Questions about EQ (slope, node tree restrictions)

Morphoice

is there a way to set the slope of a low/high cut? what is the actual slope? 6dB? 12dB? Is the curve accurate or an approximation? I can only get the curve steeper via the Q parameter, but that also makes a boosting bump
why can't I add a parametric EQ to a waveform generator's FX section? It only works on a Synthesizer Group. A scriptnode fx is possible, however.
does stacking filters.svf_eq scriptnodes have the same CPU cost as all the bands on the parametric EQ? they seem to have the identical function, then that would be a suitable workaround

Matt_SF

Nope, the actual slope is 12dB/octave. You can increase the slope by chaining multiple filters in serie.
The EQ is not polyphonic. If you put the WFgen in a Synth goup, you have to put the EQ in the Synth group's FX chain. The FX in the WFgen only accepts polyphonic effects == the FX are applied on each voice.
If you're using a WFgen not in a Synth group, you can add the EQ.
Unless you're using 50 svf_eqs, I guess the difference doesn't matter

aaronventure

@Morphoice said in Questions about EQ (slope, node tree restrictions):

does stacking filters.svf_eq scriptnodes have the same CPU cost as all the bands on the parametric EQ? they seem to have the identical function, then that would be a suitable workaround

Yes, if you're using the SVF Parametric EQ preporcessor. Otherwise the parametric EQ is biquad, which has about 20% lighter CPU hit.

aaronventure

Also you can make a variable slope filter by controlling dry/wet of individual filters in the series past the first one.

Some notes about the snippet:

probably better to handle it in a script where you mathematically specify the sections
There's no perfect crossfade using the Xfader which is why the response dances up/down

X-fader needs a true equal power curve (you can confirm that RMS isn't perfect just by crossfading two identical signals. Having a curve output that allows for instant serial addition like this could also be cool.

HiseSnippet 2005.3oc6bsraaaDEcnjobrcd1lrW.YiSSffjkhdftvJ9gZMZsspkqa5J2wjijHLIGVxQ1VonqZWjcEn6xt9Izscm+.RAJP9B5ef6WP5LjTT7k0qXKGmLx.Fbdb4b48dtONBPrtIVBYYgMAByuaWCDP3lhM5pSZuZanhNXi0.B2VbSnEAYl1YpU5Z.srPx.AgjeAaBg4lAX+4rkWApB0kP8mB.1CqHg9ZEMER+YqW8qTTUqAkQ6pn4a2EptgDVeUrJtCUeRJlEX.kND1BsEjssDhfuDZ0FH7YhYeZQDpT9lxxv74kKVrXkxUJTIGrY1BvCxUrRgkJTZoxkZBDRstrBAa1f.IHK5McErb2FswGq6b.6oXobfJhMHGnA8jclFrZaEU458LNV.fvL06apR5Xptu3lJxJdy22jcW6ER2WB+FMgDCRkxMFpjfOUZFGU5dhMjLULH8WgoOKHtgN0C1DR8M9UEm8BDNUbULcC5jLZvCQ0LoC7DXwhYy9jzz+8nOuYGcIhBVOMVeKLAss9hOZ9eZ94l+mmOc3kZ1L10XGiIVUEYF6xL3f4fDbQ8NZGfLeR5ifpcPdaj93GzllZzroRNO091HVeCcEx1FH2w0vpxLaE65nd.fqYid02twZPBj4TbmitOCjIQgoNBqgNhFF33hlSbMj0gDrAMPHh+ihbvxcTgjfvIVfl6BT6Q.eHyQoaoP55OPbLvXYGHFaTUw6IVWgH0NdcLQL5H0RcYnitQl2Rb8lMQRj9J3Lh0d9jFFlcrCCm2MLjdlrC+ltCSW644hM.7kh8A6FlHCnIZWbcUX2EsfZFpncn53SRefJV5vFJu.EMBwvQEVgsiEkZC00QpVSRfTpKnjSfcvcHJ5s1DRLUNAHHtUGsFz75RnUc0N5bBIXwPNiyxFyvEMP5x1CdK8i6h4XiEbWLWuE8E9sEhbL17Pa2g60.gYcr8V1F48adRNvyTUwGuJVyPwEkR8A1yUGq10nMVWQhMkyN5ooOSC2gdHtpKsDztPEUFrtQGKZhB4s0aP2rccOgj6gLsruwyJlMC8OJLeKrLcoYqAknF1t0gj1r3EVtGJ3DYlQxCjFPW8iSbJ.FXY1csWoRGCGUq0XVCVgRlKqNzjJFMMNKCTBgj9xH4.ZYingjIbN5Mr1ispDTErGCTvDx9TXaWz4gHEaqyKpgkcT5.OS222yj2Nh9T3S7A+L7towy3SiuknDjhZ2mv9ePk9A1JMMhHi+sDUqCdKFhhOieEGzOo3tPyVHhkaUBcjjCFzFgrgLyzHa1c+iQjb.OmGa9aHtlY2uCQrOFZ5fS5UqYWl1X4cEyxnKiNoW.x5zXbYZ8K6sK7Tw7EyD7yQamIS8lUXWVYU2wYZ9XeiA.evHlSn2HpVsohtqcueV8MgmDZtyVtAAYvRa4eeMNDcriiv+NYNqYEctAQt0qgZB6nRhL+y5PvZzzUxd9eW29s86yxecvuW5xvuutXgxmue2ablLO9iAbvRWGvAUll3.G+d+w0+n.Gj65.Nn7TrNf91AwAWy869UWWPvrNcu3Ui8b6HyxPUgDSOT9c2yk5U+dq2xN2nfj45cFg5moWuIIBzMEauFLcXHcHFWqTdxNrVoBcz1lh6zSM2WS4DpmLv4eWu.hSZBkQlwb7gkOdyie8nGcZJ8qfM2kHTyc2gwWvmw20Y6zBNHzlSwXj5ATDE2YyFTf0wLBo8izR3ehfQdCKPrEy5GHPLkMi7fGxneOYVry+ddsJNKYzPqVQHFr.EJYhxzZPf3Vg4CLD.TBOHiSx0.VsjC1p88mF0pQsPAsZ+wqd0esbDq1os9yGT0AyY6slPilO8a1QxqV82pNR52O7oq8lkc9J+Zngwj1TJ3gUxBUiWIKTcH50Egc6lh6frPjwDxAlDlfwTAvMiKK.bRy35I6Xlw08nw5n8MvgYg9IhMUTYZZFu0i6r8VbbHNGtWgZNmT+KGb2BteUKAhuRMXbRgpQvIu7MUGorRU+Es+1E+VyD8icP5RcAQBEhElPmeP32p4y+lkCoWm95W+vQSu.O7+9Ga8Jo32.hjvMdX6YKOj3okhIdZopiW16wMOz3m5YTKnLzTOU+2W7qwYprm+BRuNa4AYp72lvTyTMHUZNw00YMrKOFXp.Eb8ZkXrJ35IEufKuf6kTA2ocardbiC7HEMtJ.KwQqi8v7bBz1du1mCXwhkvYo2OHbtfGowhScFm2MnkrHmxYDv1BdzgJdww4bAuT9E+.fzYB+gW9sWbVmi1WIFuH3zm04BdLGKN0octfGywhbdmbdmbdmbdmuS7N82NAm3IulKm34PIdFBWLZssGgtzDx7rx6aLOKeUy7rLm44.XdV9xf4Y4ObYdVly7jWE7Z.yyxWgLOKyYdxYdxYdxYddAw7rLm4IulKm44zg4Y4Ik4Y422XdV5pl4YINyyAv7rzkAyyRe3x7rDm4IuJ30.lmktBYdVhy7jy7jy7jy77Bh4YINySdMWNyyoCyyRiByyy8weVm29CN0danREAbwaPF4zIgM.yDqAH3uv33e5iUvPutClTIWZhkL24H4JX7gZP62NKS16Ioow6RFMnjIdeWx3LuyMrmgltU28qaXS13z4.GEF9noHqrujTvaUDAWZREL+jJXgIUvmNoBVbRErzjJX4gKH6cwlahHF3kFtWecm9sE50Q.6mHx+CB7hGL.

variable slope.gif

Matt_SF

@aaronventure said in Questions about EQ (slope, node tree restrictions):

X-fader needs a true equal power curve

Alternatively you can use linear Xfading and math.mul nodes :

aaronventure

@Matt_SF that's so stupidly simple, thanks

Mighty23

@Morphoice said in Questions about EQ (slope, node tree restrictions):

is there a way to set the slope of a low/high cut? what is the actual slope? 6dB? 12dB? Is the curve accurate or an approximation? I can only get the curve steeper via the Q parameter, but that also makes a boosting bump

one pole =-> 6db/ octave
ladder filter & moog -> 4 pole -> 24db/octave
Linkwitz-Riley filters -> 4 pole -> 24db/octave
other filter is 2 pole -> 12dB/octave
https://forum.hise.audio/topic/8784/qol-write-slope-information-for-each-filter-node-in-scriptnode-and-filterfx

https://ccrma.stanford.edu/~jos/juce_modules/classdsp_1_1LinkwitzRileyFilter.html

Questions about EQ (slope, node tree restrictions)

17

2.0k

10.9k

94.6k