@Lindon Would it be possible that the single error that both you and @JulesV are experiencing is because of a drift between the plugin parameter value stored in the DAW project and the internal preset?

Whenever you save a DAW preset, it will

• create a .preset containing the plugin state (exactly as if you would save a user preset)
• store the values of all registered plugin parameters

and dump that into the binary data blob stored as DAW project.

In a "healthy" plugin, these values should not drift - the plugin parameter value is the value stored into the preset. But if there is a mismatch (either because of a new plugin parameter that wasn't stored or because of some initialisation error at startup, it might create these issues.

@ulrik of course. Git is the single tool across all software development that pays off. I couldn‘t imagine doing any non toy project without version control. How are you tracking your progress?

I'm testing pretty much this setup:

1. global mod => LFO
2. hardcoded FX in master chain with network:
   • 1 parameter, set to combined
   • modulates a osc.freqRatio
   • extra_mod slot has a MatrixModulator connected to the LFO
   • no preset, nothing, just the default state

Load VST3 instance => pitch wobbles. Before the fix it didn't.

@ulrik said in No sound from HISE Controller on macOS:

I'm not using git

lol why?

Fixed on develop. There was a build-flag mismatch that prevented all incoming buffer content to be propagated through the signal path in the HISE standalone build configuration.

That affected the HISE controller playback as well as the latency check - which is also fixed now.

@ustk no my test never touches that codepath. I'm loading a compiled network in a hardcoded fx - all your changes are modifying the script FX modules. What's your exact setup?

@ustk ah yes sure, now I'm back on track. Let me just check the current example as exported plugin, but I checked this exact setup back when I fixed it last week.

@ustk but that's not the hardcoded master FX - just tried again with a simple dsp network that modulates an osc with an extramod - it initialises the connection correctly. We're talking about loading the project fresh in HISE, right?

But your fix look good, no reason to call this with a samplerate of zero.

@ustk hmm, you sure? bypassing / unbypassing just calls reset() which shouldn't change the connections...

@JulesV Are you using plugin parameters? And is this affecting only controls that are linked to plugin parameters or all values?

@Lindon yes send over.

So your particular problem is that plugin parameters + macros + preset recall is not working (mastervolume +30 = 1.0 normalized)?

@Lindon yes but please give me some guidance on what to look for. Also if you can dumb it down so I don't have to download gigabytes of samples, would be great.

Did you use the actual processBlock method where you shuffled the buffers into a Libraries.load() module?

Also the script is completely empty, all you're doing here is declaring some parameters in the XML.

I wanted a simple gain that can handle multiple channels.

If you were stuck with the "script" part in scriptnode, use that instead:

HiseSnippet 1249.3ocuW8zSiaDEelDb1krzpcUaumaksBgBfS.odfPHv1nVfHbJp8Dch8jjQwdFK6w.oU8du0y6s9QnW6M9Nz9AoeCnuwyjDmD1.Dw1fDxu2ad9868+wshDtz3XQDBWr8vPJBulkyPtr+g8ILNpYCD9SsNgDKoQkzrpOLjDGS8PXb92oXfWcET5u+c+5DeB2kNgEBcgf4R+NV.SNgaXsuk46eLwi1lEj4z10Z5J3GJ7EI.dxaUFERbGP5QOknNVNKz2Ph6ivekkmcUJw11yktcWp8dU5r6N6rmaU6xdkqtMsR0tjxU10tqMBW3HOlTD4HIRZL7RqK7F5zWbMWafKXwrN9TEwVHGvxZ1nC6y78ZMJ3DiP3UZMITkWGp9bqSXdrw7mDxdcpfRSzHaPCmaQPZqm.jvYfzJZH8FKG2HVnbhDEddkUSNjA6RfbSVnnOKBeq0gB3.b4lAjAzii.hwJrd0xk2nD7u290cS3tRlfWRvOUHomwW+sE+khqV7WKVZVQc6duxTlIR36SitWwpxgnEo357jfNznMJcEwOgN9ff6OcLsviKl5p85LGTvaxYxyBoF5iE9dpXk544y.HSXCd56a1fHIpjhgGbtPZjjofCtA8JnMPmhV0pAMdfTDhPym+fJGgWhOQNc4jpQyH.hGSkCUIJdLSNLai3SnFq7BqwdrP7MVsXR292OFycOXDhTeLvnoy7SrNpaWpqbB.Ww53eXYaCK+jaCKZZCAap6+zjklACi5+9MqI05gQzPRDssnkOY35wjfPe54.D2nTGeg6.G1OSmuAITif5pSrtaeBmS8iWl9nBOSylPmKRjLduSHxH1MPB4zj.GXrtK8PC5PXKbAUKjltrhVUV3P4doD2A+LB2RQiMB2ZFgaqnyYDt8LB2QQm2HbmYDZqnWwHzdFgUTzVFgUFILSG+oT40hnAoU.lmQ3WpS2woI1K6dC5.eew0GJBBYl9BHsmxqkveXXeAm4pXoOwnfyAAhDvFF+ZRSU8QU.iBVv9v1DlupGyIIFlZ4cF2AdOoKgw4ufFEmZyWXUdS3Onm6TgGH5EGSbgz7vVDYeUyqZPHzoPi1zcbGSVuHaQqdYbVop24ns1ZbAtSfJJo1YqJeZQh.sfMJpgg4v4yLbT2+nnfoC4TFdMqlwWnj5R7QWnJPUJkZEcvN0EJnNZAqdJ3Nk27Jvahna1a1KqnwsQiECYbtr.DkA8P89XJvTmv3F.NZt1Od6IjalgGTCIog57VJu+38u+u12Y.8ZMvMbus2e9E0zdU5EqZSuQ1Vj9tfptq.3nroZ8gKqC0GfUO8shbR57tHQRXJwb1tAsKIwWNG+CRjh.nMOcBMJiWUbduBJBm0qp860dTd0O8YM968MCACDBYeXpvGv0JXkdevGkeYW698K6ZOfq7bjfVy5bZL0X5Oxoooq9vYbt0xTq776mvTmoaPN5Fvpbhelc7vdtFrXBrQv6C3lS4UOXnJKb+m6t6KuM81TbZ51qXykqLjlYYM8F2SONzn4kFWP0EhAAjzwzK2cz9edQFZQKxvKZQFZQKxvysHKf3FItzUee.Ur8kobfAb7zAkqBe1GPWZKzUyVDE.ebykttS+plSwsWVE2YYUzdYUrxxpX0kUwcWVE26gUT8smlg6p9DX1cqiRK7w3i35lUXcH5+.8tIYVB

The code that sends the macro value as plugin parameter is already in HISE 4.1.0.

What changed in that commit is that HISE registers a plugin parameter at index 0 for the UI control (basically what the compiled plugin is doing) so before that value was sent out to nirvana (or the DAW if HISE runs as a plugin), but now it loops back to the first UI component that is defined as plugin parameter.

HISE_MACROS_ARE_PLUGIN_PARAMETERS=1

This solves the problem my registering the 8 macro controls as plugin parameters, so changing the first macro just changes the first plugin parameter (=macro) => no op.

So the proper fix would be:

if USE_BACKEND && HISE_MACROS_ARE_PLUGIN_PARAMETERS==0 => don't send the macro value as plugin parameter

if USE_BACKEND && HISE_MACROS_ARE_PLUGIN_PARAMETERS==1 => send it

Note that the compiled plugin is completely unaffected by this. All it would break is the behaviour of HISE to send out macro values as plugin parameter without that flag.

I saw it until I added a DspNetwork with a Gain parameter.

I totally overlooked that you added UI controls to the script FX like only a madman would do... Why?

Also the script FX doesn't contain a setForwardParametersToNetwork(false) (or whatever it's called) call so they are dead in the water anyways?

What's your use case? I think there's a valid issue with the metadata system only looking in the DspNetwork to build up the parameter metadata list (so the fix would be to also check the UI controls in there.

@David-Healey yes I saw that - the reason for this is that HISE registers all macros as plugin parameters in the backend by default which then loop back to update their value.

It should not affect exported plugins though, but it's definitely a weird glitch that I'll look into. Is this a regression? The backend macro -> plugin parameter connection is ancient (about as old as HISE).

@David-Healey hmm, works, your snippet didn't include a network with a Gain parameter though - can you load this one?

HiseSnippet 1534.3ocuX80TabCDWmMBBNoYZloe.byzGb5jwXCDfYxzACF6T2veb4bnvST46jwpbmjm6zEvMMS62f9bdqeM5a7onO2OB4aP5Jo6ruCLDOdRiYFvRq1U+Vs69a2j1ABGZXnH.YUnyvATj0Cv1C4x906SXbTqcPVODuGITRCJZ1Z6gCHggTWjkU9Wn1vZw4P5OueysIdDtCc7VHzQBlCcWlOSNd2A0dIyyqIwk1g4m5zqVqkifWW3Ih.7jGWAMf3bN4L59D0wxgQeOIrOx5awqs9F8VsBsJwoR0UVaipqubkkojUd15qub0UVY8Uc6sV2JNarLxZ9FtLoHvVRjzPj0baKbGZ2WbA2bAGwBYc8npEUQ1vMa1tovyU4hpcQ06y7bam7PEh.qzd7yVdyy1Wg2i4xFs+3muuTKn3XMR+.ZkKK7xmAdUSCuJof2DfjUJHMmAROBa6DvFHGKQgm6iawgnYOBDmRCEyYQ49GKbcAbBtrrO4bZy.XwHMJsVkJOsH7qm77BEfXUnrnQulGW76JpSbJeFU1nWOpirziMxJ173GqN+RK80EeIWzsZrl5uCpkbcfh0E9CDbXQoGqkpzS+kxgJgbYfvqNwyqKjWTRv0hh2VcCvWCEdzxCBXfIRPVYUZpRLi6w3zh8h3NRlfWLqAJ4jb4Os3qIdQzmT3MEVbjQ..rkTFv5FIoYMcxwedg2VHks2WHoGvKosBHo30E0q2DkEiFOZvDEqpXBtKEKwi76RCR6CpCBYEYS0v2dpV5JAGSvI0AE7Vbl7fAT9sUefhinpzxXTAGUpSJeXbRosGykFfXPt2BXcX.oAbLav+9geeyqPup0NDIIwLfEgaY.MPxTNf0NzWCbKlb8Ew6PCOWJFnOabbD3z9nW5kioe1u1vQK9ierlOy00i1VDxTOviO1xWArgCrY+ZBV+q28t+dS+zLgmbkO4xzrfgQ85wfcv3htai7EtwP1g0k5gP2r1EXMDtQdDYVpDU5Vr.ElRW+ppQ4.VGllP9SF+xzBwGgayjN8mLFyMALBA2+OvXLq7WfMTQiA3b3lG+4iBt.NgoHl6cDi3j3dsN.+lFG2owg6u0tmZW+vVs6712z9vC9gF06bZyC1cmFG91C6SFDJbYgxkpKBnKkX+k1KxSx.+jyodpzjx+R3M59ke5p4mvS6ghHIie1dDfATkGuejuMzk1gFeiJBEqbpRey5Jp0pnqMk6pW7A3Srvpp0VwBqlHLEqwbSAqw75hgLjFvm6cEZJzEhC9BgrO3QWy.0D0PSEk0Owbk8u9sexTo7h3j4jtl5JNHp7BQv45zo3uirtmI2ITanS6cIZKOOwEJZNVbQFjCo2qsva3f9BNyQsk4DIgns7EQJNQSbZbE51dBmyMzYlPFLjUGBySUvZGEB77tGvsA6nmryJ+QzfP8ct.tRY3Gn.deMg1BMINPx1v1D3kAXBTsNfPDMnrynxuzdQ5J.yDdokprYxnfFbAtiulUGFDTkD2lD.ZASlnZFjyJeplClhQ0JfpIm5he.tU3QJoNDOzQl2c3gPcKlGasKLu5nXrejGJiyT.6C+srZ+ahZy4ua7ZkKM5PoftBpIqfWf8X7itdVwdjKu1dueS6r8fzerOmdgA1oOoxkV.aLPG5kxNB82gD0WC3Qeo3Wwco8fPkKfryLiZaG08EAhnA5E2352g1i.DN2X+shjB3oxP2e6tYAEnhqf+z6wPtc1HQiKgakS7R0VBBo6vBI.62s4yYbQyHHbpdXqvrKiK.Z4ljJLxQa4N9oGssPbtOQWaOaC1LWpAat+GiWL03GeSsL8a+zMWSoZ28bMtSLIQSQOxQVbpHoScotSm2.ybMIuIE7q8m0lH.gV.FGYdr9edZZztvGsivX6+ygSGROoVlGU1jQko2xTLX63tKiM5U+1L9noKmtqTfSFkBjGuabB.F2lv+7LxfOwIPbpiomqpr7d5c.zv0TyKh2SstX0a1pEdCXm53j0T2PwkmUEWYVUb0YUwmMqJt1rp35ypha7wUT8eZRbyDUJJzRncCMmokUCtg1FZ.i9ODdN.kA

(the 300 extra bytes indicate that this includes an embedded network).

Have you switched this somehow?

DOUBLE_TO_STRING_DIGITS

should default to 8.

looks like it's trying to use the full double precision logic when converting single float numbers to a string.

Guys I really need more to work with. Just stating "it breaks older presets" or "it seems to be upsetting HISE" gives me absolutely no idea what to look out for. I get it has something to do with macro modulation and plugin parameters, but it's a shot in the dark until I get more context.

• what does break
• ideally a snippet, if not a DETAILED list to reproduce the setup
• what does break mean? crash? values not restored? weird couplings between UI elements?

Ensuring backwards compatibility is the number one priority of all bug reports, so I'm on it as soon as I get enough info.

@Christoph-Hart ah nevermind, xcbeautify is definitely not codesigned - I don't codesign other binaries lol. But the projucer should be notarized & codesigned.

It's a setting you have to deactivate somehow.

Note that the hise-cli export command does not require xcbeautify anymore.