Sorry, I was thinking about gain knobs. I’ll do my tests then. Thanks!
I think the problem is in the pregain knob. For a clean sound I have to turn off drive channel and put pregain at minimum.
So pregain and drive on/off are the only values I change and the spike is there.
So plugins should declare the order of parameters loading (so that MOD platform knows which one to load first) since this is crucial somehow, right?
Hi @redcloud
I’ve tested the mutant mimicking your settings and notice that it exists a big volume jump ( 34 dB ) between the clean sound and drive sound so I imagine that it is expected that this change can be heard when abruptly changing from one snapshot to another.
Nothing was connected on the input of the device so the level meter plugin is showing only the device noise level.
I´m not sure if my settings here are in line with your setup though.
Clean (mutant output peak hold at -84.4 dB)
Drive (mutant output peak hold at -50.6 dB)
If you are on the Dwarf I know that device has a noise gate at the input and a compressor at the output, so the fact that you are hearing a spike could be related to these settings. I Hope this help you troubleshoot the problem.
The yellow vertical line is the sample start
The red lines are the CLEAN snapshot
The green lines are the LEAD snapshot
You can see that there are some spikes on the yellow line caused by the audio player looping the sample.
The thing to notice here is the blue rectangle. When changing from LEAD to CLEAN, a 2-steps spike happens: a first huge spike followed by a lower step then followed by normal CLEAN level.
I have builtin noise gate level at minimum (-70 db) and no builtin compressor active. Also ground loop compensation is off.
Could you test it with the sample Sine_wave_440.wav here and make a recording to check the waveform?
This is the setup I’ve used.
No matter what’s the volume difference between two snapshots, I expect a clean/straight behavior.
If CLEAN snapshot is at -10db and LEAD is at -20db, I expect that switching from one to the other in A/B loop test should act like this
A(-10)…B(-20)…A(-10)…B(-20)…A(-10)…B(-20)…A(-10)…B(-20)…
but what I see is something like this
A(-10)…B(+10…0…-20)…A(-10)…B(+10…0…-20)…A(-10)…B(+10…0…-20)…
Thanks a lot! This looks less evident than my case anyway. Could you share the waveform?
Just noticed that also something messy is happening with the Mutant user presets. I was unable to delete some of them I created in past days and I noticed that when switching from CLEAN to LEAD snapshot, all the knobs were moving in a fraction of a second then moving to right position. This would explain the A(-10)…B(+10…0…-20)…A(-10)…B(+10…0…-20)…A(-10)…B(+10…0…-20)… behavior.
Rebooted the Dwarf and then I was able to remove the presets. Now the glitch is barely audible. Maybe user preset management is broken for Mutant?
Is there a way to automatically clean snapshots.json file in order to remove all the referencer to presets and plugins that do not belong to the pedalboard? Don’t know why but I see references to stuff that are not present in my pedalboard
I’ve recorded the glitch (audio only) with a white noise and a wave generator.
What we are hearing it’s indeed an audio spike followed by a sound gap as the image of the sound wave shows.
White noise
Wave genarator
Hello @redcloud ,
did you try to reload the browser page after creating/deleting a snapshot?
Updated plugin settings are saved when a snapshot or pedalboard is saved, but the GUI in the browser will not show the updated settings.
At least not for me MacOS 10.14 and Safari/ Opera…
Hope that helps regarding the presets you are changing.
Greetings and God bless, Marius
Yes, tried several times but in the end I found that the bug 1 was due to “undead” user presets I was unable to delete from Mutant user presets list .
Now bug 2 still remains.
The sound gap is reasonable (even though gaps have all different volume) while the spikes aren’t!
I expect that gaps are all same volume and duration (A) but we have spikes and gaps with different volume (B)
Hello @redcloud
Analysing the wave forms again I noticed that the spike only happens when switching from the snapshot with the low volume on the pre-gain parameter to the snapshot with the high volume on the pre-gain.
On the opposite way the spike is not there.
That gave an Idea that I’ll test later that is set the pre-gain volume at its lowest possible setting that is different from zero.
Ok… now I’m confused.
I have tried to reproduce the glitch and now there is no spike happening. I’ve tried several times with different very low and zero volume pre gain settings and with different generators (pink and white noise and tone generators at 440 Hz).
Resultant waveforms was always like the one below.
Thanks for the detailed reports guys!
I had a look as well and sadly there is not much the plugin can do.
The issue seems to be on the mod’s side not handling snapshots too gracefully.
The order of parameters changed seem random and on top of that, not all changes are applied on the same frame, resulting in these jumps
I already reported the bug on mod’s side.
Thanks again for reporting.
Sorry to distract from troubleshooting discussion, but just wanted to mention that I’ve changed my current preamp emulation in my main pedalboard to Mutant, and I really like the resulting sound for my melody chapman stick strings. I am no way proficient in designing sound chains, so not sure if my opinion really counts, but I would like to thank authors for providing such a plugin, feels like it made my music life a bit better.
I am new to the Dwarf and this AMP really is one of my reasons to keep the Dwarf despite the insolvency.
This amp still is the best one out there for my heavy metal/trash band sound.
It is the amp that has the best favorable tone / amount of tweaking ratio for me.
Though, it has me bummed out over the CPU usage on my Dwarf
As soon as I try to combine with a few other plugins, I get cuts of the sound due to the CPU maxing out.
No, the other effects aren’t CPU heavy reverbs or a ton of routing options and when I turn the plugin off, I go back to the lower levels of pc usage I’m used to (100%-> 40 à 50%)
I’ve made setups with other amp modellers using both inputs for guitar and vocal, each with their reverbs, deays etc and have it all running smoothly.
So yes, I did enough tests to isolate the Mutant as the cause of the issue.
(The only other plugin to do it this hard that I encountered is the RMPro ‘mastering’ plugin), which is also unusable for me because I can’t add it to my setups.
I as confronted with this issue again yesterday, trying the MOD assistant . One of the things he proposed was a simple 3 plugin setup with the Mutant and a reverb. The reverb was pointed out as probable cause by a users but I disagree.I have less issues with load with that particular reverbthan with the Mutant.
What strikes me is that I seem to be th only one, which would mean either nobody uses it actively or only the large Duo users use it, without cpu issues?
(or I have a specific fault in my cpu?)
Is there something I can do?
if anybody wants to run spcific tests, let me know?
Could the Mutant be optimized?