Various presets & tricks

[Bakerman]

1. Hotel California - Verse 2

Generate 5 harmony parts with one guitar. Video/audio demo: https://forum.fractalaudio.com/threads/hotel-california-verse-2.135713/

Global scales

Four global scales need to be configured. If you want to use scales other than 1/2/3/4, Voice 1 Scale & Voice 2 Scale parameters will have to be adjusted in Pitch 1 X & Y states.

GS# 1  2  3  4

A   4  4 -4 -5
A#  3  3 -4 -4
B   3  3 -4 -5
C   3  5 -4 -3
C#  4  5 -4 -3
D   4  5 -5 -3
D#  3 -3 -5 -7
E   3 -3 -5 -7
F   3 -4 -5 -7
F#  3 -4 -3 -7
G   4 -4 -3 -7
G#  3  3 -4 -4

ADSRs

ADSR 1 & 2 thresholds should be low enough to trigger consistently on beat 1 of measures 9-15. If they're set too low, the first note of measure 9 may not always trigger properly. Check for that and adjust as needed, using the same threshold value for both ADSRs. ADSR response can be tested by watching Vol/Pan 1's volume parameter on front panel.

Vibrato

Bending certain notes more than 50 cents will make harmonies jump to wrong notes. Reduce vibrato width if this happens, or try a floating bridge for vibrato above/below resting pitch. Make vib. rate 8th triplet (3 cycles/beat) or so for smoothest transition to MTD trail in scenes 3-6.

Tempo

Preset is saved at original song's tempo, 74 BPM. Sequencer and delay used in scenes 5 & 6 are tempo-based. Tap the tempo somewhere earlier for best sync if playing without a recording/click.

Scene changes

A scene increment switch is ideal. Follow the chart here (play only full-size notes) for scene changes:

https://i.imgur.com/TDvK4mx.png

Scene 3 should be triggered slightly after the beat to avoid sending a short blip to MTD. Aim to be a 16th note late. Anywhere from a few ms to an 8th note is fine.

Scene 4 can be triggered anytime before playing the higher G# in the measure. (Scene 3 also gives the correct harmony on the low G#, and there's no glitch when changing.)

Scenes 5, 6, 7 should be triggered exactly on the beat. Scene 8 can be on the beat or slightly early if preferred while sliding out of previous note. Scenes 5 & 6 depend on the sequencer beginning in Scene 5, so if you stop playing while testing it may be out of sync when resuming. (Go back to scene 4 and play into scene 5 for proper results.)

 

[MrGuitarabuse]

Pure genius!!

 

[Bakerman]

2. Real transient detection

With envelope controller as ASDR threshold modifier (assignable w/ FracPad, adjustable in Axe-Edit afterward) the ADSR can be triggered with nearly any picked note, instead of only when crossing a fixed threshold.

The threshold needs to hover some distance above the current level and have its upward movement slowed enough for picked notes to cross it. Default modifier curve w/ offset around 8% and envelope attack around 15 ms works.

Possible uses:

Autowah/envelope filter effects

https://dl.dropboxusercontent.com/s/ldm1zeey5oxzuhq/ADSRwah.mp3

Reducing/removing pick attack with a short delay and ADSR volume control

https://dl.dropboxusercontent.com/s/g1fkk6c33xbi1uu/ADSRattacksoften.mp3

 

[Bakerman]

3. Stereo double-tracker

An ADSR configured like in post #3 can run a random LFO (at maximum 30 Hz rate) just long enough to change value once shortly after (most) picked notes, then stop. This can randomize parameters without creating any obvious modulation as notes sustain. Here ADSR 1 runs both LFOs that way.

LFO 1 randomizes pitch detune amount, Amp 1 input trim and Filter bypass. LFO 2 is quantized to 5 steps and selects exactly one multidelay line at a time, or dry signal by setting MTD mix to 0. This creates varying timing differences between the "dry" (MTD) and detune signal, which adds around 15 ms latency on average. MTD & Pitch are balanced L/R with Amp 1/2 input select L/R as well. Filter swaps L/R so neither channel spends more time ahead of the other on average.

The "wide" preset gets a slightly larger range of delay times by only using 3 MTD lines, and switching to Pitch voice 2 (with small additional delay) sometimes (50% chance) while MTD mix is 0. This keeps the average latency as low as possible, instead of delaying the pitch voice and using a wider range of times in MTD.

You might hear clicks from the MTD switching if trying this with cleaner tones or volume rolled back. With enough gain they don't seem very obvious.

Demo clip

0:00-0:34 = Double-tracker JVM, 0:35-end = Double-tracker wide

https://dl.dropboxusercontent.com/s/2o5cselyfafiehx/doubletrack.mp3

MTD modifiers overlaid to illustrate switching and 5 possible LFO values:

 

[MrGuitarabuse]

Nice! Gotta try these out. Thanks

 

[MKULTRA]

Very clever indeed. Thanks Bakerman!

 

[crg123]

3. Stereo double-tracker

An ADSR configured like in post #3 can run a random LFO (at maximum 30 Hz rate) just long enough to change value once shortly after (most) picked notes, then stop. This can randomize parameters without creating any obvious modulation as notes sustain. Here ADSR 1 runs both LFOs that way.

LFO 1 randomizes pitch detune amount, Amp 1 input trim and Filter bypass. LFO 2 is quantized to 5 steps and selects exactly one multidelay line at a time, or dry signal by setting MTD mix to 0. This creates varying timing differences between the "dry" (MTD) and detune signal, which adds around 15 ms latency on average. MTD & Pitch are balanced L/R with Amp 1/2 input select L/R as well. Filter swaps L/R so neither channel spends more time ahead of the other on average.

The "wide" preset gets a slightly larger range of delay times by only using 3 MTD lines, and switching to Pitch voice 2 (with small additional delay) sometimes (50% chance) while MTD mix is 0. This keeps the average latency as low as possible, instead of delaying the pitch voice and using a wider range of times in MTD.

You might hear clicks from the MTD switching if trying this with cleaner tones or volume rolled back. With enough gain they don't seem very obvious.

Demo clip

0:00-0:34 = Double-tracker JVM, 0:35-end = Double-tracker wide


MTD modifiers overlaid to illustrate switching and 5 possible LFO values:

Wow. I'll have to try this out when I get home. At first I thought the wide version was the emulated double tracking and the first part was true double tracking. Didn't realize until I saw both patches that they were both emulated double tracking.

Great job!

 

[simeon]

i was looking for a way to randomise a delay time for something i was working on the other day (a kind of infinite jets emulation). thanks for this solution, man!

oh wait...you need fracpad to do it? damn...

modifier control of adsr threshold has got to go on the wishlist...

 

[Bakerman]

oh wait...you need fracpad to do it? damn...

Only for initial modifier assignment. After that the hardware or AE Control: Modifiers menu can access them. Hardware won't jump to modifiers on control menu parameters this way though, so use AE then.

 

[hedua]

Thanks!!

 

[Bakerman]

4. Improved amp XY and preset switching

(Better no-compressor setup here: https://forum.fractalaudio.com/threads/various-presets-tricks.135770/page-2#post-1630877 This compressor version may still be useful when one or both amp types don't load as quickly as possible. The vol/pan method uses a timed volume boost/return configured to match the fastest possible switching time.)

The amp block does a long fade-in on XY and preset changes. More details here: https://forum.fractalaudio.com/threads/amp-xy-preset-change-fade-in-unnecessarily-long.133263/ No response from FAS on this yet.

One way to reach full volume faster is heavy compression. Place a compressor immediately after amp block and set as shown below. This can be used assuming you switch amp XY state via scene changes. A scene controller assigned to compressor bypass should be at 100% for all "Amp X" scenes and 0% for "Amp Y" scenes or vice versa. Then whenever amp XY state changes, the scene controller value changes and engages the compressor for ~200 ms. You might need to adjust compressor threshold and level for best results. Try aiming for around 16-18 dB gain reduction when hitting a chord, then (before assigning bypass modifier) adjust level for equal volumes engaged vs. bypassed.

This improves preset switching too. Instead of using a scene controller (limiting switching options) you can set up a group of 4 presets with comp bypass controlled by a different Manual ABCD knob in each. Set the source knob to 100% and the other three to 0% in each preset. Now you can switch freely among the 4 presets and the comp will engage each time. This works because if the new source knob was at 0% in the previous preset, the modifier/damping system will jump to that value then move to 100%. You could set up more preset groups like this, but switching between any two with the same ABCD source wouldn't engage the comp. If you're setting up to get through a specific piece this usually wouldn't be a problem. Once you need a fifth preset, choose a different knob from the most recent one and work from there.

To have both preset and scene switching options you can start with the ABCD assignments then (use FracPad) assign a scene controller to the chosen ABCD knob in any preset. Scene controller should follow amp state like before, starting at 100% in scene 1.

Here are 4 presets with the ABCD assignments and some random amp/cab choices. The "Comp switcher D" preset also has the scene controller - Manual D assignment and amp/cab XY set for scene 1-2 switching.


If you want to build new presets from these, just keep the compressor and check threshold/level as mentioned earlier. Level was a bit low so I increased row 2 in the output mixer. You may want to set that back to 0 dB.

 

[Hoosierdaddy]

1. Hotel California - Verse 2

Generate 5 harmony parts with one guitar. Video/audio demo: https://forum.fractalaudio.com/threads/hotel-california-verse-2.135713/

Global scales

Four global scales need to be configured. If you want to use scales other than 1/2/3/4, Voice 1 Scale & Voice 2 Scale parameters will have to be adjusted in Pitch 1 X & Y states.

GS# 1  2  3  4

A   4  4 -4 -5
A#  3  3 -4 -4
B   3  3 -4 -5
C   3  5 -4 -3
C#  4  5 -4 -3
D   4  5 -5 -3
D#  3 -3 -5 -7
E   3 -3 -5 -7
F   3 -4 -5 -7
F#  3 -4 -3 -7
G   4 -4 -3 -7
G#  3  3 -4 -4

ADSRs

ADSR 1 & 2 thresholds should be low enough to trigger consistently on beat 1 of measures 9-15. If they're set too low, the first note of measure 9 may not always trigger properly. Check for that and adjust as needed, using the same threshold value for both ADSRs. ADSR response can be tested by watching Vol/Pan 1's volume parameter on front panel.

Vibrato

Bending certain notes more than 50 cents will make harmonies jump to wrong notes. Reduce vibrato width if this happens, or try a floating bridge for vibrato above/below resting pitch. Make vib. rate 8th triplet (3 cycles/beat) or so for smoothest transition to MTD trail in scenes 3-6.

Tempo

Preset is saved at original song's tempo, 74 BPM. Sequencer and delay used in scenes 5 & 6 are tempo-based. Tap the tempo somewhere earlier for best sync if playing without a recording/click.

Scene changes

A scene increment switch is ideal. Follow the chart here (play only full-size notes) for scene changes:



Scene 3 should be triggered slightly after the beat to avoid sending a short blip to MTD. Aim to be a 16th note late. Anywhere from a few ms to an 8th note is fine.

Scene 4 can be triggered anytime before playing the higher G# in the measure. (Scene 3 also gives the correct harmony on the low G#, and there's no glitch when changing.)

Scenes 5, 6, 7 should be triggered exactly on the beat. Scene 8 can be on the beat or slightly early if preferred while sliding out of previous note. Scenes 5 & 6 depend on the sequencer beginning in Scene 5, so if you stop playing while testing it may be out of sync when resuming. (Go back to scene 4 and play into scene 5 for proper results.)

Amazed again

 

[Bakerman]

5. Enhancer bypass modifier

The enhancer block has no bypass mode parameter, so a bypass modifier can't be created in the normal way using front panel or Axe-Edit. It turns out that the expected sysex command (like any other bypass modifier except use enhancer block ID) will create one. Here's a preset that's just the enhancer with modifier assigned. Use Control: Modifiers menu of front panel or AE to adjust modifier settings.

 

[Bakerman]

6. Intelligent bender

Controlling a voice harmony value in pitch block's intel. harm. mode allows smooth bending between scale notes for pedal-steel type effects. Glide time value determines the bend rate, but higher values won't sound great the rest of the time. Notes will drift into pitch whether or not you want them to. The solution is a glide time modifier that stays low normally and moves to some higher value when using the bend controller.

With a switch, glide modifier can stay at max for most of its travel, with damping set so it returns to minimum just after the bend completes. Maximum glide time determines the required damping time.

https://dl.dropboxusercontent.com/s/8nxzln6o5i24jur/diatonic-bender.mp3

With a pedal, modifiers can be adjusted so pedal speed has some influence on bend speed. This uses a smoother arch-shaped curve with higher maximum and less damping for the glide modifier. The pitch won't begin changing until 50% in either direction, so this isn't quite like a whammy pedal harmony setting, where pedal position is directly tied to pitch. If you move the pedal at a steady rate it should seem like the bend takes about as long as the second half of the pedal sweep.

https://dl.dropboxusercontent.com/s/z7aagnojm5vqx5q/diatonic-bender-pdl.mp3

For more elaborate pedal steel impressions you could control up to 4 voices with up to 4 controllers. This also works with custom shift mode, changing voice scale instead of harmony value. The voice will bend between whatever shift values are set in the scales you're choosing.

These presets just have one voice set up like in the sound clips, Ext Ctrl 3 bending from +3rd to +4th in B major.

 

[Bakerman]

7. Twisted Sister - We're Not Gonna Take It

Pitch block & scenes to cover solo/harmony break while playing lowest part:

Scene 1 = Pitch off

Scene 2 = Pitch X, intel. custom scale

Scene 3 = Pitch Y, intel. E major 3rd/5th

Scene 4 = Pitch Y, harmonies to 6th/octave

Most of the harmony section is some close voicing of an E or F#m chord. A custom 6-note scale containing both chords (E F# G# A B C#) works for those with harmonies at 3 & 5.

The first few notes where that scale won't work are B G#m A chords, all root position, so E major 3rd/5th (Y state) can be used. Selecting scene 4 makes scene controller 1 change harmonies to 6th & octave.

For the A chord near the end, Ext Ctrl 3 is set up to change pitch key from E to A, which will get the right harmony. A momentary switch would be ideal.

Scene changes can occur throughout the dashed line in chart where present, up to the next note. This is because playing E or F# gets the same harmony from either XY state.

For easier switching scene increment could be used instead of direct scene selection. Eight scenes are enough to get through the solo including no-harmony scenes before/after.

 

[yek]

Cool tricks, Bakerman.

Perhaps post the compression trick in the AX8 forum section as well, to draw more attention to it?

 

[Bakerman]

Cool tricks, Bakerman.

Perhaps post the compression trick in the AX8 forum section as well, to draw more attention to it?

Do you know if the gap/fade is about the same on the AX8? If it's a longer gap (zero output) and more reasonable fade time, the compressor might not help, or might need to stay engaged longer.

It would be nice to get an example XY switching clip to know if the settings above are close to ideal for AX8.

 

[yek]

Attached: switching Amp block on AX8 between X and Y (using top panel), with similar Amp settings.

 

[Bakerman]

Attached: switching Amp block on AX8 between X and Y (using top panel), with similar Amp settings.

Thanks. The gap there looks about equal to the III. Fade-in seems slower than the III but much better than the II, and can probably use a lower comp ratio.

Want to test scene 1/2 switching of this? Hitting around -5 dB on comp meter should be enough to make a good ~15 ms ramp to full level. Try aiming for that by adjusting threshold first. Increase ratio & tweak threshold again if the improvement doesn't seem very noticeable.

 

[yek]

Sure, will do, asap.

BTW, your measurement of the III may be outdated, when it comes to the speed of channel switching. I'm not sure though. Maybe someone can post a clip of the current state of things.