bkFX v0.8.0 - Free Effects Suite - NEW down-/upward compressor

[toothnclaw]

i don't know, in logic i never had the problem and there's as well some guiless 3rd party plugins in studio one where this works. on those i can finetune with shift held or using the mousewheel, or i can type in the desired value. i assume it's rather the parameter stepsize you allow internally, as well as defining a default value, when ctrl-clicking the knob, as this action just sends a "set default" value, which is what you have to adress in the code for the respective parameter. at least that's how it _should_ work.
however, it's really crucial to be able to get the exact values in whatever way (even with the host internal gui), as, like i stated above, if i can't set the perfect delay factors, i can't really use the delay, at least not in host sync mode.

In FLStudio, all knobs are fine tunable by pressing Ctrl and dragging. E.g. on bkDelay, I can adjust delay factor from 1.500 to 1.501, 1.502, 1.503 etc.

So maybe it's your host that doesn't allow this, not the plugins.

[karrikuh]

no, you missundertood me. <snip>
so what i meant was, that there should be 2 additional modes, which essentially summ up the signal which is exclusively going to the delay network, to mono. that way both sides of the inputsignal, left and right are respected, as the delay network recieves the stereo information summed up to mono. the signal into the delay network, right after the summing, then must be hard panned internally to either side (selectable with the pingpong knob on the gui), effectively overcoming the above mentioned problem. the signal _before_ the plugin then can be panned as one desired, without the delay being negatively affected in the way that it would miss the information from one side... that is how all professional pingpong delays are done (i.e. eventide, boss, roland, korg, etc). i hope i made clear what i mean now.

Yes, that's exactly how I understood your original post. Summing the input to mono and then panning either L or R before feeding into the delay lines, but keeping the dry signal stereo.

and another delay mode would be really cool, actually easy to do:
a cross delay. meaning one side input, other side delay... tha way one can do great jarre delays, with the synth itself being panned to one, and the delay outputs the delay only to the other side. basically just like the stereo mode in pingpong, only without the taps altering the side one after another...

Since you can achieve non-standard effects like this using your DAWs signal routing capabilities, I'd rather avoid introducing yet another parameter just for this special purpose. Or maybe I could rename the "ping pong" parameter to "routing" and combine it with the "Jarre mode"?

there's no way to do such a delay in any host, it's the delay that has to do this. essentially it works like this: <snip>

Of course you can do all this by using bkDelay as send effect in your host, and inserting bkChannelTool before or afterwards in the send channel. Using bkChannelTool's "Flip L-R", "Width" and "Pan" parameters gives you essentially all the flexibility you need for this and other effects, such as narrowing the stereo width of ping pong delays. Whether such functionality should be integrated into bkDelay for convenience may be discussed however.

and another thing on the delay:
in studio one one can't "reset" the plugin to default values, but that may be the fault of studio one. there's other plugins that behave the same.
but more crucial is, that i cannot seem to get the offset factors to perfect 1, 2, 3, etc... as soon as i moved the knob to whatever value and release it, i can't go to perfect factor values, as tzhe knob just won't let me get to the value, as well as that the plugin doesn't accept numbers that i directly type into the value field. this actually is the case with all bk plugins in s1.

Yes, with your hosts default GUI, targeting accurate values is difficult, a GUI version of course will allow to snap to integer factors, maybe keyboard value input.

i don't know, in logic i never had the problem and there's as well some guiless 3rd party plugins in studio one where this works. on those i can finetune with shift held or using the mousewheel, or i can type in the desired value. i assume it's rather the parameter stepsize you allow internally, as well as defining a default value, when ctrl-clicking the knob, as this action just sends a "set default" value, which is what you have to adress in the code for the respective parameter. at least that's how it _should_ work.
however, it's really crucial to be able to get the exact values in whatever way (even with the host internal gui), as, like i stated above, if i can't set the perfect delay factors, i can't really use the delay, at least not in host sync mode.

The delay factor parameter is floating point internally, so there are virtually no steps. Using fine-tuning in either FL Studio and Reaper allows me to adjust the factor accurate to the last fractional digit with ease. That being said, keep in ming that the current GUI-less versions are only temporary, and the functionality and convenience of default GUIs varies greatly between hosts. Even if a plugin provides extended functionality like reporting default values, many hosts will just ignore it.

[brok landers]

karrikuh, just a short one, im on mobile and it's hard to hammer the text with the phone - i understand all points. completely agree, exept that i think the way i described the correct ping pong mode is mandatory... but of course, these are your plugins, you just do what you have in mind... don't feel offended or forced, i was just interrested... i am eagerly waiting for the ensemble to be fixed... it' extremely close to the rea thing...

[karrikuh]

0.4 Beta (2013-06-16)

• bkChannelTool: reworked "Width" parameter (200% -> pure side channel)
• bkDelay: new "Input" parameter: allows to sum delay input to mono or invert input stereo channels.
• bkDelay: replace "Stereo Spread" parameter by "Width" parameter (works like in bkChannelTool)
• support parameter value text input (unfortunately only supported by few hosts, e.g. H. Seib's VSTHost)
• support Cockos REAPER extension for text input of automation break point value
• bkChannelTool, bkDelay, bkEnsemble: fix crash in x86 builds
• bkDelay, bkModFilter: fix another potential crasher
• bkEnsemble: slightly randomize LFO rate across instances
• ditch bkPassthrough again since bkChannelTool uses almost no CPU as well and can thus be used for the same purpose

[DuX]

Whoa! Thank you, karrikuh!

Just a little something that I'd like to commend you about that I noticed: your plugins when fed 32-bit audio output 32-bit audio, when fed 64-bit audio they output 64-bit audio. Intentionally? Anyway, great little detail as there is no bit depth conversion at all with your plugins. I don't think I've seen any plugins doing that. Not that I test for that, too, why would I? The difference is kinda negligible, but little differences tend to pile up in a busy project.

Cheers!

[miedex]

0.4 Beta (2013-06-16)

• bkChannelTool: reworked "Width" parameter (200% -> pure side channel)

instead of summing L+R at <- 100% i think it would be good to have it remove the side bands. there is a real need for propper mid/side balance in these types of plugins

[karrikuh]

0.4 Beta (2013-06-16)

• bkChannelTool: reworked "Width" parameter (200% -> pure side channel)

instead of summing L+R at <- 100% i think it would be good to have it remove the side bands. there is a real need for propper mid/side balance in these types of plugins

But that's just what "Width" does: blending bitween M/S channels:
Width = 0%: mid channel (mono sum) only
Wdith = 200%: side channels (L-R and R-L) only
Maybe it should be labeled differently?

[DuX]

I think no. This seems quite logical to me.
Width 0%=mono, Width 100%=original/stereo, and 200%=sides.

[miedex]

0.4 Beta (2013-06-16)

• bkChannelTool: reworked "Width" parameter (200% -> pure side channel)

instead of summing L+R at <- 100% i think it would be good to have it remove the side bands. there is a real need for propper mid/side balance in these types of plugins

But that's just what "Width" does: blending bitween M/S channels:
Width = 0%: mid channel (mono sum) only
Wdith = 200%: side channels (L-R and R-L) only
Maybe it should be labeled differently?

If I have the output of 1 mono sound going into the left and right of bkChannelTool then there is silence at 200% because its a pure mono sound and aparently bkChannelTool is rejecting the mid band

If I have the output of two totally different sounds going into left and right of bkChannelTool then there is no silence at 0% even though its a pure stereo sound because its apparently adding the left and right channels together

so it looks like you have the <--100% doing "width" and -->100% doing "mid/side". i doubt anyone would complain if you changed the width control to propper mid/side rejection and added a new width summing control. both are useful features.

[karrikuh]

If I have the output of two totally different sounds going into left and right of bkChannelTool then there is no silence at 0% even though its a pure stereo sound because its apparently adding the left and right channels together

so it looks like you have the <--100% doing "width" and -->100% doing "mid/side". i doubt anyone would complain if you changed the width control to propper mid/side rejection and added a new width summing control. both are useful features.

The problem here is that "totally different sounds" is only how you perceive stereo information of your signal and is not necessarily related to how mid/side signals are usually defined mathematically (i.e. arithmetic sum/difference of channel signals). In principle, you could come up with infinitely many ways of defining M/S channels, the one that is standard probably being the most trivial definition, but may not reflect human perceiption well in all cases. I would be surprised to find an M/S processor which gives you the results you're expecting.

[DeepUniverse]

Big thanks and respect!
We think there is no need for the GUI. Sound, quality, stability and features are more important things.

[DuX]

Big thanks and respect!
We think there is no need for the GUI. Sound, quality, stability and features are more important things.

Agreed.

[Rah]

Useful plugs, thanks a lot

[SaganTech]

Hello karrikuh,

You wrote:

support parameter value text input (unfortunately only supported by few hosts, e.g. H. Seib's VSTHost)

So I was curious and surprised that there was even a effString2Parameter VST opcode. So I implemented it in Metro 7.1.2 and your plug-ins work beautifully with it. Thank you for that... except one thing....

For non-numeric parameters such as the 4th parameter of the bkPitch plug-in, if I type in "random" or "fixed" I see this in the debugger log:

First-chance exception at 0x000007fefceabccd (KernelBase.dll) in Metro.exe: Microsoft C++ exception: std::out_of_range at memory location 0x0012e988..

And also the parameter does not change to "fixed" or "random"

Jeremy

[karrikuh]

Hello karrikuh,

You wrote:

support parameter value text input (unfortunately only supported by few hosts, e.g. H. Seib's VSTHost)

So I was curious and surprised that there was even a effString2Parameter VST opcode. So I implemented it in Metro 7.1.2 and your plug-ins work beautifully with it. Thank you for that... except one thing....

For non-numeric parameters such as the 4th parameter of the bkPitch plug-in, if I type in "random" or "fixed" I see this in the debugger log:

First-chance exception at 0x000007fefceabccd (KernelBase.dll) in Metro.exe: Microsoft C++ exception: std::out_of_range at memory location 0x0012e988..

And also the parameter does not change to "fixed" or "random"

Jeremy

Cool to see this useful feature implemented, I wished all the major hosts had it since I feel if that was the case the need for custom GUIs for bkFX would almost totally vanish. I once issued a feature request for REAPER, was rejected unfortunately. Maybe we can change the dev's minds if more users raise their voice?
http://forum.cockos.com/project.php?issueid=2657

As you have observed, the effString2Parameter is only implemented properly for numerical parameters as I don't see the point for the others except for consistency. It shouldn't however crash or something, just do nothing.