.4m (~11 feet). This allows delaying room mic IRs that have been trimmed to remove the leading silence. The alignment graph features a Zoom control that changes the abscissa between 3ms and 12ms.

"Mic distance is just a delay control, it doesn't alter the sound if you are using just one IR." [64]

"It's no different than using a delay block and dialing in a very short delay with mix = 100%." [65]

If you want to use this parameter but don't want IR coloring, use a FLAT or NULL IR. The last stock cab in the Axe-Fx III and FM3 is a flat IR.

(Matt) "The flat IR in the Axe-Fx III is NOT one that was downloaded. We created this file from scratch when we were working on a bass rig for one of our most celebrated endorsers. Intended uses include adjusting a DI and/or IR with Mic Distance, or combining DI and IRs into the same virtual preamp and room sim (which gets you a pretty great Nile Rodgers tone, by the way."

(GM Arts) "This is about mixing 2 signals: one without delay, and the other with a very short delay. 0.06ms is way too short to be perceived as a repeat; the effect is filtering caused by mixing these two signals. To keep things simple, we’ll apply an equal mix of the same signal and another delayed by 0.06ms. An easy way to experiment with this in the Axe-FX is with a Flanger block, with depth and feedback set to zero, and mix set to 50%. Adjust the delay to 0.06ms (not 0.6ms) to hear the effect with a mono signal. This produces a notched frequency response with complete signal cancellation just above 8KHz, with the -3dB point one octave lower at just over 4KHz. The signal is restored over the next higher octave (8KHz to 16KHz), but bear in mind that most IRs will not have much response there anyway, so this effect is mostly a blocking filter over the range 4KHz to 8KHz. So if you have a cab IR that has some response over this range, it will be perceived as a loss of some treble response. For many, this will remove harshness in a way that’s difficult to achieve with other filters. Others may find this effect too much. You can soften this effect by decreasing the delay and/or changing the mix ratio. Decreasing the delay raises the frequency at which this cut occurs. For example, a 0.05ms delay blocks response over the octave 5kHz to 10kHz. Lowering the mix % decreases the depth of the notch. Similarly, applying a delay to a different IR than the un-delayed block will “jumble” and reduce the final response to some extent. If you increase the delay (typically from 1ms and above), you’ll hear the combing effects as multiple notches become low enough to hear in the range of “guitar frequencies”. This sounds like a flanger or chorus without modulation, which shouldn’t be a surprise given we’re experimenting with a Flanger block. So why does this delay sound produce a tone more amp-like? Most players prefer their amp tone off-axis, meaning that they’re avoiding the direct harsh sound directly in front of the speaker, where high-frequencies are beamed. This filter simulates that effect. It’s also similar to standing slightly off-axis when using multiple speakers. Sound travels at roughly one foot per millisecond, so there is a very short delay between sound from different transducers. As Cliff stated, it also emulates recording techniques with mics placed at different distances from the cab. How to calculate? To find the frequency where this rolls-off high frequencies at -3dB, it’s simply: Hz = 1000 / 4 /delay in ms. So for 0.06 ms: 1000 / 4 / 0.06 = 4167Hz. Complete cancellation occurs at double this frequency, 8333Hz, and builds back to -3dB a double this frequency again, 16666Hz. Bear in mind that with higher delays, there will be audible effects from additional notches above this calculated frequency." [66]

DynaCab: Cab Type, Mic type, Position, Distance

Available in firmware 22.00 and later for the Axe-Fx III with the Cab block set to DynaCab mode.

These parameters let you select the speaker cabinet type, the type of microphone (condenser, dynamic, ribbon), the position (cap to cone) and the distance between mic and speaker.

Depending on a configuration setting in the Amp block, the Amp block automatically selects the correct Speaker Impedance Curve for the selected DynaCab in the first IR column of the Cab block.


The Cab block in firmware Ares and later displays an alignment graph showing a zoomed time series of the IRs. This allows visual adjustment of the mic distance. When using IRs that have not been minimum-phase processed, this facilitates aligning the IRs.

If you're unable to line up the IRs, you probably need to trim an IR to remove leading silence using Cab-Lab or the editor.

"I've added a time display to the cabinet block which shows all four IRs on the same axis and allows you to adjust the mic distance (delay) of each on that display. This allows you to precisely time align the IRs." [67]

"The graphs show the first 128 samples of the IRs. When engineers/producers record cabs in the studio they go to great pains to make sure the mics are time-aligned. A shift of just a few mm can make a big difference. In the virtual world we can do the same thing. The graphs show the time series and you can adjust the virtual distance of the microphones to change their time alignment." [68]

"Producers/Engineers often use two mics and mix them together to intentionally create phase cancellation. They'll move one mic away from the speaker a bit to get a longer delay which causes comb filtering when combined with the other mic. This is analogous to the Align tab on the Cabinet block." [69]

"It's analogous to moving the mic in and out." [70]

"Don't miss the fact (as indicated by the small text labels) that the data shown on the Cab Block's ALIGN page (and in Axe-Edit) is windowed from just 0 to 3 ms. That's the very beginning of a full UltraRes capture, "zoomed" so you can easily align the critical peaks in this region for close-miked IRs. If you are looking at the raw data for a "ROOM" or "FAR FIELD" IR, you will very likely have NOTHING on the ALIGN plot for that IR, because the sound will have taken longer than 3ms to reach the capturing microphone. Not that you'd ever need to align a room, but you will need to use Cab-Lab if you want to visually inspect the contents of such IR files." [71]

"You should align impulses ideally but sometimes a little misalignment adds character." [72]

"Another way to tame the high end is the studio trick of placing the mics at different distances. Use the Align page to do this. As you separate the IRs in time it will put a notch in the high frequencies." [73]


"The files are initially aligned to the same reference point -- a universal standard across all DynaCapture IRs. In other words, IRs for mics that are farther away have been time compensated to be aligned with IRs for mics that are closer. You can use the controls on the Align page in the usual way, however, to add time to any individual IR." [74]

Deliberately mis-aligning DynaCabs can add a desirable twist to the sound. See the video below.

Mic preamp and channel strip simulation

Microphone preamps, channel strips and tape can create pleasing musical "distortion". This might range from subtle “warmth” to full-on “nasty”. Mic preamps and channel strips also offer tone controls which change the sound. The Cab block in the Axe-Fx series and FM3 includes controls to simulate these effects. Not supported on the AX8.

Types include: Tube, Bipolar, FET 1, FET 2, Transformer, Tape 70us, Tape 50us, Tape 35us, Vintage, Modern and Exciter. The three TAPE types have different equalization time constants. [75]

Some of these are showcased in FM3 factory preset: Nasty Pre Slam.

The simulaton is switched on/off through PREAMP TYPE. Set it to NONE to switch it off and save CPU usage. [76]

If PREAMP TYPE is set to anything other than NONE, it's active. But it will have no impact on the sound if DRIVE and SATURATION are both turned fully CCW. PREAMP TYPE affects only DRIVE and SATURATION, not the tone controls. [77]

DRIVE controls the gain of the simulation. SAT controls the ratio of even/odd harmonics.

PREAMP MODE selects between ECONOMY and HIGH QUALITY. In High Quality mode, oversampling is employed to prevent aliasing which results in higher CPU usage. [78]

"It sets the oversampling rate for the preamp emulation." [79]

The preamp tone controls are the same as the "3 Band Console" type in the Graphic EQ block. [80] The tone controls are applied AFTER the selected preamp type. [81]

Important: LOW CUT, HIGH CUT and FILTER SLOPE are independent of the preamp. If the preamp simulation is switched off, LOW CUT and HIGH CUT are still fully operational.

"Probably not something you would use for clean sounds. A common technique for rock music is to push the pres, console, tape, etc. to varying degrees to get compression and "sparkle". The trick is getting just the right amount. Too much and it sounds raspy and nasty." [82]

"The VU meter shows the level into the pre. Select a pre Type and turn up the Drive. As the VU approaches the 0 dB marker you will begin to overdrive the pre." [83]

"0 on the VU meter indicates onset of clipping. It's not the same as your plug-ins in that regard. The problem with plug-ins is that you don't know where the onset of clipping is since the headroom isn't specified. Our way is superior since 0 dB indicates the point where things are clipping. The other way you have no idea where things start clipping. So 0 dB on the Axe-Fx is NOT equivalent to 0 dB on a typical plug-in." [84]

"I've done a lot of testing with isolation cabs. The big thing that happens is that the mic distorts, especially when using an SM57. This adds some crispness to the high end and some compression. I've found that I can duplicate that effect very closely by using the FET I preamp type in the Cab block and turning the Drive up until the desired compression is achieved. I set Sat to zero." [85]

"I always use a little bit of preamp drive in the cab block. All the venerated real preamps add distortion. There's nothing clean about them. It's the distortion that gives them character." [86]

"If you are trying to achieve amp > cab > compressor > tape, then you need to change the order around. A Drive block (to mimic tape distortion) after the Amp and before the Cab block sounds a bit dull. The tape distortion adds high end but putting before the cab block filters out those highs. The correct sequence would be: Amp > Cab > Compressor > Drive. The way I would do it is simply Amp -> Cab and use the compressor in the Amp block and one of the tape simulations in the Cab block." [87]

Forum discussion

Speaker Size

The Axe-Fx II allows the user to change the relative size of the virtual speaker. This is controlled through the parameter SPEAKER SIZE. This parameter appears only if the selected IR is not UltraRes and the Cab block’s mode is Mono.

Adjusting this parameter also changes the resonance of the selected cab.

This parameter was available only on the Axe-Fx II.

Microphone modeling

Miced cab.png

Available on which products

  • Axe-Fx III: no
  • FM3: no
  • FM9: no
  • Axe-Fx II: yes
  • AX8: no
  • FX8: no

About microphone modeling

Microphone modeling was available in the Cab block on the Axe-Fx II and AX8. It uses Impulse Responses.

Not supported on the FM3 and Axe-Fx III, except for the Proximity effect (see below).

(Axe-Fx II) "The mic models are actually IRs. The mic IR is convolved with the speaker IR to create a composite final IR." [88]

(Axe-Fx II) "If I were to design a Cab block today I wouldn't even include a Mic parameter. I NEVER use Mic simulation anymore. I simply find an IR I like and EQ as desired." [89]

(Axe-Fx II) "The mic options are mostly legacy. I never use them but if we took them out there would surely be much protestation." [90]

(Axe-Fx III) "What we found is that convolving a conventional mic IR with an IR obtained with a reference mic sounds nothing like capturing the IR with the conventional mic. The beam pattern of a reference mic is completely different than conventional mics. A reference mic is nearly omnidirectional whereas conventional mics have narrower beam patterns. In the far-field this wouldn't matter as much and "microphone modeling" might work. However in the near field this makes a huge difference and it simply doesn't work. Furthermore in the far field you want the response to be as neutral as possible so in this case there would be no desire for mic modeling anyways." [91]

Wicked Wiki article

Microphone models

57 DYN — based on the Shure SM57
58 DYN — based on the Shure SM58
421 DYN — based on the Sennheiser MD 421 II
87A COND — based on the Shure Beta 87A
U87 COND — based on the Neumann U87
E609 DYN — based on the Sennheiser e609 Silver
RE16 DYN — based on the Electro-Voice RE16
R121 COND — based on the Royer Labs R-121
D112 DYN — based on the AKG D112
67 COND — based on the Neumann U67

NULL — doesn't apply microphone coloring, but it enables the use of the PROXIMITY parameter
INVERT — inverts the signal, allowing for interesting effects in conjunction with the DELAY parameter
NONE — disables ALL mic processing in the Cab block, including proximity

The NONE and NULL types both disable mic coloring. A mic is still involved though, because IRs are always captured with microphones. Even when a neutral mic was used to capture the IR, such as an Earthworks mic. When capturing IRs, the mic is most often placed very close to the speaker, so the result is a close-miked tone. Still, selecting NONE is the best way to prevent adding additional EQ-ing to the tone.

Forum member Moke created Tone Matches of the mic models in the Axe-Fx II. Available here as IRs


Miced cab.png

This adjusts the proximity of the virtual mic to the virtual speaker. Higher numbers translate to the mic being closer to the speaker (near-field), causing an increase in low frequency response (more bass). Lower numbers translate to the mic being further away from the source, with the lowest number providing far-field coloration.

On the Axe-Fx II, proximity only works when a mic model has been selected, including the NULL type. The AX8 and firmware Ares do not support mic modeling, but do provide a PROXIMITY parameter.

The PROXIMITY FREQUENCY parameter lets you tune the frequency range over which the proximity effect occurs.

In DynaCabs mode, the Cab block doesn't provide a Proximity parameter. The Distance parameter already covers this.

Proximity in Wikipedia

Speaker-related parameters in the Amp block

H amps.png

The Amp block provides a number of parameters which are closely related to the Cab block, including: