Bug? Volume jumps in certain presets

Chromatizm

Chromatizm

Power User
FW 6.00, but I've experienced this in other FWs.

When switching scenes, sometimes there are volume jumps and additional sounds like the previous scene is active for some time after you've switched scene, and the target scene sound starts with a certain delay. Not all amps/preset display the issue. Attached preset is Tim Pierce's G.O.T. FM9 (not FM3) preset with amps replaced for Buttery, which has this issue when switching between scenes 1-2-3. For example, when switching from 3 to 2, it seems like the volume increased for a fraction of a second while not resetting the "boost" setting, which results in a short loud sound, then the volume and the setting for the boost switch of a target scene are applied and it goes back to normal.
If you will set the Input boost to "on" on the scene 2, this behavior will expand when switching from the scene 2 to 1 too.

I use Input Boost and Saturation switches is all my presets, but not all of them are subject to this. I feel like the original preset has a similar issue between scene 1 and 3, but less pronounced.
I assumed Spillover might affect this, but there is no difference between "All" and "Off" setting.
 

Attachments

  • Volume jumps - FAS Buttery (edge of break-up).syx
    24.1 KB · Views: 10
Easy to reproduce with this setup (new preset): Channel A FAS Buttery, default settings, amp volume -12dB.
SC1: Input Trim 0.5, Output scene 1 volume: 20dB
SC2: Input Trim 1, Input boost 12dB, Saturation 4db, Output scene 2 volume: 0dB
SC1 is louder, but not by much.

When switching from SC1 to SC2, there is no jump in volume.
When switching from SC2 to SC1, there is a huge jump in volume at the beginning, but then the volume "normalizes". Sounds like the output volume "opens" to its maximum value, and then the target value applies.

If SC1 Output block volume is set back to 0, the volume jump when switching from SC2 to SC1 is still there, just not that apparent. The SC1 in the case is so quite, that the volume jump is the only thing you hear when switching.

If there is a Delay in the chain, its tails will also have a huge boost for the first fraction of a second, then they will go to the expected target volume.
 
That's not a bug IMHO.

The preset has an overload of controllers attached to Amp block parameters.
That seems just too much for the FM3's processing power to handle during the few milliseconds inbetween scenes.

* I checked: the FM9 and Axe-Fx III handle this without hiccups.
 
Last edited:
I pulled both Pierce's original FM3 and FM9 presets into my FM3, and they work without problems.

I iterated through each modifier assignment in the Amp block, removing one, testing to see if it made a difference, resetting the preset, and moving to the next. The CS 2 on Input Boost appears to be the problem. Why? I dunno.

Perhaps it's a bug, perhaps that many modifier assignments reached a tipping point. Perhaps @Admin M@ has some insight.
 
For now, just remove that modifier if it's causing you any trouble (and the rest of them). If I recall correctly, Tim also worked with @austinbuddy who often adds these controls, but Tim doesn't use them on his FM9 IIRC.
 
Greg Ferguson said:
I pulled both Pierce's original FM3 and FM9 presets into my FM3, and they work without problems.
Click to expand...
Sorry that I didn't make it clear - I took Tim's preset (because I like it very much), and made it a "template" for my own presets. That's where I discovered steps to reproduce - when I assign control switches, the issue presents itself.

Tim's presets are totally fine as is - the issue is reproducible when you attach modifiers!

Admin M@ said:
For now, just remove that modifier if it's causing you any trouble (and the rest of them).
Click to expand...
The G.O.T. preset "as is" is totally fine, no complaints. Nothing to fix about it!

As I clarified in this post, you can create the same problem with another preset.

I've checked it right now with the "Rotary Tapeverb" factory preset (CPU 45% by default). Removed all blocks except for Amp and Cab. Copied Amp channel A to B, set "A" for Scene 1, "B" for Scene 2. CPU usage is 25% in this setup.

Assigned CS1 to Input Trim, CS2 to Input Boost, CS3 to Saturation.
Set CS1 "on" for Scene 1, the rest are "off"
Set CS2 "off" for Scene 2, CS2&3 are "on".
Add a little volume for Scene 1 to match it with the Scene 2 volume. Switch between scenes, witness the issue!
 

Attachments

  • Volume jumps - Rotary Tapeverb control switch.syx
    24.1 KB · Views: 4
Last edited:
I have a preset where CPU load changes from 72% to 80% when switching scenes. Can it be connected? I never saw this before. Spikes when switching - yes, but not 72% CPU on one scene, 69% on a second one, 80% on the third one. Same AMP and CAB channel combo is used. Scene #5 has different amp, same cab, 75% CPU.
Should I share that preset or it is known and expected?
 
Chromatizm said:
I have a preset where CPU load changes from 72% to 80% when switching scenes. Can it be connected? I never saw this before. Spikes when switching - yes, but not 72% CPU on one scene, 69% on a second one, 80% on the third one. Same AMP and CAB channel combo is used. Scene #5 has different amp, same cab, 75% CPU.
Should I share that preset or it is known and expected?
Click to expand...
CPU load within a preset can be impacted by a number of effects types depending on the settings of those effects.

Look for blocks that are changing channels on scene changes.

Reverb and Drive are two common effect types that can vary a fair bit. Some Drive block types might use only 3-4% where others could be 8-10%.
 
I have notices the same thing switching from an overdrive scene with one kind of reverb in block, A to a clean scene with another kind of reverb and Block B, and I get that volume bump and you hear the Reverb, and then it disappears into the clean scene. 45% cpu on fm9
 
i Should note that this only happens as Chromatizm describes only when chaining from gain scene to clean scene. Reverse is fine for some reason.
 
You must log in or register to reply here.
Back
Top Bottom