event.drag has a "dead zone"

[ustk]

Hi,

I have an issue when dragging a panel with event.drag (not allowDragging property)

The panel is moving as I want, but you always need to "unstick" it from its position before it moves.
There is a "dead zone" of 5px that you need to pass when dragging, otherwise, the panel sticks to its position.
it is the same when coming back to the starting point, like if the panel is hitting something 5px ahead, both in X & Y...

What is even more strange is that it doesn't do that with the allowDragging property...
can I do something so I can have precise movement?

Here's a quick snippet

HiseSnippet 874.3ocsVs0aSCCE1oaFQKTIPBwC7jEOfZEio1w3hzDBXsaPErQDEPaOgbSbaslicjiyFEz9Oy+fwwwIKoiRE8g4Gp74xmymO2b80p.VRhRi7p+kYwLj2swCmIMS6MkxknA8Qd2GCNP4jXpjIHgZ5jIb4DztyhoIIrPjm2Zuy5qW80QYqe+5coBpLfUpBg9lhGv9HOhaJ05+lOvEh8ogruvip381uYPfR1SITo.uVC2AESCNgNgcH05VMLx6F6ExMJ8PC0vRPdquqJb1vopyjN++FOgORvrBcQCgCxodekHzxXqVTuobQnew8OAAmheYzXMWz3d3C3g7K0WFUtSlARIhpwCuZKidcqRuN++zyqB8V2Qu6hGFn4wlRKVtcK7.ogoGSgTPUZ47EUyuFtmB7PZ1LhdBaeMHbIhVc6zYCB7S6cZ1nYCHQjXHmR0Dea9uK4UjBrSXldpnXkDDZ8Pm4G5P4D1LgYNPklv5QEhQPJr03TYfgqjsXmBfZ+qlMH4qrM7wDmkMskYUMWtyREI6ri.hj+U.d7NgZDUrYrJgaO9iHOlTdNGsyUAe7x.e7bfOdmqvQ6p754mCpkkRaboC+0x9QWh4m0YoFa6tAm23716TDBIJ4gJC6SxVsa7qF0abdiqZY73EYxl8zJgfoWjUaindIvZISiFwza.QRQJqvOndb9Bd7+tfuZ+XfqRphiJ4.I27oXl7e0FfxK+fcecPepgZ6Lx0A9EyzFtkBd8YmBibb8I0w8YImXTwY9lWyBMJlLqMK5hroUDG5UtI1khQ+n5LrYUENiGZlB2Tus.goL9jolBInZWcVQMOPkGf6I3vtMHuWcJSSdDoewTzQJcHS+YZHOMo5o6zOj+y4FhxMrnhwh0uABcwEyqcqppQ+83BXnkJLUPMyOIyN9N2.jrmajgcrfDJwmUc790x3s+W5dWrO2DLcw7s1B3KTebcy27GKZh2a7XVfojrqi2+nq+WFPeVkZfpoCnFMGpXwGlFMDJGBX.SjPcrsizqls2wI2wJaiLCYxvLgKfUtwtVYubicKLhhnAZ02CbiArOGcyLM.mjYuJWGefUlzEkMZvw6H3MxuGDXCDOAX9hQr0Ji3oqLhsWYDOakQ77UFwKVYDubIHr+gj2lZTQt1BPg+dYCH871SRgpprNDze.5cxmkC

[ustk]

No idea of what I can do? It is very annoying as I need precise settings…

Or, another way would be using the allowDragging property, but I can't find a way to disable the property on the fly. It doesn't work with a JSON property change inside an event in a mouse callback, and I don't think it is the good way to proceed anyway…

Cheers

[dustbro]

yeah... there's something a little sticky in the top left corner. Doesn't happen on the Canvas, but exists on the Interface preview

[ustk]

Yes in fact this zone moves wherever the new panel position is, so you can never replace the panel at its origin

[ustk]

event.drag doesn't trigger before reaching a displacement of 5px

simply attaching a variable to event.drag shows it goes from 0 to 5 with no transition
and 5 to -5 in the other direction which makes a 10x10 dead square

[Christoph Hart]

There‘s a threshhold to distinguish mouse clicks from drags, so most probably you‘re hitting this. Has this a real world implication?

[ustk]

Thanks Christoph,

Yes I am using the movement of a panel for controlling parameters, like frequency, gain, and Q (for the latter, the panel stops via event.ctrlDown) so the "dead zone" is critical, and moreover, you can't go back at the starting point, so it makes any setting quite difficult.

I just tried by the mean of the allowDragging property, and it is the same...
I mean, the panel moves smoothly from the first pixel, but the movement detection still has a 5px threshold due to the event.drag in the mouse callback

[ustk]

[orange]

@ustk Very cool looking eq curve. This is a cool way but I think it's time to use GUI functions of the parametric eq unit in HISE GUI editor. @Christoph-Hart ? HISE need this

[ustk]

That's right it would be very cool

I am thinking to another use of moving the panels, so it would be cool to have the option to remove this threshold

[ustk]

@Christoph-Hart I'm coming back on this "issue"
The dead zone of 5 pixels when dragging into a panel is still annoying because you cannot make small adjustments, for instance, with a vector knob.
The value doesn't change, then jumps once you've reached 5px in X or Y.
This dead zone stays active when you come back on this 5px square without releasing the click and add 5 more px the other way. This creates a 10px "glitch like" behaviour where you have no possibility to come back at the starting value.

I know this allows to differentiate a click from drag, but it is so annoying with knobs, and even more with graph EQ like above...

[Fortune]

New Draggable Filter Panel floating looks very good and usable, thanks @Christoph-Hart

• Can we disable adding new nodes by the user or limit the number of the nodes?

• Can we link each node to a panel (I will make panels that contains Q, gain, type, freqq knobs for each nodes) So when the user click the node, it will make dedicated panel visible. I thought that because we can't adjust q value from this floating tile.

[Steve Mohican]

I made this but the node does not matched unfortunately; @ustk @Christoph-Hart @d-healey @orange @ulrik

HiseSnippet 2106.3oc2Y80baabDGTRvIhoxw4OSlN8oKZZ6.FIQQPKm3F2NFVThNZhjEqnrbx3wiGHfij2HPbv.fxh0QS6L8k1OB4M+RmoeE5S0eG5W.2uA449hyt2Q.bfBRlR0VMS4HSIr2t29a+ys6dvsB4NznHdnVoY2cX.UqzOSu8P+3dM5Yy701XMsRk0269OdKZLMTa0gA1QQTWsRkl9t35klcFMwme31qZ6Y66PyHoosGm4P2j0mEmQsk0Wy77ZZ6R2k0Wg6Ur1vg62f6wG.XYZ8ZZA1NGX2kdOajsoz0Jck0cYw7v1w1wzHsRyrJ2cX6d7m5K4eOVDaeOJ9foVaXijjax8bQDiT0Zzi441JwlizzJo2JyCLszC7w5awbYozy7DWSr.ISBU+QooNK3YpBuZSN7Jo.uYjv6Cza6DxBhyVAw16ougODh5XCg.UXI4Uap+yUzavAN7iq129.ZyP3gTILtYsZKRV4F0pbq4JOWYHPDESNzNjf1KJVH2yiFZR9cDQxQ0tz3s3tC7rAC1XdgWQgs4ws4z1j5S1lTerM4a91V1ts78.oSrCP9F79Abe3Ai4SXX9wsgV1gPJTbH6Iq+6yYAq2oC0AjLGCmP7C746GAh8vGIVnCOjXvfmqcKBi7aI0gesvBUlq7ylqLA9HX+grGcp37qAFleAC1BlUPUoJS0HjSgGngsm29vA.CgYU4VkOFU9xKCVCyOlDxGDy7oyUNwpQQEKsibEiNC7chYbeitUFgL42cq1AN+cGOOiZGYBgc7SBNN42ckPBOTZH+UT0tgzgEIA5rNBL6srA2anseWpgzvp8HzErms2.pQExxj5nNIeFItGKBW4AL23dFUVj.4fiQaIyTvga+PX6S33qnrt8hgMboSpQySnQE0kH3hDSEElPMQci6wV2yiEDQMd3QKRFJE0r1iv.iLgABMawGDQIMVMWPQPLMZlFUnGBICUJ+LgFXcHRBUc7XNGPcIe22QjDbCs6VQvzyxvk.yt1w1UGyeKk4HvhKx2ZpZaY6wvh1igJ6QlCaIyw1kr+JMTGk33Eg4OSAqpBlFmTXeI3eJ7OrxIE.1emdHLcMJb2xVTr1w3WIQHluGbvfjDBjUTLbRNYtHjgA3.OIOxr73N1dDlK5jEZnJy2kdz1cxDJ+4.LN9o.+PJ2xMCoOQ0OMatpLnYemX3g8GDSMxuDJIgr.AyYyuxp19ta2oCHaBXklI0Cx6VdYribhJmXsgBM4ZSt8o42gz.rlC5tO9VoNVt+83wzs8Mv76YKebYx3K0oSgqkU8uvkwAFBOKAM7GzeeZXBbSXD5blu0r9o2ZVcxAGY8aEF49a3yh2Nf5eZMr0FUzGZcO0HTArFKZbe0QMta6wbggoXP+42QGaGXpI.b5vP8dgk1DKb8Lg+x+9eF9YdKs6uwZvIHb7gQvAfX.MLlgVeo0nGBykIGlXV80nQGDyC.mTZSJXZh3QSBIUZSOtMzUo6tLvKgp980UIYpcD3TKgPeXxe7TrvS13c+UqdhpHoTd4.qQXSBjOQeMnTWWaHNzj4Aykzx1m5ose2jQBkaKKl1Oixqd0q9iYTpOhoqKL9R2TWTxb9zvyFty+kjw51unfkMvy0vhKYNW4iwXXpinrzQLWxLWBPgdfx5ImBPiOal2gpOLANAny.+oI8GfnyuPuA1CHZQxWwOjFR90DgeA7zZA7fAAa6uCtEBlv46xbPydEvcH11X5QwmjppY8Nu1LREiZMqLi5EGZEESCZy9CT0g+6amw+e4eY0W8pAqXcdzbcEM+x9mOMCpJuleg1IGeFFhOYnS0I6w5fiV.pojaDZbLY+HV7PU89VYb+IEtefdKVrSuhw6TEfWrbzaY7N5xSyoKGqNCryn27atn2Tp3KxU60cSoxIE2ZLH7P55OQfi2WO2Q+bP5dC5Kq5fcGJgEQvte0F4.svhpHAS0RzHg54uAru60GmvJYDtlUyl6ttOVhS5GlNueX5IyOTPXPNzeWXJtPFVJFLm1voeGJDE7gJVBqZJrDf74ZIkSaSEVonR5qFsnYhK.WzLYQ0P8Ukt22SuMLPk35TBG7mHdl7.6Coj6R8ogXRq4o8RCd92+8+iI8kFDLwuzfschA0uKLJaT.ORsRw+rMsOaWnJTjp1teDEm2ZG7TjJ8Fb6vBW5GtcSvHKTl11wCBEGGuSe9.+3bG+l9+xb9Be6.y7S5WdwkWw2omL3VfGbpVm3EfcMHMtefGcc+CodvXSBL9gvbRcrG3EmPMed8VbedPOtOyQMnuCVqoaWZnJ1Kzffoysw94IT9XqcndTa0D3eo0lPhmcH3mnWPewYWEon20TgwqettDtD7vH4mTMMGUYRWeylaOJ2B9KRpQX95v3KuMdzd.02Iin0JVXomMTlrfYgU45vC6OpSSZI9MocAEoZz6Rggefpj4RMZ2mygq5hyOmVP6428e+QVax4Ao8Hx1jV8fjA4swxUlZPTLueBVDulz1vPRxAfen9muRU3SmEvu+MMfut91uYdV6MWV2TSXV2GIhjogLLldYVnnfwKjUTytz5nBEi+1Pe8nDZBsqs.FpkIfo+iUSBzzx1z6Itna5Z+oGXISnx2djYMpn0dpWu7ucimZksSBEKyW9U50E4KJeNbr3+S19MWjW6bE3SOV9+x.+rmm.e8KRfWS6sVf+4Ge0O05RO5clWUY5+O5pJmCnbYLt9kgN5a6DxeriL4DOq7tBJfc6K9OsaV8svmIoueKc8ZUqo0GNm7XGGzcuD3eJVl5W.Yt9EPlUt.xbiKfLe9EPlu3BHyMOSYvRT2Y.Lxf7nHPn05xh+kTtXp1OhORhfx

[ustk]

@Steve-Mohican Because the value you handle shouldn't be linear but logarithmic.

[ustk]

@Steve-Mohican I've made my own GraphEQ a few months ago before @Christoph-Hart made the curve EQ accessible.
If I knew at that time I wouldn't go through all this pain But I learned so much!

[ulrik]

@ustk that's really good looking!

[ustk]

@ulrik Thanks

[orange]

@Steve-Mohican

Yeah you can use logaritmic or other non-linear equations for this kind of implementations. Here is my method, I used a "Hyperbolic Decline Equation" method for the frequency cordinates.

Cheers

HiseSnippet 2270.3oc2Y07aabbEeojWaK5p7QgKPPALvTghhkVjT6tjTTr1ElRhTQBIxhUT0IAFFFq1cH4.sbG58CIw3XfbL8TOzK9OGeonG5k.zS8T8eAE9dOn7lY3xcVJZIFAXijPCKw48wLu2u27dyaF0xmZiCBn9JYV3fgCvJY9UpsG5E1aydVDOkcZ.zUa9rm5QcvJaLbfUP.1QISl4+TF6LKbME9m27fMrbs7rwIjTTdDkXi+bReRXB0V0+Lhq6VVN3CH8kjtb8crodaRcoQfoLuptx.K6ir5henESr4TUxb8lNjPpe6PqPbfRlqsA0YX6dzS7Dx+HR.4PWLafgRaXhDj2h55vrX12U1rGw0oUrKGnnjQsUB.Lu..ts5tDGxX5I.wGwYfRzPFOxL2EYdFxlm9TMOioYdYjLuqILuOVsssOYPXBGlscK0c7Bw9crrSGlDxpLWoaptIEjvKrXeqiva4CCFqgV0pUyiJWQO28VLK6ePnHHDcrkO5K+pVVNs7bQ+ITr9cwgaR6Of5ACzVJVfkD5lnYKKeHzE5SdVy+rAnNeWES4lc5fsAMSIvRwKs7TbjG8v.P0G+DNyNTejFAFqeODAcejI7qkWN2hYe9hYQvGt3Ol7j2ps9Yf.KsrFYYibrkSVmhALI8B8otaZ45dHr4Si6ZLAeAa0WYEvkHdgHeZTHwCuX1XWmoKm09BNZch7rCITOst4FYZhe1sXGXy+5ttZ5mZrlN+Srgb9e1UXSrLBMwuBJND65ROYZ5vvqSAOeWK.k8s75h0D9l9SXnvirbivZ4PqfLYqJ5tnvdj.Fmuf3D1SKWdjd9IoUvXr4wl9gvzGKw1XR2dgvDV37qnw4VQokKVw7HiJ4OG03kaRLqoqKYP.V6wmlGMTnpQkmviMhMNPzYWZT.Fs4FoCLbpiCoiiL3igcDIacHcPBREscI1GgcPey2fDDb7s5lSH0ySLNtg6XEZUbBPWnzofaOM.1P1ASligSaNFJMGInVAiIlkjuMNdGDi97X8ckrUYEGGrjDu.7eI4Gl67J.yucOlY5nM0YaRluPXjihSDOWHCAEGGD0WzriyQyCaz.KgEXF4XtTaKWDwgAy70nHwyAe5dcRTJU5PR772AJkiM.1arkO9YvV0sgJn9GRgPLpA1laI6hC6QcXlFR5SpZSL.Z8PXvgQgXszr3SLZYDaKdZNaX44rWmNft4QZZFEWqrgwpqV0vnZIS8ZMWV2HGnCOpOfdBHALKZZELJZVt1ZUqXTS2zTeUilEzKk6tBTYEsBlEqtZMiJ0JWsrdYSSCXdJkC1V.JtxaQY1m6MoChhIfcgjlUVgcbN.PMnA3B6iC.fE9x1Pl2X.J0dsYFe3y6LfOLOa0J0VynrQES8RUqTVfPLLoVwRlUqVtlwZF50JUhgcl4t6XjiiMPRQQ8xqUoZESypvrTasbrROZikpTQFjTAlefcMiUYSxTTBVvKehm.PSP1wEd7wCXGHnMYhp3DtWL9rN9wZw4BTuGRCw64okK6yytP1WjEMIqNclJuQGb4h8mJaVmV9WjhZdQ8OD6Gm7EKHzxQ5dZTe68zH2xks3vWIAod63QB2a.16s0oixnSrgddlajUAhFx634CF0wSaWhC1Wg.M1bCU1Y4FJbCdbWj+y+QckYVYyTJ+sstgZck+xNMfxdr9tFYNfIBULBILuOSC7wPCshtvVPsAN3nP5.triJDAsgIVzEi6QyxC6xWyawa+B1S.1voRMIWenbGymvNpXLgO46p2iW2eLku+qqGhOMLtM4Et9YmcFmAID2WP0jQ992+95oIKS8PpO.E6a4PhBjWcA81juNUW7YTk7uECG0hrv+1xkZAc7z8.BrKf4lenpLISYW8MOHkqNBfEnInFwEZFsAIXfq0PkC6lX0kJUZ0y6KK7nyNdJN9baeVWFYdXLyZp7SrWZ7FscbV5Ohlnwy7bQ1gcnBvrfAjdNQbP+bwg2DUOYgKwVXnFLmy.Kevqj2QHE3Ujn+A+nPxRyNRtdjCgttmk6v.XIm.Iaddj7ku7k2YJHY2tcuSBRVZ5HY7xLIJZdkAwytTPLcBv+5UmcmQnvkmDlUMt.sLbp8pKHGT4uNoY+5n5Vrlvi6oDRg9spax5ZLHOZa5wXeze.0.ZZrKD6TFPGDMXOu8YSAWHQIxjvRVUiUgqeYZTIkmIQdl2RciKsbYhW25D4JOmTuuUBuu6eWuu7k6KWelMgqeoEcSVlAeqjITOsInz3RVyyeuX314TmHWqvzWYm0.xHFvYdotaL69udAjvgxoVuStG+rZterZKRncuoauyME6kcb46X6czqhrnp3d6IF60T25KupOAxOhWnYNIaIa7guaF4eLt4y31wGplpfdJS5gQ8Emrv5dISFvyXscpOB.Q+eyxuhQvXDgW+e+e+s5LBlwsE7e96yyITJ8ac44TNgvGUeqsNnomE3aBbX9z3v7yFNLkvf3AE5Bcj5SfjCUvcZCkGrwPTvCppw8p4XU+Di0YiYlTaL2KUy.MHb1HlFwP.ioQLS4Psp.dypN93CwSKMtLeJrciHXCwDsK7p5s7gtjvmLJPkjV8yUL41BO4Vpeg0wX9aXwAkeCebGpeezmhgJRr7XiK3AR+9Y8ARGLyOP5d1gfIbfukGbcs.rg7LOAOSYdMvgQdoEWPJkTL2aKv8jj6l0iIZJSDNeM0jAiSMS6Rjpq+5GHRTZiAuwYu.avYYXm7dkVQvcR4uZhgrhIjMkIO4NqYrpS5yEjj+mcOR76uyBme1L2K4crWH1FaS5OvE2z6XrKbKKtM9qgqU0wJxMLlZ5bpcodzA8ndDa4MX6yJ82sKN0lno5PqGFZwZAKlxsquO1EaEHkm86q+4PGFV96x9Cdb0vBiYGKtn30mnJLWDqP.5WF8vL+uf5g4cQFJTq7m.Q52GGw99XM5aY6Seps3EgXod2jSA7aO9eUwET2kMFM9cjTU0KpqzGJf+Taa1tiB.9LccLuB5T5JnS4qfNUtB5r5UPmpWAcV6B0gc145Qgz9h8+.gVMEu.QFoFrU9AjroNXE

[clumsybear]

@orange said in event.drag has a "dead zone":

you can use logaritmic or other non-linear equations for this kind of implementations. Here is my method, I used a "Hyperbolic Decline Equation" method for the frequency cordinates.

I have a question regarding the non-linear equation. As I understand, the non-linear equation is using the knob values as coordinates to map them to the logarithmic frequency response of the Curve EQ. So, if the user is trying to automate the knobs, the values would not correspond to the actual frequencies in Hz, right?

How can we show the actual frequencies, not the linear coordinates?

[orange]

@clumsybear Yeah the real values won't show up in DAW's automation track, but on the plugin GUI, you can show the corresponding value in a Label (with using the same equation)