Gonzalo

I was wondering if there is some kind of way of connecting HI to control RNBO parameters in a Raspberry Pi. My idea is to be able to make a voice module for eurorack that can be controlled by gamepad. Thank you for your help!

Hello,
I was wondering if there is some kind of way of connecting HI to control RNBO parameters in a Raspberry Pi. My idea is to be able to make a voice module for eurorack that can be controlled by gamepad. Thank you for your help!

Hello,
I was wondering if there is some kind of way of connecting HI to control RNBO parameters in a Raspberry Pi. 

controlling-rnbo-parameters-with-hi-gamecontroller-in-rspi

yes, you can use the evdev python library to send game controller data to RNBO via OSC

run something like this as a python script on the rpi

from evdev import InputDevice, categorize, ecodes
import liblo as OSC
import sys
gamepad = InputDevice('/dev/input/event7')

# send all messages to port 1234 on the local machine
try:
    target = OSC.Address(1234)
except OSC.AddressError as err:
    print(err)
    sys.exit()

# start the transport via OSC
OSC.send(target, "/rnbo/jack/transport/rolling", 1)

for event in gamepad.read_loop():
    if event.type == ecodes.EV_KEY:
        keyevent = categorize(event)
        print(keyevent)
    if event.type == ecodes.EV_ABS:
        if event.code == ecodes.ABS_X:
            if event.value < 115 or event.value > 135:
                left_horiz = event.value
                left_horiz_norm = left_horiz / 255
                print(f"left horizontal: {event.value}")
                OSC.send(target, "/rnbo/inst/0/params/mix/normalized", left_horiz_norm)
        elif event.code == ecodes.ABS_Y:
            if event.value < 115 or event.value > 135:
                print(f"left vertical: {event.value}")
        elif event.code == ecodes.ABS_RX:
            if event.value < 115 or event.value > 135:
                print(f"right horizontal: {event.value}")
        elif event.code == ecodes.ABS_RY:
            if event.value < 115 or event.value > 135:
                print(f"right vertical: {event.value}")
        elif event.code == ecodes.ABS_HAT0X:
                print(f"dpat horiz: {event.value}")
        elif event.code == ecodes.ABS_HAT0Y:
                print(f"dpad verti: {event.value}")
# Add additional elif statements here to print values for other ABS codes

Thank you John!! I will work on it right away!! Will keep you posted!

Controlling RNBO parameters with HI gamecontroller in RSPI

taavet

Hi is anybody here who is also using Max to create some NFT's? How is it going? I've sold some in HEN https://www.hicetnunc.xyz/th001/creations . Seems that all those years of experimenting could finally pay off.. :D

Hi is anybody here who is also using Max to create some NFT's? How is it going?  I've sold some in HEN https://www.hicetnunc.xyz/th001/creations .  Seems that all those years of experimenting could finally pay off.. :D

Hi is anybody here who is also using Max to create some NFT's? How is it going?  I've sold some in HEN https://www.hicetnunc.xyz/th001/creat

creating-nft's

Hi Taavet that link doesn't look like its working anymore, can you share with us? Curious about how the new RNBO library might overlap with your process.

creating NFT's

Alex Van Gils

Hello everyone, this post is to announce an experimental integration that we are excited to share — an adapter for Unity’s Native Audio Plugin. Using RNBO’s C++ source export, this repository will help you build a plugin which provides an API that facilitates integration of your RNBO device with the Unity game engine.

As described, this is experimental software. Please use at your own risk and know that anything in this repository could change in the future. However, we very much want to hear your thoughts and see what you make.

If you find problems or have feature requests, please track them using the GitHub repository’s issues page.

https://github.com/Cycling74/rnbo.unity.audioplugin

Hello everyone, this post is to announce an experimental integration that we are excited to share — an adapter for Unity’s Native Audio Plugin. Using RNBO’s C++ source export, this repository will help you build a plugin which provides an API that facilitates integration of your RNBO device with the Unity game engine.

As described, this is experimental software. Please use at your own risk and know that anything in this repository could change in the future. However, we very much want to hear your thoughts and see what you make.

If you find problems or have feature requests, please track them using the GitHub repository’s issues page.
https://github.com/Cycling74/rnbo.unity.audioplugin

Have fun!

Hello everyone, this post is to announce an experimental integration that we are excited to share — an adapter for Unity’s Native Audio ... 

rnbo-unity-audio-plugin

Experimental.. risk.. fun.. what a nice combination! :D
Thank you so much for this additional extension or extension of the addition; have been checking the forum once a week in the home that this announcement would pop up (as I hadn‘t yet managed to combine RNBO, Unity and Android)..
Will dive in right away!

Hey Alex, I'm building RNBO-generated audio plugins into a multiplayer theatre-VR-installation-performance hybrid, and it's working great so far using Unity and Quest 2.
Thanks for this nice starting point!

I do have some questions though that don't fall into the category problem or feature request. Do you prefer that I post them here, or also via GitHub issues? 

Hi Jan, it’s so cool that you are using this for your project — it sounds awesome.

Please feel free to share your thoughts here, and if it makes sense to move some of them over to the repo as an issue, we can do that later!

Yes, since the announcement of RNBO I had been waiting to be able use it in our Unity VR project.. Good timing! 
It's an audiovisual experience for (for now) 4 persons in a shared physical / virtual space, where they can engage and interact sonically. And play with a room-sized sequencer...

I'm mainly using about 6 RNBO-synths per person, which I control individually via OSC. As all players are networked and hear/see the others, the result is that on each Oculus 2 there are at least 24 synths playing simultaneously, on spatialized objects. And that works fine!

Right now they play continuous sounds, but I'd like to make them a spatially distributed synthesizer, so would like to send midi notes to them from Max, Ableton or both - or use the spatial sequencer for that.

One question that I'm having is this:
The setup I use now is a hacky modification of the  CUSTOM_FILTER example code, via local (and spatialized) Audio listeners. But I'm unsure of how to include the Synth Helper as in this example... Doesn't that reside in the audio Mixer? How does that combine with the spatialized setup?


Hi Jan, thank you for this! If it becomes reasonable to do so, it would be super cool if you shared your project. 

I understand your question. Basically, if you are making a custom filter, like the CUSTOM_FILTER example, you don't need to use the Helper in order to get access to those methods like SendMIDINoteOn() -- you can get the methods from the Handle directly. The Helper is really there to help you access an instance of your plugin which has been loaded onto the audio mixer. 

Here's an example where we send a MIDI CC message to our custom filter called "synth." You can use the same principle to send MIDI notes with SendMIDINoteOn(), etc, as in the MIDI example.

Certainly let me know if I can help further or if this doesn't answer your question.

using UnityEngine;

[RequireComponent(typeof(AudioSource))]
public class OrbPlayer : MonoBehaviour
{
    TestOrbsHandle synth;

    public OrbPlayer() : base() {}
    
    void Start()
    {
        synth = new TestOrbsHandle();
        
        // MIDI channel 1, controller number 1, value 110
        synth.SendMIDICC(1, 1, 110);
    }

    void Update()
    {
        synth.Update();

    }

    void OnAudioFilterRead(float[] data, int channels) 
    {
        if (synth != null)
        {
            synth.Process(data, channels);
        }    
    }
}

Ah thanks, that de-confuses me in that matter, and it works!

Another confusion I have is about the updating of Parameters.
Does that have to happen in each Update() cycle for *all* the parameters, like

void Update() 
{
    synth.SetParamValue(param_ratio, ratio);
       // and then quite some more like this
    synth.Update();
}

- even when the parameter doesn't change in-between?

Thank you for the question --  You don't need to update the parameter value every frame, when Update() is called. You can call SetParamValue() outside of the Update() method.

using UnityEngine;

[RequireComponent(typeof(AudioSource))]
public class DrumFilter : MonoBehaviour
{
    
    RNBODrumHandle drums;

    readonly System.Int32 numerator = (int)RNBODrumHandle.GetParamIndexById("numerator");
    readonly System.Int32 denominator = (int)RNBODrumHandle.GetParamIndexById("denominator");

    public DrumFilter() : base() {}
    
    void Start()
    {
        drums = new RNBODrumHandle();
    }

    void Update()
    {
        drums.Update();

        if (Input.GetKeyDown(KeyCode.T))
        {
            ChangeSubdivision(Random.Range(1f, 4f), Random.Range(1f, 8f));
        }
    }

    void OnAudioFilterRead(float[] data, int channels) 
    {
        if (drums != null)
        {
            drums.Process(data, channels);
        }    
    }

    void ChangeSubdivision(float numeratorValue, float denominatorValue)
    {
        drums.SetParamValue(numerator, numeratorValue);
        drums.SetParamValue(denominator, denominatorValue);
    }
}

hey all! anybody succeeded to make this work with unity 2022? i think the .bundle gets imported incorrectly, the plug-in is not seen from the mixer as effect. 
All working fine in unity 2021. 
cheers
Stefano
(edit: solved, see below)

Hi Stefano, thanks for reaching out... would you mind sharing your experience over on the GitHub repo's Issues page? 

I would expect the plugin to work with Unity 2022 versions. If you could share your exact Unity version, and maybe a little more about how you are importing the built package and what you see on the mixer. 

Hi Alex, thanks for answering, it seems i cried wolf too early: I made a couple of new test projects and all went well there. It seems i had something wrong with the first project i experimented with. The only oddity i see is a field in the inspector of the imported .bundle  (see image below), but the plugin actually works fine, sorry for the false emergency.

Thank you for following up, Stefano! I've tracked this for some sleuthing here: https://github.com/Cycling74/rnbo.unity.audioplugin/issues/29

Hello! I'm here just to say that I'm thoroughly enjoying learning RNBO (and Max) for use in Unity!

Ran a few "hello world" tests just to make sure I understood the process and hope to build out some fun prototypes in the coming days.

Is this forum the best place to discuss, suggest, ask general topics? (Any obvious bugs/issues I can take to github, but curious if you have a preferred channel for general convo).

Thanks for creating the RNBO -> Unity Repo and docs! 


Hi Tonzari! Glad that you are having fun checking this out, and certainly, we can chat about the topic here. Fire away : )

Hi @Alex. This is great and I've been having a lot of fun experimenting with this implementations. Thank you very much for the thorough documentation!
I want to be able to use the spatialization option that unity provides with my plugin. Seems like Jan figured out a way to do it. Is there a more direct way to implement the spatialzation? Maybe a way to route the signal around? I'm thinking something along these lines, where you add an empty file to the audio source and then multipy and adc~ input by the end of the signal in rnbo (see here https://github.com/LibPdIntegration/LibPdIntegration) 
I haven't tried this yet, but I wanted to ask first and see if this id something that was already adressed. Thanks again!



Hey @NCDLMN, I'm glad you've been having fun! Certainly feel free to share anything you've been hacking on, it would be cool to see/hear what folks have been doing. 

Probably the easiest way to take advantage of Unity's spatialization features is to make a "Custom Filter," as in this example: https://github.com/Cycling74/rnbo.unity.audioplugin/blob/main/docs/CUSTOM_FILTER.md 

What's sort of cool about this is that because this requires an Audio Source, you can naturally take advantage of the 3D spatialization algorithms that Unity provides for an Audio Source. And if you change which spatializer you are using (in Unity's settings), that change should apply to this Audio Source as well.

Does this point you in the right direction? 

A

If you'd like to see this implemented, the "humming orb" from our demo project ( https://cycling74-assets.nyc3.digitaloceanspaces.com/rnbo/unity/example-projects/RNBODrumkit.zip ) uses this strategy. 

A

Hi Alex, thanks for your reply. 
Yes the CUSTOM_FILTER example works great. Just a couple of questions. I have a plugin that requires a buffer (a granular synth) and I'm able to make it work as in the BUFFERS.md example.  But I cannot use both the buffer script and the filter script in the same object (I dont get any sound out.) Probably has to do with the instace number? But becasue the filter does not generate the GUI for the plugin, I cannot access it. Does that make sense? The overall question would be, is there a way for the CUSTOM_FILTER example to generate the plugin GUI automatically? Or would I have to program each of the parameters again, as in the TRANSPOR_TEMPO.md example?  



Thank you for the question -- so, I think I have a sense of what you are trying to do, but I think there might be some slight confusion. If you make a custom filter from your RNBO device, you don't need an Instance Index -- that Instance Index is what we use to distinguish which instance of a RNBO device, loaded on an Audio Mixer, a given component should manipulate. 

If you make a filter, as we do with the CUSTOM_FILTER example, we don't load that RNBO device onto a track on the mixer. Instead, the Audio Source associated with your filter has an Output field in the inspector, where you can select your desired destination for this audio.

So you don't need to generate the GUI with the param sliders, instance index, etc. This is all for the benefit of the Audio Mixer. You can just create your filter and set the Output of the Audio Source. If you want to add buffers to your custom filter, you can do so just like the BUFFERS example, and when you create your

[SerializeField] AudioClip Buffer;

you should see that field in the inspector of your filter component. 

Does this all make some sense? If not, please feel free to share your script, either here or in an email to support@cycling74.com, and I'll be happy to try to help.

RNBO Unity Audio Plugin

leemorgan

I have an audio-generating patch (or, if needs be, just audio) that I'd like to come out of two different iPhones - synchronised. The audio is different for each phone.

It occurred to me I could use RNBO to generate code for an app - or two different websites? - to play the audio - and synch them by having them start on the next whole minute of their respective clocks (and just loop from that point)

My question is: does RNBO have the ability to access the internal clock/any universal clock? Unless I've misunderstood what RNBO can do.

Or is there a simpler way to achieve this that I'm missing? I feel sure there must be.

Advanced Max user, beginner at RNBO....
I have an audio-generating patch (or, if needs be, just audio) that I'd like to come out of two ... 

syncing-two-different-iphones

To synchronize audio to multiple devices, you might consider generating your sound with one device and then "casting" it to multiple devices, much like web radio works. We provide the source for the "runner" that we use on the raspberry pi, it can also be built and run on other platforms. I know of at least one person who is running that code in a virtual machine in "the cloud" and then casting it through a webpage so all viewers of that webpage have a more or less synchronized experience.

https://github.com/Cycling74/rnbo.oscquery.runner 

I'm re-reading your posting and realizing that you're asking a question that a few other people have asked, how to get the current "wall time" into a rnbo patcher.. I suggest you use a parameter and in your integration code you can set that parameter with the current time as you see fit, you can then use that to sync however you want inside your rnbo patchers.
RNBO itself doesn't have a way to get the current time, but it should be easy enough for you to do in javascript.


Thanks Alex - though seeing 'easy' and 'javascript' in the same sentence is a little off-putting.  Oh well, time to learn I guess! 

Syncing two different iPhones 

Ernest

A long time I played with capturing the stream from two audio waves with known wavelengths and drawing them as lines in real time. However, although they looked like mesh objects, I was unable to convert them to real mesh objects. That is, I could convert them to OpenGL mesh objects, but the resulting object has polygons in a spiral, whereas one really wants a blob-type mesh so that one can apply materials to the mesh object properly.

Thinking about it again recently, I realized I was going about it the wrong way. Instead of making a continuous line for each of the two audio sources separately, I should be stuffing each two synchronous values from the audio source into the same {x,y,z} mesh coordinate.

I started to figure out how to code it, but I'm not quite sure how to set individual mesh coordinates. I'd rather use a FOR LOOP in jit.gen I guess, processing say 512 samples from a stereo audio buffer for each video frame. But who exactly do I set individual mesh coordinates in an OpenGL mesh?

A long time I played with capturing the stream from two audio waves with known wavelengths and drawing them as lines in real time. However, although they looked like mesh objects, I was unable to convert them to real mesh objects. That is, I could convert them to OpenGL mesh objects, but the resulting object has polygons in a spiral, whereas one really wants a blob-type mesh so that one can apply materials to the mesh object properly.  

Thinking about it again recently, I realized I was going about it the wrong way. Instead of making a continuous line for each of the two audio sources  separately, I should be stuffing each two synchronous values from the audio source into the same {x,y,z} mesh coordinate.  

I started to figure out how to code it, but I'm not quite sure how to set individual mesh coordinates. I'd rather use a FOR LOOP in jit.gen I guess, processing say 512 samples from a stereo audio buffer for each video frame. But who exactly do I set individual mesh coordinates in an OpenGL mesh?  

Hi folks
A long time I played with capturing the stream from two audio waves with known wavelengths and drawing them as lines in real time. 

creating-opengl-mesh-object-from-two-audio-waves-in-real-time

Well I started lookiont at jit.fill, but I can't find any information on how to order the data in a jit.matrix, and the git.gl.mesh object has over a dozen attributes for @draw_mode, on which I can find no documentation. 

Hi.
For a nice "draw modes" explanation, see the following tutorial.

Anyway, I'm not understanding what you mean by "the resulting object has polygons in a spiral, whereas one really wants a blob-type mesh so that one can apply materials to the mesh object properly."
jit.gl.graph receives a matrix with the geometry and draws it as a 3D mesh. Of course you can use other jitter objects, even jit.gen to draw it yourself according to your intended purpose.
Can you explain a little more your intended geometry?

Thanks! Well when I wasd piping the signal in with jit.catch~, with left channel to x and y channel to y, it puts each point in the buffer sequentially until it fills up, so you get a spiral line strip. So what I really should be doing is building a series of strips with overlapping vertex points from the last ring, I guess. 

The jit.graph object puts the rings at equidistance on the z axis. What I want is to make a blob with the second audio channel controlling the z axis. So the strips also have to wrap back to their start. 

Like a ring waveform? jit.gl.graph also has a "@radial 1" mode. If it's something more specific, you would have to configure a matrix to represent the intended geometry. I'll try to post an example here.

Not a ring, I was thinking of a blob. It would be a sphere if both waveforms are sines. It also requires a variable matrix size across time to accommodate changes in frequency, unless the waveform is pitch shifted to fit the same frequency. 

Here's my starter patch on this, which fills a buffer~ with a sine wave, then uses zl.stream to build a message of the values in the buffer from peek~. That makes it easy to integrate with any kind of source audio. Then it stuffs the message into jit.matrix with jit.fill, But I haven't got as far as putting the data into a jit.gl.mesh mesh object yet. I did get as far as figuring out the origin of the matrix is in the bottom left corner of jit.matrix and jit.cellblock. 

I got a jit.gl.mesh coordinate matrix input to draw SOMETHING, but I still don't understand how to arrange the data in the input jitter matrix. I tried putting the x y z coordinates in an array in one plane. That doesn't appear to work except for drawing lines. Maybe each coordinate is in a separate plane, or maybe each axis is in a separate plane. This is going to take some more trial and error. 


I finally drew a polygon by setting matrix values manually, Im not sure exactly how, but it's in the Z axis so you have to twiddle the window handle to see it. So none of the tutorial objects supplied draw like it says they are going to, which is possibly because it relies on some undocumented feature that has changed. which could also be the ordering of point data in mesh coordinates. But it appears from what it says that the x, y, and Z cordiantes are in three separate planes, with the values for each successive point in successive rows in the matrix, with each column drawing a separate shape. That would explain why my experiment is in the Z plane, because I didn't set any third-row values yet. But now I need to take a break from it and start from scratch again. 

On the third try, I can draw ordered points,, and from combining bits of this from my first I attempt, I can finish my project. Being deaf, I won't be able to design it for real-time audio. Probably I'll make some harmonic wave generators  to create some interesting non-random organic shapes with ripples slowly moving across them. Thank you for the assistance. 

i dont know if you know this but the message plane -1 allows you to see all the planes at once in a jit.cellblock, in this case x y and z

also in gen you can use the cell operator to access individual points, which might help you?

Than ks Tikoda, I didnt know about the cell operrator in gen, that's a really big help. I looked at various draw_mode settings, and it seems I need the polygon value to texture the triangles rather than color the edges, so it looks like I just draw each triangle in a separate row, which will be easy with a gen~ FOR loop :) 

Here's an example modulating a sphere shape with 3 different frequencies captured by jit.catch

By the way, I couldn't get the planes to display in multiple rows in jit.cellblock, but I was just trying to figure out if the jit.fill was working. I did merge the first piece for putting a sinewave into a buffer into this prototype before writing any gen~ script for it. 

It transpires there is a new appendix B in the Jitter Reference Tutorial that contains the information I was looking for. What I was actually having trouble with was plane 12, which I should be able to figure out now. 

I could get the "planes -1" working on jit.cellblock by setting colwidth manually after setting a 1000 msec delay from loadbang. I still have axes disappearing which I remember is an old problem and I guess I will be replacing them with my own gridplanes manually again. 

Pedro, thanks for the contribution, but I was looking for a mesh shape not nurbs, thanks again )

I added my fundamental subpatch for drawing a stage again, here it is. It includes the simplest possible preconfigured objects for a stage, with the grid set around -1~+1 for width and height, and 0~-1 in depth. scaling the render context by 50% so the grid aligns with the viewport edges. 

Now I can manually draw individual lines into jit.gl.mesh. Looking in the forums, no one had done this before. When draw_mode is set to "lines," one has to set the beginning and end point of each line as consecutive pairs of points, one point in each row, with the x y z coordinates in the first three planes of a jit.matrix object, then pass the jit.matrix into the first input of jit.gl.mesh. Here I drew two horizontal lines on the back and side axes of the stage. 


This stage adds an iterator that converts lists of [x y z] coordinates into an OpenGL mesh (in this case, the stage's axis lines). The message at the top right generates 12 lists, which are each fed into SETCELL messages for jit.matrix, which then passes the matrix into a jit.gl.mesh object for rendering.  So now I can use the same iterator to generate meshes from wave sources. 


I added a camera with the preferred standard orthogonal view, adjusted the alpha blending of the grids, and moved the stage back so the origin is in the absolute center of the stage  It's still got handles to move the view by clicking and dragging in the too. At some point I will again add a secondary viewport for setting the lights, with a cone and rectangular box to represent the camera. I've got quite a bit better at it since Maz 6. But storing all the position and angular data with the preset is still quite a bit more work, so I'm deferring on that bit. 


For audio I am importing my BEAP waveset oscillators, made with my synthcore v1 library. The v2 library on my site at https://yofiel.com has antialiased  saw and triangle oscillators too, so I plan to update the BEAP code later.

The waveset oscillators require that the patch be saved and opened from the same directory as waveset_all.wav is saved, which is in the zipfile below.  

So now the BEAP oscillators and gen~ object are in place to draw triangles. 

Unclickable elements are hidden when the patch is locked. The stage is slightly adjusted so that the grids surround the origin, from (-1, -1, -1) to (1, 1, 1), and the camera repositioned for a default orthogonal view. You'll need the zipfile from the prior message as it described for the oscillators to work.

 Currently the generated mesh is normalized and translated to fit in the viewport, which I will be adjusting so that the mesh grows and shrinks with changing amplitude. But I still need to feed the CV voltages of the oscillator pitch frequencies into gen~ as well so I can scale and rearrange the signal data into triangles or quads. So I am thinking actually it needs another two input buffers which I resample to fit a particular number of mesh points, which will need to capture audio data down to 50msec so 20Hz waves display properly. 

Then I'm thinking the easiest thing to do is draw triangles with shared vertices using history operators in gen~ to capture the last edge. I think that means the output buffer needs to be triple the number of sampled points. 

My plan is to make two hemispherical shapes, one for the upper half of the signal, and one for the lower half, then position them one on top of the other one upside down, which will simplify generating the normals for the mesh polygons, because all the faces will be on the same side of the mesh. That's before adding materials, an enhanced camera stage with secondary view with adjustable lighting, and adding the analog oscillators to the BEAP gen~ code.  then I guess I put FFT onto input audio to extract the fundamental frequency so it can work with other audio sources. 

There are examples of people using a similar method to place points to draw speaker positions for multi-channel audio set ups. in this case they used a coll to write the point positions.

Not sure I totally understand what you’re aiming for here but I’m curious to see the result! 

Is there a reason you’re not using jit.catch to turn the audio into matrix data? is this uzi/peek list method doing the same thing but Just on one plane?

the problem I had with jit-catch is that the data is not in the right order for creating surface polygons. 

As a next step which makes something useful in the process, I am thinking of changing the height of points on a regular 2D mesh by combining the left and right channels.  The leasdt number of points are possible with tri_grid, in which each concsecutive point forms a new triangle with the prior two points. So one can tesselate ordered points to form a 2D mesh like this, generating 8 triangles from 11 points. Drawing each triangle separately is simpler to program, but it would need 24 points.  


So for the waves to move smoothly on the display,  the mesh has to generated from an integer number of waves at the fundamental frequency of the input signal, with the wave samples starting and ending on zero crossings. 

As the input is currently a synth, the input frequency is known, so the output is taken from the input buffer truncated to the following length from the first zero-crossing sample:

waveCnt = floor(output_buffer_size / input_freq); 
sampled_buffer_length = waveCnt * input_samples_per_cycle 

From which the output samples to generate mesh point values are taken, for example with a 1024-point quad mesh: 

output_sample_interval = sampled_buffer_length / 1024

So samples are taken from the input buffer's last zero crossing backwards,  at output_sample_interval for wavCnt samples. 

A 1024x1024 mesh is 1M points for each frame, which at 25Hz is 20 points per microsecond, so it probably needs some serious parallelization and code optimization. But that captures audio waves down to 25Hz at standard audio rates. I guess I should make the mesh size a variable lol.

A simpler approach would simply be to resample from the last single input wavecycle, and only display one wavecycle regardless of input frequency. So I'll start with that and a 64x64 mesh so as not to hit performance issues, and write the code so I can enhance it later. 

From that,  I can construct a loop in gen code to resample the input buffer values so the wave fits exactly on the mesh length. if I got it all right., then the waves will fit exactly on the mesh. I don't think spline interpolation between the input samples will make enough of a difference, and either linear interpolation or the nearest sample will be enough, but I'm not sure. 

Thus my draft code for resampling is this. I still have to adjust the output sample order for correct mesh tessellation:

Buffer bX("bX"), bY("bY"), bZ("bZ");
Data inL(50000), inR(50000);
History idx(0), odx(0), inLz(-1), inRz(-1), zCrossL(0), zCrossR(0);
Param qmetro(50), freqL(20), freqR(440), bufLen(64);
l0 = dcblock(in1);
r0 = dcblock(in2); 
idx += 1; 
if (idx >= 50000) idx = 0;
if (l0 >=0 && inLz < 0) zCrossL = idx;
if (r0 >=0 && inRz < 0) zCrossR = idx;
inL.poke(l0, idx);
inR.poke(r0, idx);
inLz = l0;
inRz = r0;
wavelenL = samplerate / freqL;
wavelenR = samplerate / freqR;
ivalL = wavelenL / bufLen; 
ivalR = wavelenR / bufLen; 
odx += 1; 
if (odx >= bufLen) odx = 0;
u = 50000 - (zCrossL + odx * ivalL); 
v = 50000 - (zCrossR + odx * ivalR); 
if (u < 0) u += 50000;
if (v < 0) v += 50000;
w = inL.peek(u);
x = inR.peek(v);
bX.poke(w, odx);
bY.poke(x, odx);


Of course this is filling the output buffers continuously, when they only need to be filled at the qmetro rate for the screen refresh, but it gets me going.  

The resampling buffer to fit a single-cycle input wave onto the OpenGL mesh is basically working. Here I set the draw mode to "triangles"  with 64 points. but I'm not sending duplicate points, so they aren't joined together.  


So in view of the lack of interest in anything except telling me what to do instead, I dropped the project. Here's the final code for the above screenshot. You may try to persuade me to do something else, but at this point I'd want money lol, I'm retired and I don't need it so no worries. Cheerio )


Hi, Ernest. I was following with curiosity the documentation of your process here. I like reading these topics that are written as a sort of public personal journal on a particular project.

I would like to add that the perceived lack of interest on a particular topic in a forum can be just that, perceived. And sometimes, even if there's no feedback being given at the time of writing, there's always someone down the road that will find it as invaluable information. So, thank you for your efforts!

Creating OpenGL mesh object from two audio waves in real time

spin lud

Hi guys, I have to prepare a live event at a climbing gym. I would like to add some reactive visuals to the music, to make the event more cool 😎 

I am total newbie at Jitter, just started to watch some tutorial videos few days ago 

My general idea is to play randomly small clips of climbing videos as a base layer. On top of that, I would like to add reactive visuals to the music, generative shapes/particles systems or whatever. I am not sure if it makes sense to blend these visuals with the videos (like a crossfade) or make them modify the video itself, if this is even possible. As final touch maybe some postprocessing effects like bloom etc.

Do you guys have any suggestion for examples about similar concepts (videos + reactive audio visuals)?

Also any suggestion about what devices/effects to use?

Hi guys, I have to prepare a live event at a climbing gym. I would like to add some reactive visuals to the music, to make the event more cool 😎 
I am total newbie at Jitter, just started to watch some tutorial videos few days ago 

My general idea is to play randomly small clips of climbing videos as a base layer. On top of that, I would like to add reactive visuals to the music, generative shapes/particles systems or whatever. I am not sure if it makes sense to blend these visuals with the videos (like a crossfade) or make them modify the video itself, if this is even possible. As final touch maybe some postprocessing effects like bloom etc.

Do you guys have any suggestion for examples about similar concepts (videos + reactive audio visuals)?
Also any suggestion about what devices/effects to use?

Hi guys, I have to prepare a live event at a climbing gym. 

looking-for-ideasexamples-climbing-videos-reactive-audio-visuals

Looking for ideas/examples: climbing videos + reactive audio visuals

Are there any significant differences between Mira and Miraweb? 

I have the Miraweb, tests working very well on a M1 Mac and Android clients. I wonder if I miss some functionality using the Miraweb instead of Mira on an Ipad? I have no Ipad to check out myself and did not find any statements what would be different, if anything. Could some one tell? Thanks!

Are there any significant differences between Mira and Miraweb?  
I have the Miraweb, tests working very well on a M1 Mac and Android clients. I wonder if I miss some functionality using the Miraweb instead of Mira on an Ipad?  I have no Ipad to check out myself and did not find any statements what would be different, if anything. Could some one tell? Thanks!


Are there any significant differences between Mira and Miraweb?  
I have the Miraweb, tests working very well on a M1 Mac and Android ... 

mira-vs-miraweb

Sorry, I'm new to all of this. I asked ChatGPT 3.5 about this:

"Mira and Miraweb are both related to Cycling '74's Max software, a visual programming language for music and multimedia. However, there are some differences between Mira and Miraweb:

Platform Compatibility:Mira is designed specifically for iPad devices.
Miraweb, on the other hand, is a web-based interface that allows you to control Max patches from various devices with a web browser, including iPad, Android, and computers.


Functionality:Mira is tailored for iPad and may have specific optimizations or features that make use of iPad capabilities.
Miraweb is more versatile as it can be accessed from different platforms, but its functionality may depend on the capabilities of the device and the browser being used.


Testing Environment:As per the user's comment, Miraweb has been tested and works well on an M1 Mac and Android clients.
The user is seeking information about potential differences when using Mira on an iPad, which they don't have access to for testing.


Community Input:The user is reaching out to the community to gather insights on any significant differences between Mira and Miraweb, especially when used on an iPad.


Without specific details from Cycling '74 or the community, it's challenging to provide a comprehensive list of differences. Users in the Max forums or community discussions related to Mira and Miraweb might offer more insights based on their experiences with these tools on different devices."

ChatGPT is quite useful here and there, but you always gotta take things with a grain of
salt. There were instances where it told me a lot of a baloney.


Mira vs Miraweb

using Ableton suite is it possible somehow to open Max in standalone without passing through Ableton? 

 in Contents->App-Resources-Max (MacOS) but it loads without license activated (so saving is disabled). Only when opening it thorugh a M4L device in Ableton the license loads correctly.

Hi,
using Ableton suite is it possible somehow to open Max in standalone without passing through Ableton? 
I have tried to open directly Max.app in Contents->App-Resources-Max (MacOS) but it loads without license activated (so saving is disabled). Only when opening it thorugh a M4L device in Ableton the license loads correctly.

Hi,
using Ableton suite is it possible somehow to open Max in standalone without passing through Ableton? 
I have tried to open directly ... 

max-in-standalone-with-ableton-license

you have to buy max license to use it without live.
max run as Live's "slave"  blocks all audio drivers beside Live
and disables all midi devices, which are main drawbacks.
Otherwise, you can  open or create max patches as usual. 
midi problem can be solved by using imp.midi externals,
exporting collectives or standalone apps is disabled,
but not impossible 


Max in standalone with Ableton license?

Mathieu LENGAGNE

I just made a little volca beats controller for an upcoming electroninc session I have, I thought it could be usefull to some ! There is a resizible step sequencer, two probability based random triggers with their own time subdivisions, swing, random velocity for more interesting hi-hats and easy access to all avaibles CC of the volca beats. It can easily be modified to implement more random stuff if needed :)

VOLCA_BEATS_CRTL.maxpat

Hello everyone !

I just made a little volca beats controller for an upcoming electroninc session I have, I thought it could be usefull to some ! There is a resizible step sequencer, two probability based random triggers with their own time subdivisions, swing, random velocity for more interesting hi-hats and easy access to all avaibles CC of the volca beats. It can easily be modified to implement more random stuff if needed :)

Enjoy !

Hello everyone !

I just made a little volca beats controller for an upcoming electroninc session I have, I thought it could be usefull to ... 

volca-beats-midi-controller-patch-sharing

VOLCA BEATS MIDI CONTROLLER - Patch sharing

Ullstein

maybe old-fashioned problem, but I need to specify the port to a midiout-object, but it does not accept the message "port n" (in my case 'from Max 1' is # 13 in the list = equal to letter n). It does not accept any other letter and always stays connect to the first port which is 'AU DLS Synth 1'. Adding the letter to the object as "midiout n" does not work either. Connecting "midi info">"umenu">"midiout" shows no difference.

Any advice available anywhere? Many thanks

maybe old-fashioned problem, but I need to specify the port to a midiout-object,  but it does not accept the message "port n" (in my case 'from Max 1' is # 13 in the list = equal to letter n).  It does not accept any other letter and always stays connect to the first port which is 'AU DLS Synth 1'.  Adding the letter to the object as "midiout n" does not work either.  Connecting "midi info">"umenu">"midiout" shows no difference.

Any advice available anywhere?  Many thanks 

maybe old-fashioned problem, but I need to specify the port to a midiout-object,  but it does not accept the message "port n" (in my case ... 

midiout-does-not-accept-port-message

"from Max 1"  message to midiout is all you need.


I know, unfortunately none of the midi-objects react to any of these messages. I think it is some kind of weird bug. and I will have to deinstall Max and install it again. But thanks for thinking for me.

Actually I just try it on my MacBook and it 's the same. Can it be a problem with M1/M2 Macs?

what do you see when you doubleclick midiout object ?
and what is with midisetup ?


double clicking the object shows me the list at port 1, but I can select the port I want and close it. It will work, but it loses the information when I quit Max and restart it.
Sorry, don't understand what you mean with "midisetup", there is no such object. The Apple AUDIO-MIDI setup shows me the connected interfaces and instruments. But indeed there is one problem, I own an Euphonix/AVID Artist Mix console, that also has 4 MIDI connectors. Although I deinstalled the AVID-software for it, the MIDI- symbol is still in the MIDI-setup and I cannot erase it.  I dumped it from the MIDI-devices folder in Library/Audio but it still remains in the MIDI-Studio window.


OK, I restarted the Mac and the symbol is gone, nevertheless, the MIDI-objects do not accept port messages on two different computers. I wrote to cycling for support. Many thanks again, you are always very attentive and friendly giving your advice. Thanks for that

midisetup is in options menu, hard to overlook.

I don't quite understand this :
double clicking the object shows me the list at port 1

what is the list at port 1 ?
you should see a list with all available midi output destinations.


Max patch can't remember selected midi ports, if you have problem with messages, maybe try to write output as argument



I also remember that in some weird cases prepending name
fixed the problem, for example
name "AU DLS Synth 1"


Sorry, I wasn't aware of the main menu item. I disabled all other ports and put the letter for "from Max 1" behind midiout and it worked. Then I reales all MIDI out ports and the midiout-object staid on "from Max 1" So that is a workaround that will work.

The solution 'midiout "from Max 1"' works but is not stable.

Anyhow, thanks, I can manage with the workaround by disabling all the other ports.

this is rather strange behaviour, on shoud not need to disable other ports.

Maybe it has to do with creation of virtual midi ports.
I never used that from / to max ports, but rather relied on IAC bus,
which is system created midi port(s).

OK, Thanks, I will check this, although in this specific case, I can do with disabling other ports, since it is for a dedicated installation, where the computer is supposed to only work for Max...

midiout does not accept port message

Ben Bracken

, available now for download on the Max Package Manager.

This one just has a single improvement relating to the "Start RNBO on Launch" Preference, fixing RNBO server startup issues when this preference is disabled.

https://cycling74.com/releases/rnbo/1.2.3 

We've just released RNBO 1.2.3, available now for download on the Max Package Manager.

The release page is here:
https://cycling74.com/releases/rnbo/1.2.3 

Happy Patching!
Ben & Cycling '74

Start RNBO on Launch Preference: fixed RNBO server startup when disabled

Cycling '74 Forums

Hi folks, We've just released RNBO 1.2.3, available now for download on the Max Package Manager. This one just has a single improvement ...

We are happy to announce the release of Max 8.5.6 and RNBO 1.2.1.

Thank you John!! I will work on it right away!! Will keep you posted!

Hi Taavet that link doesn't look like its working anymore, can you share with us? Curious about how the new RNBO library might overlap with ...

Hi NCDLMN, Thank you for the question -- so, I think I have a sense of what you are trying to do, but I think there might be some slight ...

Thanks Alex - though seeing 'easy' and 'javascript' in the same sentence is a little off-putting. Oh well, time to learn I guess!

Hi, Ernest. I was following with curiosity the documentation of your process here.

Hi guys, I have to prepare a live event at a climbing gym.

ChatGPT is quite useful here and there, but you always gotta take things with a grain of salt.

I see, thank you

Hello everyone ! I just made a little volca beats controller for an upcoming electroninc session I have, I thought it could be usefull to ...

OK, Thanks, I will check this, although in this specific case, I can do with disabling other ports, since it is for a dedicated ...