Add Disintegrator effect to LMMS #5161

LostRobotMusic · 2019-08-30T02:11:43Z

(Made this whole thing in like 30 minutes, haha)

I'm not going to try to hide it, this effect is a blatant ripoff of Ableton's Erosion. You could probably even call it a clone. That was less because I intended to copy them and more because I couldn't think of anything else I'd want to add.

Disintegrator (and Erosion) is an effect that delays the sound with a modulated delay in super aggressive ways to get some very digital-sounding distortion. The first mode modulates the delay with white noise, which is filtered by the two filtering knobs before it does any modulation. The second mode is identical to the first, except the white noise is stereo rather than mono, which sounds much wider. The third mode uses a sine wave as the modulator, which sounds very similar to FM and can result in some very cool timbres.

Notice that the two filtering knobs are replaced with a frequency knob when the sine wave mode is selected.

This effect is quite basic and doesn't make any changes to other parts of LMMS, so I'm assuming the review-and-merge process will go by fairly quickly.

The effect is also 100% sample exact. I think. If I did that part right.

(And sorry for commit mess, Squash-And-Merge should prevent that from being a problem)

Master

LostRobotMusic · 2019-08-30T02:33:24Z

I just realized the code is quite lacking in comments. I'll fix that.

LostRobotMusic · 2019-08-30T02:45:48Z

Alright, I added comments to the code and also removed the "detuneWithOctaves" function which I accidentally left in there. Should be ready now.

plugins/Disintegrator/Disintegrator.cpp

plugins/Disintegrator/Disintegrator.h

plugins/Disintegrator/Disintegrator.cpp

plugins/Disintegrator/DisintegratorControls.cpp

…last commit).

LostRobotMusic · 2019-08-30T17:37:10Z

An absolute galaxy brain (known as Oncill) suggested that I add a new mode that modulates the delay time depending on the audio input. I tested it out. The result can work as an aggressive exciter, as well as something that (although is very violent and digital) is unnaturally good at enhancing the general timbre of any instrument you feed it. The filters are still enabled for this mode, which filter the input audio modulator (but not the "carrier", if that's even the correct term in this case). The new mode was added in the latest commit.

plugins/CMakeLists.txt

LmmsBot · 2019-11-24T00:44:55Z

🤖 Hey, I'm @LmmsBot from github.com/lmms/bot and I made downloads for this pull request, click me to make them magically appear! 🎩

Linux

Linux (AppImage): lmms-1.3.0-alpha.1.79%2Bgfb5471d-linux-x86_64.AppImage (build link)

Windows

Windows 64-bit: lmms-1.3.0-alpha.1.79%2Bgfb5471d17-mingw-win64.exe (build link)
Windows 32-bit: lmms-1.3.0-alpha.1.79%2Bgfb5471d17-mingw-win32.exe (build link)

🤖

{"platform_name_to_artifacts": {"Linux": [{"artifact": {"title": {"title": "(AppImage)", "platform_name": "Linux"}, "link": {"link": "https://11785-15778896-gh.circle-artifacts.com/0/lmms-1.3.0-alpha.1.79%2Bgfb5471d-linux-x86_64.AppImage"}}, "build_link": "https://circleci.com/gh/LMMS/lmms/11785?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link"}], "Windows": [{"artifact": {"title": {"title": "64-bit", "platform_name": "Windows"}, "link": {"link": "https://11788-15778896-gh.circle-artifacts.com/0/lmms-1.3.0-alpha.1.79%2Bgfb5471d17-mingw-win64.exe"}}, "build_link": "https://circleci.com/gh/LMMS/lmms/11788?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link"}, {"artifact": {"title": {"title": "32-bit", "platform_name": "Windows"}, "link": {"link": "https://11787-15778896-gh.circle-artifacts.com/0/lmms-1.3.0-alpha.1.79%2Bgfb5471d17-mingw-win32.exe"}}, "build_link": "https://circleci.com/gh/LMMS/lmms/11787?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link"}]}, "commit_sha": "88ea50dd3e2812efd1d71dc1158de5691eb0a098"}

LostRobotMusic · 2019-11-24T00:58:42Z

Don't test this out yet, I found some weird bug with the self modulation mode that isn't appearing in my copy for some reason...

LostRobotMusic · 2019-11-24T02:18:21Z

Alright, all merge conflicts and bugs have been removed. Feel free to test now.
Here's a video of me messing around with this effect: https://youtu.be/nzUV3gDL4Wo

LostRobotMusic · 2020-04-19T14:23:21Z

I just realized this effect sounds different when the sample rate is changed. Fixed it in the latest commit. 👍

plugins/Disintegrator/Disintegrator.cpp

PhysSong · 2020-06-08T02:31:16Z

@douglasdgi REMINDER. Did you see my reviews?

PhysSong · 2020-12-10T04:20:38Z

@douglasdgi Is this ready for review?

LostRobotMusic · 2020-12-10T04:41:12Z

@PhysSong I was planning on rewriting this, it's incompatible with sample rate changes and isn't written very well in general. It shouldn't take more than a few days to rewrite (maybe even just one day, it's a fairly simple plugin), I'll get to that as soon as I'm done with the Phaser changes I've been working on.

LostRobotMusic · 2020-12-27T18:45:47Z

@PhysSong I was wrong, turns out I did make it compatible with sample rate changes, just not in a nice or neat way...
Anyway, the whole code was kind of a mess and it turns out it was super buggy as well. The last commit fixes everything. The filters should actually work correctly now, and the Self Modulation mode should sound significantly cleaner.

Also, the Amount knob is now in units of milliseconds like it should have been, rather than samples like it was before. Should be ready for review now.

plugins/Disintegrator/DisintegratorControlDialog.cpp

curlymorphic

A nice plugin, seam to get a lot for your money as they say. It is really simple and effective. I have left some comments in the code, I feel these are minor amendments, leading to a substantial improvement.

My other concern is the name "Disintegrator", Lmms already comes bundled with a LADSPA effect of the same name. Maybe now would be a good time to reconsider the name, as a name change would not be viable once release.

I have kept my review to the audio processing code, I have not reviewed any UI components, or comment on style.

curlymorphic · 2020-12-31T16:53:11Z

plugins/Disintegrator/DisintegratorControls.cpp

+	m_highCutModel(20000.0f, 20.0f, 20000.0f, 0.01f, this, tr("High Cut Frequency")),
+	m_amountModel(1.6f, 0.0f, 10.0f, 0.001f, this, tr("Amount")),
+	m_typeModel(this, tr("Type")),
+	m_freqModel(100.0f, 1.0f, 21050.0f, 0.01f, this, tr("Frequency"))


Where have these range values originated from? The lower end of 100Hz seems reasonable if that suits you desired effect, the upper limit seems a strange value, I can see this is not limiting to the audio spectrum or any common fs/2

100 is the default value, not the lowest value. The lowest value is 1. The highest value... I think that's a typo. Pretty sure I meant 22050.

curlymorphic · 2020-12-31T16:56:52Z

plugins/Disintegrator/DisintegratorControls.cpp

+	m_effect(effect),
+	m_lowCutModel(20.0f, 20.0f, 20000.0f, 0.01f, this, tr("Low Cut Frequency")),
+	m_highCutModel(20000.0f, 20.0f, 20000.0f, 0.01f, this, tr("High Cut Frequency")),
+	m_amountModel(1.6f, 0.0f, 10.0f, 0.001f, this, tr("Amount")),


Depth may be a more suitable name for this control.

curlymorphic · 2020-12-31T17:29:13Z

plugins/Disintegrator/Disintegrator.h

+
+	void sampleRateChanged();
+
+	inline float realfmod(float k, float n);


What is the reason this function is declared inline?

curlymorphic · 2020-12-31T18:06:26Z

plugins/Disintegrator/Disintegrator.cpp

+{
+	m_sampleRate = Engine::mixer()->processingSampleRate();
+	m_2PiOverSR =  F_2PI / m_sampleRate;
+	m_lp = new BasicFilters<2>(Engine::mixer()->processingSampleRate());


The necessity for the creation of a new filter whenever the sample rate changes is an unfortunate historical requirement. This pull request creates new filters but does not yet handle the disposal of any existing filters, thus causing a memory leak.

I am suggesting using std::unique pointers to solve this. lmms has a handy function, that did not make it into the c++ standard until c++14 make_unique. if you update the two lines of code to :

m_lp = make_unique<BasicFilters<2>>(Engine::mixer()->processingSampleRate()); m_hp = make_unique<BasicFilters<2>>(Engine::mixer()->processingSampleRate());

and include stdshims.h

The declaration in the header would need to be amended to

std::unique_ptr<BasicFilters<2>> m_lp; std::unique_ptr<BasicFilters<2>> m_hp;

We recently switched to C++14, FYI.

curlymorphic · 2020-12-31T18:09:48Z

plugins/Disintegrator/Disintegrator.h

+
+	float m_sineLoc = 0;
+
+	BasicFilters<2> * m_lp;


please see comments in cpp file, these could be unique pointers

std::unique_ptr<BasicFilters<2>> m_lp; std::unique_ptr<BasicFilters<2>> m_hp;

curlymorphic · 2020-12-31T18:23:43Z

plugins/Disintegrator/Disintegrator.cpp

+	m_lp = new BasicFilters<2>(Engine::mixer()->processingSampleRate());
+	m_hp = new BasicFilters<2>(Engine::mixer()->processingSampleRate());
+	m_lp->setFilterType(0);
+	m_hp->setFilterType(1);


Would it be an improvement if the filter types were selectable, like in the Dual Filter effect, this would then lead to the addition of resonance and possibly gain controls, and may extend the pallet of sounds?

If the filter types are to remain constant then the use of the constant types 0 and 1 should be changed to BasicFilters::FilterTypes::lowpass and the respectave hi pass

curlymorphic · 2020-12-31T18:54:59Z

plugins/Disintegrator/Disintegrator.cpp

+				}
+				else// For when the interpolation wraps around to the beginning of the buffer
+				{
+					s[i] = m_inBuf[i][m_bufferSize - 1] * (1 - newInBufLocFrac[i]) + m_inBuf[i][0] * newInBufLocFrac[i];


see above comment

curlymorphic · 2020-12-31T18:59:25Z

plugins/Disintegrator/Disintegrator.cpp

+		float newInBufLoc[2] = {0, 0};
+		float newInBufLocFrac[2] = {0, 0};
+
+		// Generate white noise or sine wave, apply filters, subtract the


might be preferable to just call it noise, not White Noise. White Noise has defined mathematical properties, that fast_rand() does not provide.

curlymorphic · 2020-12-31T19:13:14Z

plugins/Disintegrator/Disintegrator.cpp

+
+		sample_t s[2] = {buf[f][0], buf[f][1]};
+
+		// Increment buffer read point


when implementing a circular buffer, it is common to have the buffer size a power of 2, with 512 being the next closest to the 500 you stare in the header.

The index increment and wrap at end of buffer can then be reduced to:

++m_inBufLoc; m_inBufLoc &= (m_bufferSize - 1)

This is preferable as conditional branches can be processor expensive, and may also hinder some compiler side optimisations.

curlymorphic · 2020-12-31T19:22:03Z

plugins/Disintegrator/Disintegrator.cpp

+	}
+	if(m_needsUpdate || m_disintegratorControls.m_lowCutModel.isValueChanged())
+	{
+		m_hp->calcFilterCoeffs(m_disintegratorControls.m_lowCutModel.value(), 0.5);


The updating of the filter coefficients should be done per frame using the value buffer, not once per buffer.

The current implementation is block size dependant, so the audio will be different depending on the block size. This is not desirable, and the effect should remain independent.

One of the drives for sample exactness was to remove the buzzing and zipping noises, these are caused by sudden stepped changes in parameters on a per-block basis, and are removed by smoothly updating the parameters on a per-frame basis.

rdrpenguin04 · 2021-01-03T23:10:50Z

Douglas is offline for an indeterminate amount of time. He has, in the past, requested his unfinished PRs to be closed pending his ability to work on them more. I will now forward that request again, since he won't be here to close the PR for potentially a very long time.

LostRobotMusic · 2021-01-15T18:06:44Z

@rdrpenguin04 I definitely did not request that... I did request something similar for Microwave, but for everything else I'd much rather keep them open to give others the chance to finish them if they're able to.

(And I'm back for the time being by the way, I should be able to finish this one up)

rdrpenguin04 · 2021-01-15T18:08:05Z

@rdrpenguin04 I don't actually recall requesting that... I did request something similar for Microwave, but for everything else I'd much rather keep them open to give others the chance to finish them if they're able to.

(And I'm back for the time being by the way, I should be able to finish this one up)

Sorry about that; I had assumed you had meant all of your PRs.

tresf · 2025-01-28T03:28:41Z

My other concern is the name "Disintegrator", Lmms already comes bundled with a LADSPA effect of the same name. Maybe now would be a good time to reconsider the name, as a name change would not be viable once release.

Yes, the CMT plulgins ship one under the same name and this is a good observation but I think the same argument can be made for names like "Reverb", "Eq" and "Compressor", so duplicate names are OK.

I asked AI for some better names given Lost's description and none of them were as good as "Erosion" and "Disintegrator".

LostRobotMusic and others added 5 commits June 21, 2019 15:09

Merge pull request #8 from LMMS/master

7f1c954

Master

Merge pull request #11 from LMMS/master

0bb50a8

Master

Merge pull request #12 from LMMS/master

a332d47

Master

Add Disintegrator effect

e83d44a

Add Disintegrator effect

a630f32

LostRobotMusic added needs code review A functional code review is currently required for this PR needs style review A style review is currently required for this PR labels Aug 30, 2019

Added comments, removed unnecessary detuneWithOctaves function.

2fce224

PhysSong reviewed Aug 30, 2019

View reviewed changes

LostRobotMusic added 4 commits August 29, 2019 21:08

Resolve code/style reviews

6a44a51

Changed filters to use BasicFilters.h, fixed crashing bug (caused by …

5a4670e

…last commit).

Increase buffer size to correct amount

328c765

Add self-modulation mode

6b09fce

LostRobotMusic added 2 commits August 30, 2019 12:00

Clip modulator at 0db for self-modulation mode

4556151

Minor CPU boost I missed

02b9a2e

PhysSong added needs testing This pull request needs more testing and removed needs style review A style review is currently required for this PR labels Sep 15, 2019

zonkmachine reviewed Nov 9, 2019

View reviewed changes

plugins/CMakeLists.txt Outdated Show resolved Hide resolved

Merge branch 'master' into disintegrator

ee442e8

Hippity hoppity squash and merge hides this probably

d2cb54f

Fix knob ranges

06024b8

LostRobotMusic added 3 commits April 18, 2020 14:43

Resolve code/style reviews

89f0501

Put some duplicate code into for loops

624b2f2

Remove newline in untouched file

7e231b8

PhysSong self-requested a review April 19, 2020 02:42

Make effect work correctly with sample rate changes

ee0420a

PhysSong reviewed Apr 23, 2020

View reviewed changes

ryuukumar mentioned this pull request Jul 21, 2020

Reorganising LMMS, renaming editors and classes and everything in between (cont. #120) #5592

Closed

56 tasks

Apply suggestions from code review

fa0f436

Fix stuff

1235e41

Fix Amount knob units

5665902

PhysSong reviewed Dec 28, 2020

View reviewed changes

plugins/Disintegrator/DisintegratorControlDialog.cpp Outdated Show resolved Hide resolved

LostRobotMusic added 2 commits December 27, 2020 20:05

Fix more stuff

5384c5d

Optimize

88ea50d

curlymorphic requested changes Dec 31, 2020

View reviewed changes

LostRobotMusic added the rework required Pull requests which must be reworked to make it able to merge or review label Jan 2, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add Disintegrator effect to LMMS #5161

Add Disintegrator effect to LMMS #5161

LostRobotMusic commented Aug 30, 2019 •

edited

Loading

LostRobotMusic commented Aug 30, 2019

LostRobotMusic commented Aug 30, 2019

LostRobotMusic commented Aug 30, 2019

LmmsBot commented Nov 24, 2019 •

edited

Loading

Linux

Windows

LostRobotMusic commented Nov 24, 2019

LostRobotMusic commented Nov 24, 2019

LostRobotMusic commented Apr 19, 2020

PhysSong commented Jun 8, 2020

PhysSong commented Dec 10, 2020

LostRobotMusic commented Dec 10, 2020

LostRobotMusic commented Dec 27, 2020 •

edited

Loading

curlymorphic left a comment

curlymorphic Dec 31, 2020

LostRobotMusic Jan 24, 2021

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

PhysSong Jan 4, 2021

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

curlymorphic Dec 31, 2020

rdrpenguin04 commented Jan 3, 2021

LostRobotMusic commented Jan 15, 2021 •

edited

Loading

rdrpenguin04 commented Jan 15, 2021

tresf commented Jan 28, 2025 •

edited

Loading


		void sampleRateChanged();

		inline float realfmod(float k, float n);


		sample_t s[2] = {buf[f][0], buf[f][1]};

		// Increment buffer read point

Add Disintegrator effect to LMMS #5161

Are you sure you want to change the base?

Add Disintegrator effect to LMMS #5161

Conversation

LostRobotMusic commented Aug 30, 2019 • edited Loading

LostRobotMusic commented Aug 30, 2019

LostRobotMusic commented Aug 30, 2019

LostRobotMusic commented Aug 30, 2019

LmmsBot commented Nov 24, 2019 • edited Loading

Linux

Windows

LostRobotMusic commented Nov 24, 2019

LostRobotMusic commented Nov 24, 2019

LostRobotMusic commented Apr 19, 2020

PhysSong commented Jun 8, 2020

PhysSong commented Dec 10, 2020

LostRobotMusic commented Dec 10, 2020

LostRobotMusic commented Dec 27, 2020 • edited Loading

curlymorphic left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rdrpenguin04 commented Jan 3, 2021

LostRobotMusic commented Jan 15, 2021 • edited Loading

rdrpenguin04 commented Jan 15, 2021

tresf commented Jan 28, 2025 • edited Loading

LostRobotMusic commented Aug 30, 2019 •

edited

Loading

LmmsBot commented Nov 24, 2019 •

edited

Loading

LostRobotMusic commented Dec 27, 2020 •

edited

Loading

LostRobotMusic commented Jan 15, 2021 •

edited

Loading

tresf commented Jan 28, 2025 •

edited

Loading