XHip--Please finish your synth!!

[aciddose]

"any chance of making tune controls for OSC 1 on the OSC page?"

no, since there is no point in doing so. you can accomplish everything you'd want to accomplish by using slightly more effort. generally, you do not even need to do those things though. yes i know it may seem more "neat and tidy" to detune the first oscillator downward, but it has no practical purpose. you can accomplish the exact same thing by detuning the entire synth, but you shouldnt be doing that anyway. it is totally pointless.

why wont i do this? it'll only use more cpu power. if i was willing to waste cpu power, xhip would be more like all those other synthesizers out there. you named a few :P

it isnt practical to do unisons of other waveforms without adding a great deal more controls to the interface which would be very rarely used. the cost increase of a ramp unison decreases per ramp as you add more. its O(log(n)). with a pulsewave, or the triangle/sine (which is simply a filtered pulsewave) that no longer applies. the cost is actually not even linear, but more like O(exp(n)). if you want an idea of what the pulsewave or others would sound like, just apply the main unison without any other processing going on. it isnt all that exciting. if you REALLY need unison on those waveforms, use the main unison and sample the output. use that sample as a pcm waveform.

i WILL implement this for you just to test though, so we can see how many more controls it requires and how much cpu power it takes up, then weigh that against how useful it really is. i'm not sure when. it doesnt take very much effort or time on my part, i just have other things i'd rather be working on ( http://xhip.cjb.net/xhip/todo.cgi )

the reason the pulsewidth control is multipurpose is that there is no logical use for it on any waveform other than the pulse waveform. it isnt possible to synthesize a stable ramp in any dutycycle other than the existing 0% and 50% (ramp/triangle) levels and maintain all the functionality like sync and bandlimiting.

subject; i removed your mp3? i was just clearing out the whole dir before i made the categories and started archiving stuff. you mean these? http://xhip.cjb.net/temp/public/xhipy.mp3 http://xhip.cjb.net/temp/public/xhip_food.mp3

i didnt realize before that i have several gb of space on that drive. no need for me to clean up files unless they're extreme (>32mb) or inapproperiate.

[brok landers]

aciddose, am i getting dusted somehow?
i thought there was an option to retrigger the osc's (individually) at a userdefined startpoint?
i cannot find it ...
very necessary for basses or percussive instruments ...
please let me know where it is ...

[JerGoertz]

dj_subject, that food song is the bomb dot com.

[aciddose]

"i thought there was an option to retrigger the osc's (individually) at a userdefined startpoint?" you mean reset the phase of the oscillators at a specific point during some event. no, there is no such option implemented yet. i'm planning to include that type of thing with the event routing system. is it really so important for some instruments? analog subtractives never had this type of option, unless through the osc sync input. in that case it would only have let you reset to the start position and not to a defined point. i've never had any problem without the ability to do this.

if you're wanting to reproduce exactly the same sound on every hit, the solution would be to sample the output and use it with the pcm feature, or another sampler/replayer.

i kludged it for you though:
http://xhip.cjb.net/xhip/releases/v0/b6 ... 6.11.7.dll

i added a 4th option to the osc sync parameter called "trigger", which will reset the oscs whenever a trigger (new note trigger) is received. that will reset the phases to zero, reset the position in a pcm to zero, and reset the tables and etc. so the effect is that of having fresh, completely reset oscs on note triggers.

[brok landers]

you mean reset the phase of the oscillators at a specific point during some event. no, there is no such option implemented yet. i'm planning to include that type of thing with the event routing system. is it really so important for some instruments?

ohh yes, very important.
i hated all my analog synths for not beeing able to do that.
simple explanation:
you have a very percussive bass set up.
amp attack is 0, decay is "very fast" but still audible, sustain is around 6db lower. that way you can increase the punch of the attack.
the filter env is having also a fast go on the cutoff.
now, with free running osc's, you will allways have different positions in the phase of the osc, which makes the osc sometimes click, and sometimes not, which is very audible in these cases, and very unwanted.
now, with the adjustment of the individual phases of the startpoint of the osc's you can actually set the clicking ammount and frequency (which works greatly in conjunction with the filter of course) to your needs.
and not only that. there's a significant advantage to this in addition, which i very often use:
assuming you use 2 identical osc's for this bass, pitched to the exact same tuning (or one osc is exactly one octave down), you have the wonderful ability to get a bass "stable" in the mix, frequency wise, as you can freely adjust the frequency response on that bass of the phase of the osc's to fit it in the mix, by cancelling/boosting certain frequencies.
you can _never_ do this with an eq.
when it comes to "define" a sound in the mix which has to be static, this is the key. same goes for all signals, that need to have the exact same phaseposition, i.e. pecussive stuff like electronic bassdrums and such. with these 2 features one can add unbelievable punch to such signals.
if you're creating club oriented, or generally electronic dance music, this is a must.
of course it is unwanted for other signals, i.e. if you use a supersaw with a harder attack ... thats when the osc's have to be in free run.
but to have it when you need it is a big advantage for a lot of sounds.
analog synthesis is superiour when it comes to "warmth" but if they would've got that features on top (these came with the fist samplers/rom based synths), the would've had something i (and a lot of other users) wanted these days ... i would've payed a lot these days, if my minimoog could've done this, or my mks80, or my memory moog.
these 2 features are imo the most overlooked in some synths.
i often recieve the question why my basses that i created in soundsets are so punchy and how i do that ...
this is it ...
however, of course xhipsynth is your synth and you certainly do what you want, but i strongly suggest you to implement these 2 ...

analog subtractives never had this type of option, unless through the osc sync input. in that case it would only have let you reset to the start position and not to a defined point. i've never had any problem without the ability to do this.

of course, i know ... but ask around some analog fetishists, how hard they sometimes hated these "spikes" in the sound, when they created a punchy sound ... they compressed the shit out of it, just to tame these peaks ...

if you're wanting to reproduce exactly the same sound on every hit, the solution would be to sample the output and use it with the pcm feature, or another sampler/replayer.

i did this all the time, with some stuff
but still, the phase adjustment is lacking with this method, unless of course i load these waveforms (sampled unfiltered) into a sampler, then adjust the startpoint manually, but these samplers lack of _your_ fast envelopes and _your_ incredible filter, which i really love!

i kludged it for you though:
http://xhip.cjb.net/xhip/releases/v0/b6 ... 6.11.7.dll

i added a 4th option to the osc sync parameter called "trigger", which will reset the oscs whenever a trigger (new note trigger) is received. that will reset the phases to zero, reset the position in a pcm to zero, and reset the tables and etc. so the effect is that of having fresh, completely reset oscs on note triggers.

great, thanks! step 1 onf 2 is done
again, it's your synth, and i don't want to nerve you, but these 2 are really a big underrated thing, if you ask me ...
anyway, thanks for considdering it!
i watched the development from a very early stage, and beside a few things i allways loved the way you implement/code your stuff.
it's just very accurate, very low (close to noe) aliasing, everything sounds/works just really good.

but while we're at it, i'd like to ask for one more ting, maybe it's easy:
the filter sounds good for some sounds, when the osc hit's it very hard.
mostly i'd like to use just one osc for such sounds.
but then the filter inputgain, when full, is still too low to me ...
the boost is max 12 db ... can you increase it up to 24db?
or is the filter getting unstable then?
i did this with my minimoog (opened it and readjusted the filter input gain, there is a screw for that), man, that nailled it, i can tell you ...
though it started to groundnoise a lot (and the max resonance was way lower then), but i did not care, as i was used to gate all my synths, to get rid of all the juno-chorus noise and such, and the sound was so delicious, even with less resonance ...
well, i thought i'd ask ...
thanks, and lemme know!

[aciddose]

well, you'll get aliasing with anything louder than 0db (even just using both oscs) through the filter since it isnt oversampled and it uses some very basic saturation. the effects you'll get from increasing the gain into the filter wont be the same as in an analog synth since there is zero distortion applied to the signal. the distortion is only applied to the feedback. what you'd want would be a waveshaper before AND after the filter where you can adjust the drive into the waveshapers and into the filters individually. for those configurations you'd really need to use the modular version which i'll hopefully produce some day.

as for increasing the gain range to +24db, i'll do that quickly for you, but i'm not willing to leave it like this in the main version since i do not think it is useful for most people, some problems (wrapping distortion) can occur in certain conditions (even with just +12db) and it'll require that i change the configuration of things including the patch format.

i've just offered to create a plugin for somebody which i should get to work on now, but i'll give you just a very quickly modified version of xhip so you can see what the extra gain will do without me making any other changes.

http://xhip.cjb.net/xhip/releases/v0/b6 ... 6.11.7.dll
(i used the same filename rather than make another version)

a fixed gain of +12db is applied to the mixer output here so that i dont need to make any other changes. so, to get normal behaviour you have to set the mixer values to -12db.

i'd recommend using a little of the waveshaper (log 20%) in order to get a bit more of the saturation you'd expect by driving a filter this way. this sounds pretty terrible to my ears on most configurations, but it actually sounds pretty simmilar on analog synths anyway so..

if you get terrible wrapping distortion and your ears start bleeding (or if having it so loud without any wrapping makes them bleed) dont blame me

the event routing system to allow you to set the oscillator phases to whatever position you like wont be implemented until i get a lot more work on the gui, see the todo list.

[brok landers]

hey aciddose, i checked it out, but unfortunateley, allthough it does somehow what i meant, it is not working correctly for some reason ...
i made an exapmple patch (nothing fancy, but you can hear it better this way), please download it ...

load it, and play it repetitive ... you'll hear, that the "click" is not allways the same, it's not "stable". dunno how to describe it any better ... you'll hear it, i bet ...
any idea?

[aciddose]

the filter is free-running too. you'll have to set the filter-trigger on in order to reset it as well. i was thinking i could add another mode there, the current would be "click" and i could also add "zero". i'll quickly do that.

[brok landers]

great! you're really cool ...

[afreshcupofjoe]

aciddose, has anyone ever told you how much you kick ass!

Who offers this kind of support for a synth?

You are a fine man and a patriot.

[Lagrange]

Yes aciddose rules!!

[aciddose]

ok, same file
http://xhip.cjb.net/xhip/releases/v0/b6 ... 6.11.7.dll

there are a TON of things in xhip which are "free running" and can influence the attack phase of a sound.

- use oscillator sync on "trigger" mode
- use filter on "trigger soft" mode
- use envelope trigger on "trigger reset" mode
- use envelope constant attack time mode
- use lfo reset mode
- ensure there are no free-running waves active like noise/random/s&h
- disable dc filter on the main page
- make sure there isnt any external modulation happening like velocity, bender, or cc's.

that should ensure that every single note triggers in exactly the same way, although i might have missed something.

if your host does any processing like a dc filter, eq, or anything else at all, it'll mean the output from xhip will be perfect, but as it passes through those effects in your host it'll influence the signal.

[brok landers]

i'll check it out, got to go now ...
will let you know asap, i promise! thanks a lot !!!

[DuX]

This stuff is excellent for strong, short sounds, be it basses or fast sequences, so you don't have to worry about those clicks all over the dynamics spectrum. Marvelous, Aciddose. You should keep those parameters in XHip in the future, they really help giving you total control over how OSCs behave.

Cheers!

[jacobh]

"any chance of making tune controls for OSC 1 on the OSC page?"

no, since there is no point in doing so. you can accomplish everything you'd want to accomplish by using slightly more effort. generally, you do not even need to do those things though. yes i know it may seem more "neat and tidy" to detune the first oscillator downward, but it has no practical purpose. you can accomplish the exact same thing by detuning the entire synth, but you shouldnt be doing that anyway. it is totally pointless.

Pointless? Are you saying that changing the center point tuning is pointless? This makes it so the average tune of the instrument is more in tune with the other instruments playing.

why wont i do this? it'll only use more cpu power. if i was willing to waste cpu power, xhip would be more like all those other synthesizers out there. you named a few

The synth I mentioned was Reason's Subtractor, which is probably the most CPU efficent synth I know of. Also, it's phase modulation abilities, combineds with FM make it quite flexible! If you want, you can check out this track I made using entirely Subtractors as the sound generators (even the drums). The only effects used in this track are Reason's.
http://www.mindtrick.com/music/Phat_Math.mp3

I must say, there's a lot of things I like about adxhip, and I think it's got a great sound. Though I'll be using adxhip, I can't see it replacing my Subtractor yet. Not many synths can do phase modulation on all of their osc types!

I've got to say I like Brok's suggestions, as long as they don't add to much to the CPU. One more thing... Can the PCM waveforms have some kind of phase drift like the synth waveforms do?