Hi all,

I’ve been playing around with this in my free time and making some progress.

So far I’ve populated a couple of samplers with instruments, just to test different mic counts: a kick with 7 mics and a tom with 5. I was able to get the multimic merge working, and so far that part behaves well. I also added simple per-mic gain controls in the sampler FX, which is also working fine.

For routing, I’m currently using a container. Based on what I read in the forum and the docs, I added Routing Matrix FX modules in the container, one for each mic.

Here’s a snippet:

HiseSnippet 3290.3oc6c0sTaijEVFP.1AHYxlsp8loJsY2KLa.hk7evvxDiARBUvXVLI6jJ0TTxRsgdQVxkjLCraMUMuFyc6ivd6dWdjl2fc6V+X2pssvVXL1P6aBce59zmyQm9q++jiLMT.VVFlbwRbx0M.bwVfux051muy4xPct82Emgkl9oW.sOMkDWwqaHaYAT4hEa52gKQr3yv4762dSQYMYcEP6r339jATAb.rNztctGU3CPMs2JqBNAVmnzYJruhg9NFZFMQRyz7o3ZHqbg7YfCkwEaJdt2KacNWr+Be5p0xkQQLeU47RaTcib41HqZ5ZpxRoAJUqJKAPjyjsVNtXytmJz1vrhsrMvhK1LEMTutx4F+jtaC7InErpF.mPjqBpkcy9sFZpXUDmK2NmC0TOx2LYwgX5QsMZS6ZzdAeInJrU9sMdOygfP6ZPZ.iMUPwa5.hmHo3khP75hHEiPjlwUj9F9JJlvF1sofkmmvuutMvrlL56Don3VVtoVHA+NFnRnauVc4K.u0DknUMRtdpTqHjIUpk2LQBGuj0TA0.l6HqoUE8sxJosYS.l3qe8G1emOHTZ+cp7cBhas6mObEAos1oL5eRu0Ie7vC26.Di1p76WQH6VGWtboUDxs0m163hBUde4iOYEg7toNn7guKAxqvxV3vOV5TLSOEyTgsDxuo.pYNdaTybR4RzMElyo83bl.bNaW4LlO9bNKRA7HcF563a+ATde4G2zKqKzqhc7CjWUa8RMsAd44q8qNz9knlgoPRHh+o1T.J7WCZLPY8pWs7+JQ7W+5+gfsgvE.PCAcTmFKAz2WCEHx6RESnNTQHYIwUDJIsh.vVYs0Va4DwuT1DS4vl0qBLQMA7UhHSah3t59ZMZZcdR2O1mAr2qVMfhcxWV6pO.Ut.aHNsDT4kBupMKVF6.D2yL4VceWJDC1wndCCcThjuDUjPYhOO9B7GWyBWScaSCMemsjF5evktGgk2LdheNQBemh6XqeKGFBqeXF9+DxxCUv1dnxPv5er7Pv5GJSZY8C3q8pA6aQBntFTGHTqothMzPWftfIU7knUDtTVCgdfrjZFJxZBPcUvUHKhut3jtbs10.Kjtlou3PyQx1111DVE0YLYPRXV32Da5JaXWkhmbXEgS1qxINoPnvNcl+NASCb+YbGFQmOf3d7nrZzz1qKO96FpjXGX720SLJ4BJzcisWIcKE1bK9RjTzQ86tosnK5Bt.tXq8TLkadkP0lV2fjVDUhaTPQEp6xIhPnhIhtqT1t0MqiqtC6PMbGdyDjQslPharhGizOn9YkjQepuht93l1yrfLXcwAjvb1UuOT0ayPrp5ONYRmhf5+mZsrXG73HZ91Eamuedk368KwO2VXPeV5srfrYcUTBz.3B0I+6focJTPqx5HtkP.8iT0LA0MtDTAnqh9XpCbpdRD1Dd389proWQPD2IzozvZB9sT73jUUVUs6sQb5FoyR52B+LVUCUSwVmvUzfBfTOUxfMe59RA6j2cS45BmwJFg+vgF1fx5IWNApMPTDnIUqVWo40YTCX1Ux34YaFVESp6.66644WPzbICNAU9dOAUx4Oq3BsPTPC880g1ka.z60rp47viPSlcIOoBUTamoxtj2TYqnAUAlbPzLVeNO0rGD4bjcxEevEE9HQymO+0HwmzCI9jYHwmrCI9jaHwm7C.eJ1z11P2gOufuygRo4zu8l9gS+N9NFqa.XDspQNUpaiqHIejFR7I8PhOYFR7IaW3yG2eWYaY7JY8vAPXCM.l1PLrSrcAWBU.tqqMN+t.qKrMZvEapVSbAsnb21eQ+U8JqCzbZ944c9aQtqZ2feegqIa8eBpZedqLJXT3b.7ryI1xhqKDKAQasvf.KcE4NbPzr6Vntgpm9n.qhj15jamxm+Zc4qHERKaPiJv+ouY6e+q+5+8MVMqUChJEOufZQT8UU0.GYXAwn6sqqzWsAWYiKEZwwbMjMQJAg1zx.wEMUTpWp3u7o6AUbmx2ApX5dphxibUbdd28OoO0xEGjgZ5gV90yF4Z4L7keOGzFT2e+.WfWLuTlLqKkYiTCeUOaO6lpOxU8Y4waa0HT4y0KkWv7dv61cu5Fgpe9do9EZdeo9GbaU+3C1DnHzBURm+8K36RVZ68OjZbxe4u0w3jYJzawycmhCtqA9KFfn3Cz705gbijL+gCJ9wJ2NwtKVU+4Xbfb0Vywv4u68f8+4VFR7NoQ70UB+4MijnXp0EkDywU8r1e0yHkecwMRkZ8TAbGhymMUZQQo78ouvMMyHxwuQ83G3YFMa+OlZvYIOoLyHoAAYI3D3mzGPMxpd5G.CnFYkOyChATir5m8Aw.pRCDzuzMB8mf2+rJuqg+QRdmGG7y3KYn1TS1N3oSi+l4Q.a+HORX7w9pirqWStgHCsirteEwug+Hnsx4cWFmpKxHG2chL5cP+Kx6dpEsEvY3e6Ob2bp9jM+783xY7T9KPSIBOsJTN.ygw8ynQee+LhM8cw0qfer65Uv5N8Pn6Tr45r8iyiOKQ.pIws8h7DGXSPgX59VHDC2FD3rK4h8GOrY8JnNTJ.jIPGAdiOkkXOCuyptoSgS++P+vmLVqDdDEIIJRQTBmdJDmvDcRv0lsowom1inShXsIlgjsYnXaVRhYoHlijXNJh4IIlmh35jDWmh3FjD2f1HDvDI1gMJnQh1JIJEfrDM4zAHmllb.SkHssRLfwpcJtAwST5ggmnOQZKb.CLs80wfNimaZFJeXGyIuGwrT9vLOwgumX5G6dhQGSbVO2zbT9vNNey4QLOkOLySrWdhYXdhi9Qmm2yMccJeXGmu3dD2fxG9gtmXVlm3D07DS34m5lhvK1026I9jEo8iG68Dyw7DeDshkE7cToWYmqu2h9joWa2nvSLOySjs149DSbIeGU5U24568TexsVeGom3255I9b9J1l.45HeiJx0anAL8bIsbSg8KGc6LZB9Z0jqplKeVjnBzLjUINBC7g+UnXyZ0.lAy8+TvoE2ttQS8.MomFcLnAP1tj6wm3Qb9BGe76LMZ1ftVBEb1kuSLkUt.YTHU0x5.TeOaRUcGSCKqZHsxgWVjjNpo4YXSV6bNFbIvzJXdn9gj8.wKm4iV.b+cnhem01k1oURchrSe9o3cSKRkVhJcZpzYnRmkJcNpz4am10fVRtA1GIty9meZ85cbWmms+foP7w4PWHJYQ3YsZC2bl3dPgS.63cf27XbeYrBDa32S+RTGOTNXY7476BpI2Ty1O2fasaICciFmanCUB5libaO6LfIor2UEZaaamaxheNunvw.MfrEf7H.O.pCjMC16cfrEhC76+rqeu9C7thq.F.SXx8jJl9wyIUz9j+bbseWa6Sv2foXHivwwsKPSNfs5uG3l8v84BcYLv8QcXLsOxPS1j1ROrlSWrY5a2AmWc6XsKKFoAYlG2ERmO8i2B4yv2jg6IQjZUFeaWWkwRTqxfyadpccUFw7mDKaUFitUYvEALzvhuEe9qLLTFFJCC8tCCcHcKGnwPCbKGXXn20XnoYXnLLTFF5DGF5P594PigF394vvP6OLzLLLTFFJCC8QDF5P5lkQigF3lk83BCMKCCkggxvPYXnit6DIMFZf6D4jHFZNFFJCCkggxvPmLtMucfgF717d+fglmggxvPYXnLLzGC2C8NvPCdOzowPCyUaId2KQo.xywzdLruvB7N2rTA78AdruevTgc+TlNr6mxLgc1p7gc1pyF14BLWXmKv7gsmVwCaOsRD15wdRnqGagvmKwhgOWhkBuevS6refh+mKtXy482Nw.0XSCzwWVYmHxfF3RfVfPhiW3uYd9RhqhiZUbQptRqhCTlQntOguT5U8h.kQn5ywWJypkeeTpZb9RYW0IRNEMwN2p3PAzpUhX0y6V8Ch3yPI5WPrne4Fh9AyE8MU9VrgH2tIyOvCDExyPoiGDUvmgRuWJ.cv4A+.LXAmGVv4gEbdtoq7Ne+7.CwQYNVv4gcs0t+ez09dhrfyCK37Ld3IxBNOrfyy3gmHK37vBNOiGdhrfyCK37LR7D6y.QA1ujEHJFYAhB9Ip.QA1FwBDEOlBDERr.QAKPTLwrqbgcgWB9erKr.QA6BuvtvKC6C5+220UYjfEHJ5iUYvEAbLIFNFCGigicOgiwBFDCKbrzLbLFNFCGahCG6gb.YHJ3XYX3XLbLFN1iHbrw+fhPTvwxxvwX3XLbLFN1XTfIf8nrlr1q1GxuIDxGFE+n8gQMG9gQE4WmT5ayqSJys60Ik8185jX2MoGX2WyQ7aTi4ELV5ETWVwz3TECcaSCG7z4cxAAqnK6buw3KgSKHxcorVy.KonNTEdphRPV0QEkhZESG0JlIpULaTqXtnVw7QshqeyUDe2q1tosQc2YcywU5n8bGwH1dsFwXZt+OHGS6jI

I’ve run into a couple of questions / obstacles with the Routing Matrix approach:

At first I thought I could access it through Synth.getEffect(), but that didn’t work. After searching the forum, I found that Synth.getRoutingMatrix() seems to be the correct method. Is that expected, or am I misunderstanding how Routing Matrix modules should be accessed?

When I use removeConnection() or removeSendConnection(), the signal seems to get routed back to channels 0/1 automatically, and .clear() doesn’t seem to do anything. Is that expected behaviour?

In this example, I added two buttons just for the first kick mic:

one connects / disconnects that mic to the main bus the other connects / disconnects it to its own aux bus

I’m not fully sure this is actually the best feature design. It might make more sense to have a single "THRU" button for the whole kit that mutes / unmutes the main output using send fxs, instead of handling everything with Routing Matrix instances. But I’m also using this project as a learning exercise, so I’m experimenting a bit to understand the options better.

Overall, does this structure and routing approach make sense?

PS:

The code is still pretty rough, I’m just testing ideas quickly for now. In this example, the rack tom is not routed properly. I still haven’t compiled HISE for multi-output or exported a VST3 to test in the DAW, so before going further I’d like to understand what the best routing approach would be.

Thanks again!

@Lindon Ok well more investigations..... and a "work around" solution....

So you need to put your Hardcoded Polyphonic FX in an Effects chain in the same container as your Gain AHDSR envelope, and add a small (around 100-150) ms of release and this clicking goes away....

So for example lets say you have this structure:
2201b459-a17a-4291-a850-4dcd9d00ca65-image.png

You must put your AHDSR envelope (here AHDSR Envelope1) in the Syntesiser Group1 where you must also place your Hardcoded Polyphonic FX

Put the FX down in the Waveform Generator and leave the AHDSR where it is? - nope no worky...
Put the Poly Hardcoded FX down in the Waveform Generator and leave the AHDR where it is? - again - not gonna work reliably....

@elemen8t
i can only get it to work using the method david mentioned:

Select all the samples, right click in the mapping window, Tools > Merge into multi mic samples

the samples do have to be ...exactly the same length for each mic position.

also, doesnt work in juce 8 last time i checked. if you are using juce 8.

@Christoph-Hart That was totally my fault again. It turned out to be the default envelope with a 5 ms attack 🤦♂ . I literally spent a few days searching the forum and trying different solutions.

Just a little bump to this topic. I believe test begins to run, but the impulse doesn't pass through and the Alert Window fails to show up reporting the latency samples.

@OstinTheKr0t

a no - you need to create a third party c++ node template

--> 1.) Go to Tools -> create C++ third party node template
--> 2.) name it EXACTLY: chipTuner
--> 3.) this will create the chipTuner.h file. Open it with Visual Studio or even Notepad and clear everything
--> 4.) Copy the code below into the file, save and close it
--> 5.) back in HISE go to Export-> Compile DSP Networks as DLL
--> 6.) You can now go into the FX Section and load a Hardcoded Master Effect -> there you can choose the chiptuner effect

My advise, play with settings like Bit depth = 4 and SR Reduction = 8

#pragma once #include <JuceHeader.h> namespace project { using namespace juce; using namespace hise; using namespace scriptnode; template <int NV> struct chipTuner : public data::base { SNEX_NODE(chipTuner); struct MetadataClass { SN_NODE_ID("chipTuner"); }; static constexpr bool isModNode() { return false; } static constexpr bool isPolyphonic() { return NV > 1; } static constexpr bool hasTail() { return false; } static constexpr bool isSuspendedOnSilence() { return true; } static constexpr int getFixChannelAmount() { return 2; } static constexpr int NumTables = 0; static constexpr int NumSliderPacks = 0; static constexpr int NumAudioFiles = 0; static constexpr int NumFilters = 0; static constexpr int NumDisplayBuffers = 0; // ------------------------------------------------------------------------- // Parameter State // ------------------------------------------------------------------------- float bitDepth = 8.0f; // 1 – 16 Bit float srDivider = 1.0f; // 1 – 32 (sample rate reduction factor) // Sample-Hold state float holdLeft = 0.0f; float holdRight = 0.0f; int holdCounter = 0; // ------------------------------------------------------------------------- // Callbacks // ------------------------------------------------------------------------- void prepare(PrepareSpecs ) { reset(); } void reset() { holdLeft = 0.0f; holdRight = 0.0f; holdCounter = 0; } void handleHiseEvent(HiseEvent& ) {} template <typename T> void process(T& data) { static constexpr int NumChannels = getFixChannelAmount(); auto& fixData = data.template as<ProcessData<NumChannels>>(); auto fd = fixData.toFrameData(); while (fd.next()) processFrame(fd.toSpan()); } template <typename T> void processFrame(T& data) { // --- Sample Rate Reduction (Sample & Hold) --- int divider = jmax(1, (int)srDivider); if (holdCounter <= 0) { holdLeft = data[0]; holdRight = data[1]; holdCounter = divider; } holdCounter--; float outL = holdLeft; float outR = holdRight; // --- Bit Crusher --- float steps = std::pow(2.0f, bitDepth) - 1.0f; // eg 8 Bit → 255 steps steps = jmax(1.0f, steps); outL = std::round(outL * steps) / steps; outR = std::round(outR * steps) / steps; data[0] = outL; data[1] = outR; } int handleModulation(double& /*value*/) { return 0; } void setExternalData(const ExternalData& /*data*/, int /*index*/) {} template <int P> void setParameter(double v) { if constexpr (P == 0) // Bit Depth (1 – 16) bitDepth = (float)jlimit(1.0, 16.0, v); if constexpr (P == 1) // SR Divider (1 – 32) srDivider = (float)jlimit(1.0, 32.0, v); } void createParameters(ParameterDataList& data) { // Bit Depth { parameter::data p("Bit Depth", { 1.0, 16.0, 1.0 }); registerCallback<0>(p); p.setDefaultValue(8.0); p.setParameterValueNames({}); data.add(std::move(p)); } // Sample Rate Reduction { parameter::data p("SR Reduction", { 1.0, 32.0, 1.0 }); registerCallback<1>(p); p.setDefaultValue(4.0); data.add(std::move(p)); } } }; }

@dannytaurus said in Loading wavetables by drag-and-drop takes a long time??:

@Orvillain Weird. I'll try it out here again and see if I get the same result.

Is yours a standard mono, 'power of 2' file?

Yeppers.

@David-Healey ill search that tysm🫡😊

For me it crashes whenever I try to load a 44.1kHz sample into it.

@Christoph-Hart

It's like this ..... I have an external arppegiator triggering my synth... I click my left/right arrows to change preset.... and because I use a custom data model, I have to tap into the pre/post callbacks.... so here's what I get:

synth notes triggering.... click the arrow....
Interface: preLoadCallback triggered - no synth notes triggering when this is running
Interface: onPresetLoad triggered - no synth notes triggering when this is running
Once the onPresetLoad method is finished, a midi note does sneak through into the synth....

Then this callback fires:
Interface: postLoadCallback triggered
This kills the previous notes, and triggers the new ones.....

Here is my full loadGlobalPreset method:

inline function loadGlobalPreset(obj) { local samplemaps = obj.samplemaps; local wavetables = obj.wavetables; local params = obj.parameters; local fxSelections = obj.fxSelections; local fxChainOrder = obj.fxChainOrder; lastLoadParams = params; // Restore samplemaps UISoundSelector.syncSamplerMenu(1, samplemaps[0]); UISoundSelector.syncSamplerMenu(2, samplemaps[1]); UISoundSelector.syncSamplerMenu(3, samplemaps[2]); // Restore wavetables UISoundSelector.syncSynthMenu(1, wavetables[0]); UISoundSelector.syncSynthMenu(2, wavetables[1]); UISoundSelector.syncSynthMenu(3, wavetables[2]); // Update all UI parameters - except the ones that are not tagged as saveInPreset UserPresetHandler.updateSaveInPresetComponents(params); // TODO: Restore custom samples // Fix-up FX menus by stable id, but only when they differ if (isDefined(fxSelections)) { for (i = 0; i < fxSelections.length; i++) { local sel = fxSelections[i]; if (!isDefined(sel) || !isDefined(sel.id)) continue; local targetId = (isDefined(sel.idName) && sel.idName != "") ? sel.idName : "empty"; local menu = Content.getComponent(sel.id); if (!isDefined(menu)) continue; // what saveInPreset restored (by index) local currentId = UIEffectDropDownMenu.getIdForIndex(menu.getValue()); if (currentId == undefined) currentId = "empty"; // only fire callback if mismatch if (currentId != targetId) UIEffectDropDownMenu.setMenuToId(sel.id, targetId, true); } } // Restore FX chain ordering (pageKey -> [4 slots]) if (isDefined(fxChainOrder)) { for (k in fxOrderKeys) { local key = fxOrderKeys[k]; local saved = fxChainOrder[key]; // expect an array of length 4 with unique 0..3 if (!isDefined(saved) || saved.length != 4) continue; UIEffectReordering.pageOrder[key] = saved; // update UI state UIEffectReordering.applyVisualOrder(key); // move panels PluginEffectReorder.apply(key, saved); // set DSP chain } } // Update all UI parameters - except the ones that are not tagged as saveInPreset //UserPresetHandler.updateSaveInPresetComponents(params); }

It is doing quite a lot... and ultimately what happens is when I switch a preset, I get one voice that sounds one way... and then another voice that sounds completely different... like a voice is being allowed to be triggered before the preset is fully loaded.

It seems to be something related to my effect menus and/or effect re-ordering.

It is hard to explain. Might have to make a video. But any immediate thoughts??

@David-Healey Thank you! Thats very helpful

@Christoph-Hart You are the best. I'll give this a try later on! Thank you as always.

@Christoph-Hart Thanks for taking a look! 🙏

Any news on this? I plan to push some other webview feature updates end of March.

Someone just sent me a link to this free AI course from Anthropic https://anthropic.skilljar.com/

@dannytaurus weird then, I'm not sure why it happens. I'll leave my setup, I added MSVC V143 in the list compared to the default settings.

VS2026Setup.png

@Christoph-Hart Considering my post right above about the issues and limitation of the scriptnode table, could the upgraded UI table be ported to scriptnode with at least a default behaviour if a customisable one isn't convenient to add?

Linking a UI table for each one we want in scriptnode is not really an option so much it is cumbersome during development...

@l4ch If you're on Mac, try opening the XML in TextEdit then go to File > Revert To... > Browse all versions.

It might have some version history saved by macOS that you can restore from.

@pcs800 Or an array of buffers for stero/multichannel