Very strange behaviour with Engine.getUptime()?

[VirtualVirgin]

I am trying to measure and record time deltas for MIDI note on messages with Engine.getUptime() and getting very strange results:

Most of the time, the "currentTime" and "lastTime" are showing as the same here, when I think the logic clearly shows that they should be different.
This means that inside the inline function, the time difference is always evaluating to 0.
But when I am using that inline function to record to an array, is recording non-zero values.
Furthermore, the array and count are set to reset if the timeDelta is greater than 5 seconds, but when even when the array shows that I recorded a timeDelta greater than 5, the array and counter do not reset.

What could be going on with the values to create such an issue?

HiseSnippet 974.3ocsV00aaaCEkxwZXRaCXEX.6Ug9jyRWpsisaxBVWWhc1L1RpQcaQAFFJXotxlHTjBTTcyXH+m2+fMRQaKpXmrNiM8fguebt7vit7RMQJHPdtPh7Be4hL.48o9SWvUyOeNlxQiGh79R+WRSgng.SgidAPDxXPF0Ec1hLbdNDi7716GLI6EzDU97me2YXFlSfJWHzqETB7yzTppx6jm8STF6BbLXVhJ+8d1XhfetfIJzDaO+1nLL4Z7L3JrIsF9neDmOG48U986dDoW7Sh61s+ICH3AvII3jjNGE2q2fiOoyQmf6c7.ncej2GMJlpDxoJrBxQdMOSDuX5bwuwsKvqo4z2w.iQGzT8JacegfEa1hFunymSYwSVoX4HcUlToe6Y0uuv+RZLcs+Jc7yKCDUgvU.8ZTmd6UidcboWaG5sEJ44PolVJ8.+oDIMSUEwvmOweLWAxDr98jKUr4hZbZS+yE5L3pCSwWCWH0FqQzZP61OJR+y9mFFFJgYQLbtprQ4aiZeZoGRgTpQW2oRaYakJcER4LJGhRJ3DEUvilAk.JSo09g+QXP8xLhOSm+g5zdUloVszDHHLPyzbACNLSR4pVOzAyCiNvkIlzuUxqXtIyU+2VUWx5RiudcdaawWixTv0FFgJvQjbpmtHRPUH4UYeZ3Mg00qbMle4WshHWnfmqOeTvUaQBkkmQWqh4VYjlD0pZ+7zn96GFncWsGyOjv.rrTPCbVA6KpfaBq47fC1THch+MQlMuiCSUMTvsFOMpyljHqHedq5cAkLZKhbTb4dgxyowvxssdzzJgvxg+MkJ2BQIMc3UbRiJvH.2rVhE7qJ2GVoML3wOlIHXVsta2EdLeJNMiA4V0Ui3V4ad6t46s6s65tIqt2I51TMIYIWqGybBWJXLPt0vFxHuOfs3EouCjOJ58XVArNQ8nm5yy7u64YtiaI1ANNIJ3i4T0yy.9cMDFsbJk9euZ7PrBaFBtzmNuLPpnFJ3MDdu9JH6Hw.+gP90JQl9RnMlWpmTKhKXXU8w2lK4VFPqA0lYZlKp6BUKbuD71yza9gMS+9ux4CktOveBUQluc91XK7UqZ+ey2k2P9Y9iRR.hphrM8u3M260g+2PkWHJTT9rKwJI8208jWUjNU+EFDPyDNGXl9TuFlNJqcaisQYlB73Ri+R+rLXGis2xfcVEDkhIRwaI1SGl6i+3ROZNwK+1k.+KM1QcPkmXb0+T8mG7VBodo1.X2cE3Q6Jvd6Jv96JvA6Jvmrq.O9eFn4q299BkH0dbBgtbxnxwMddi3Xc2XYmK5uo4kFIK

[d.healey]

@VirtualVirgin Why not store the event IDs along with the timestamp - Message.getTimestamp(). Then you can compare the timestamps directly instead of relying on Engine.getUptime().

I had a quick look through your snippet but couldn't immediately see why you're getting the behaviour you are. However I made a minimal test just to demonstrate that currentTime and lastTime will have different values in a simple scenario:

HiseSnippet 785.3ocsU00aRDDEcFnqwE0DahI95l9DnMM6hsTvFi0BTkXokTZa7slgYuKLoKyrY1YqRL8mh+G8efNytPYoenFhkGHb+3bmybmy8ROofBwwBIBW5jIQ.B+Tq9S3pQMGQXbTmVH7KsNgMFbZAgJhyw.UH8AoSUzdShHwwfOBiK9QSxX6UPoe9462iDR3TXtKD5LAiBGvFyTy81a2OyBC2m3CliXt+M2sCUvaJBEIZhUzxEEQnWPFBGRLoUvB8IR7HD9UVta0ng6f5a2vMnlmGodi50FPb872jB08HUaD3sc0ACnTD9Qs8YJgruhnfXDdk8D9S5OR7Ud1AbFKlMHDLFdn95SNy89hPeyUz3E0bDKzu2rNVLRWkdy6eEy5euvpKymcs+48wmmFvYNh7MPbgEoWwEnmWd54lid2Akv4nzJYTZUq9TIKRMOhgOOwpCWAx.h9cJOUxxEU3GXqlBcFb0FiIW.6K0FWinbMW20czeUYmRkjvPmPRrx7HtSpEMQJ0.ybDjvoJlf6H3GJTvQ7xUJ88R14xw4cNs4CYbXigf5zHk1UYcgsKYyBbJyhaAA5f9kmcHUpTxVWAaM+hEgvFQRFWUdslYUzwjxacVy4044god2DvA55kK6qKuN0qrsMm+LWZFtvc5pRN27ZEDjcutQLSOTJBCA4cF1TN4eBXYdx3AfbcmKIgIv0IpebWTwXc+Jl7BZZ1SZtDE7Nbl5nHfeexbzTcf9Wm1oEQQLxro9z4EAREyPAbK3R8Pdlny1pEDegRDoGyukhTOKH7SBIpEGPLqQlFP2CVPUZTd7XlZR90L+2lZ9Wo3pV8XJ5n6liEtCNp6TODbb5tlmY0NH.np4DbEq8+xCyhEzwhDEiOrKQIYeSq2NLYbe89YJnOcNGBMZPbAiZIy10Xa5.8AtepwuzelFzyXimFzaVPzXBUJNmlo7MaydbpGMm3oa9ss5Zrc7PoSC46yi0KWO2rnOeotEvpKKv2rr.2bYAt0xBr1xBb6kEX8+NPy+88gDkXb1XCB0sW6zUIXbaNQq.SUqneOggjUN

Edit: Just commented out some lines in your snippet, and the lastTime/currentTime are different there too

HiseSnippet 981.3ocsV01aaaCDlJwZXRaCXEX.6qB8SNIcwRY4E6ErttD6rYzkTi51hALLTPSQaSDZRAJptYLj+y6ev1QQaKJGm0MiU8ACy6tmiO7gGOxAJIglmKUHuvWMOih79T+gyE5oWNEyDn9cQdeo+qXynQcobMN5kThTkRUQGgtXdFNOmlh7718GLA6Ez.U98me2EXNVPnUlPn2HYD5OwlwzUVG7rmy37qvoTyTTY+3m0mHEWJ4xBfX65GixvjawSn2fMgsiO5Gw4SQd66GeRmNwiZeVm3wmljfa2o8oivwIoGSnsSvG0YbxYGMZDgf79ndoLsTMTi0zbjWiKjoyGNU9aB6D7FVNaDmZFjfFByr07URdpYIZrhtbJimNXohkifrLnR+10peeg+0rT1J6U53mW5HpBgq.5sSc5saM5k3RuXG5sAJ44PoFVJ8H+gDEKSW4wvmOwuuPSUiwv9jKUrwh144M7uTBQHzGNCeK8JELXEhlmFG+jH3m8NOLLTQmDww45xBkuMJ97RKjBkBPW2nFFYKkJMExDblfFMtPPzLoHZBsDPYHM2K7OBCpmldhIP7GBg85LStZBDHHL.XZtjSOLSwD5lO1AyiiNvkIlvWK3kL2D4x+aypKYcowWsJNSXsZsdFWgyjxUCLRUfiL4jQHMJptPIph97v6BqqX4.le4WsxnPpou.NgTHzaPDUkmRWoi4Vgr09rwQMqVSOM5j8BC.GUqy7CIbJVUJpANygcyJ3tvZFO3.vVqVqs3ch3ahLBfiASdMjvMKOMJwPi8aUmHYE4SaVuZXuxoaShcTZ4JhIxYozEKenI0RAwxi+aIK2BRqLU6U7BvEX.dW3cqjao3lx0iUlMd4RBlWqV2cx6KFhmkwo4VcFPrV7lc56uG9dp0dXBCURQqS1wiWv159Lm3URNmp1naCcT+S.aJJlMhpdRz6v7B5p.gVQ06u4+v82ba+RrMfbBTJ5KX5WjQEOTSYzhtVv+dc+tXM1zTbgMHtLpRyLTvqK8cvUR1VjA9co42pkYvkR2q+Iz4VlVvw55syMW5svAnA05gZ5SB0h54tWJ9+VO9+sT7Q9CXZxzMywc1.GAk5CAGWby3m42a7XJQWQvF9W8yeXtFD8RYglIlbMVqX+NTucSwrgvqIHTX1EBJ2TC5sioZwNN1L1n.CohzxA+E7svYhYr2BmIKchlgIJ4aI1JeycueboEfShx2oD3esYbTBp7zfqNOCdJvaMOKwMU2C3QaKvudaAd71B7jsE3oaKvy1Vfse+.MuT66KzxY1iMHz0C5U1Jwyqm.CUfkUqn+F.QHJzJ

[VirtualVirgin]

@d-healey said in Very strange behaviour with Engine.getUptime()?:

@VirtualVirgin Why not store the event IDs along with the timestamp - Message.getTimeStamp(). Then you can compare the timestamps directly instead of relying on Engine.getUptime().

I was trying to make something more generic so I could use it for getting the time difference from buttons as well (for making a tap tempo for instance).

[VirtualVirgin]

@d-healey

Message.getTimestamp() always returns 0?

[d.healey]

@VirtualVirgin Message only exists within the on note on callback - and any inline functions called from on note on

[VirtualVirgin]

@d-healey It is being called from an inline function placed on the onNoteOn.

Using Message.getTimestamp() directly on the onNoteOn callback also returns "0":

[d.healey]

@VirtualVirgin Weird, I'm getting output here. But actually the values don't make sense to me, I'm not sure what they represent.

[VirtualVirgin]

@d-healey There isn't a particular standardization spec for MIDI 1.0 timestamps that I am aware of so I was thinking the HISE timestamp would be related to an internal clock. I would expect the output to be in samples, milliseconds or maybe PPQ, but I don't see that reflected in your Console.print there.

@Christoph-Hart
How do the timestamps for MIDI messages work in HISE?