PhonicsAudio is online!

[Nubeat7]

it's using 1000hz in the example.

this told me - why not using a loop? so i tried to write it in ruby completely - and here is what i got - was easier than i thought i also included my ruby stepseqencer if you are interested, like this you only need to redraw one window instead one for each step...

to avoid audiocrackels at preset change i normally use a trigger at presetchange to set all audio off for some ms,
do the presetchange and set audio to on again viewtopic.php?f=2&t=1537

[nix]

Thanks for the tip on preset change,
I may implement something.
That's a really neat step LFO Nubeat,
so much thanks for sharing.
The old one crashing was a real problem,
that's why I stopped using the loop.

Bill-
I tried running it on my old machine-
It uses about 5 % at 256 samples buffer.
It was a bit odd though,
I had to recompile it coz it was using about 13%.
It settled down after that.
My buffer was 64 samples,
when I raised CPU use was less also.
My systems are-
old desktop: Intel quad 3.00 8 gig (Dell XPS 720)
new lappy: i7 quad(8 processors) 2.4 8 gig (HP Envy DV7)
If you'd like the .fsm, just hit me up, if you want to use this synth-
u can optimize it on ur systems.
Here is the file I compiled on the Dell->
{I have deleted the file-just pm 4 the latest}
U can try this one if you are keen on the synth.

Cheers, hope the latest update was good for you too

[billv]

new lappy: i7 quad(8 processors) 2.4 8 gig

Sounds great man....
I lost track about 4/5 years back, when "state of the art" was Core2 duo or Quad Core...
.... looking forward to the day when my current cpu blows up...

Thanks for the 'alt" version...
I know i can probably get a better output if i go and tweak some buffers, but i always prefer to
avoid that.
Anyway, CPU usage is not an issue on my main cpu , where Entrancer runs great in FL,
and because I work in Reason these days, when i use Entrancer( or any plug in) its always in its own
in FL with just a midi file from Reason. Get the sounds i want and go.
No conflicts or CPU issues. All good.

Its fun to test on the old laptop.
The worst is loading the X11 fsm into FS. Takes more than 5 minutes...no kidding...
I have reffered to this "crappy old laptop" lots over the years... so better clarify it....
" Intel(R) Pentium(R) M Processor
1400MHz
1.40 GHz 512MB of RAM

And one more clarification..
On my main CPU, 2.33 Quadcore, FS VSTi Module runs at 14% CPU as a plug in.
So its
14% just to run the VSTi Module....
24-30% to run X11....
37-40 to run Entrancer
EDIT:
Bear in mind, cpu comparrision between X11 and Entrancer is a bit pointless.
Entrancer preset might have all FX on... blah blah..
If I turned on the Reverb and flanger in X11, and set it to poly..i think it it could easily reach 40.

Key point i want to make is 14% for just a VSTi module..!!!

Got me thinking weather a Ruby "Midi to Poly" is feasible/possible/makes sense?

[nix]

Heya,

The Ruby MIDI to poly is an interesting idea man.

I have gotten a major update of my VST work ready.
So I took all the links down that are scattered over the forums I partake in,
and am making everything available in one download->
http://www.sendspace.com/pro/dl/trutma
It's a 66.04mb .rar d/l,
that includes descriptions, hints on use and artwork also.
It's the pack I sent to Core which he will update Rekkerd with this week.
There are 4 instruments, and 21 effects.
Hope you find them useful and inspiring.

Here are the terms of use-

They are all fully working plugs and are free to use for non-commercial purposes.
If you intend to sell the audio run through or originating from my plugs,
there is an obligation to buy them. If you are producing commercially,
drop me a line at my hotmail-> nix808@....
I won't ask too much

Here are the descriptions of the goodies->



EFFECTS->

d1g1t:-
Bit-cruher and samplerate reducer.

ECHOecho:-
Simple delay. It uses a divisor of the BPM to set the tap length,
with the ability to use your own fraction. It's easy to dial up your favourite dubs.

freQ:-
Q8 parametric EQ with phase and fft display. All bands are switchable off and on, and it is hybrid-ie.
the sliders and graph update each other. One feature that I think is handy in an eq is max gain/reduction scale knob.
It's marked dB. You can be delicate or use a higher max.
The parameters are switchable, so you can switch between ie. notch and shelf, ABing to taste.
It's quite vanilla in that it doesn't really colour, however at extreme settings the biquads will start to exhibit artifacts,
which are to be expected I understand. It sounds great when you automate sweeps, which was hard to implement without crashes.
Graphics by Grymmjack- nice one bruv

Mono2Stereo:-
Utility to route channels.
If you are recording a mono source, just insert this in the signal chain to process in stereo.

Limiter:-
I think the ratio is infinite. Tames peaks on masters perfectly well. Easy to use meters.

Octaver:-
Stream and counter-based octaver, shifts pitch down 12 semi-tones for playing bass on electric among.
It chops grain start points on 0-crossings, in aid of reducing vibrato.

Shifter_RedLight:-
3 line variable pitch shifter stomp.
Can run the lines in parralel or series. Supports PDC,
as the grains take 1500 samples to process.
It works without pops in small projects,
but when the CPU is busy, it sometimes pops, due to the trigger architecture.
Can reccomend rendering audio to samples in large projects.
It works by resampling grains.

Standard Issue:-
Collection of effects that replace my old one- Wave Mechanic.
There are currently 11 effects.
Chorus,compressor,distorsion,drive,eq,flanger,hardclip,3 phasers,reverb
Hopefully fun to experiment with fx chains.

Sync:-
Recorder VSTfx that allows you to cut and export loops, similar to IL Edison sorta.
I originally built this to use this idea I had-
to play a metronome click on beat divisions of a sound editor's selection,
so you can tell when you selection is looping in sync.
Grid divisions of the selection are also displayed graphically.
thus enabling easier finding of correct loop points.
Additionally it records and you can play the recording back internally or via virtual midi cable,
DAW routing or DAW transport to preview your loop in the mix.
You can flick left and right with selection, also double and halve it.
Zooming of time and amplitude is in there.
I will soon have it normalize, and fade the ends of the cut loop.
You need a 3rd party app to timestretch, but I am looking at this.
Max recording time is 10 min with 3 gig of RAM, but shorter takes are previewed in the mix nicely.
There is a drawback with this device, after recording, the plugin will render the sound to conserve RAM,
so it requires the user to wait for the same period of time it took to record. Likewise when saving.
It will not recall what it has recorded if you close the project, u need to save the .wavs you want.
It includes a BPM calculator, so if ur wondering what bpm a track is at, select a bar long loop,
and it will be displayed in the footer.
Kenny Buck made the GUI, making it more ergonomical also. So much thanks Kenny 8D

Synthetic Reverb:-
Built by Maik Menz and released to the Outsim community.
I've made this one available as VST.
It's an unnatural sounding algo.
Cheers Myco

Tripper:-
Auto-filter.
It has 18 filter types, modulatable by env follower or tripped env.
It's good for wahs, using audio as filter mod source and distorsion.
The filter curve is animated, so you can see also what you are doing a touch easier.
Also you can turn the modulation off and have static cut.
MIDI learn is supported. The depth knobs are logarithmic, so the low cutoff for wah is easier to hit now.
Go hacks on res and cut via controller!
Right click on knob enables learn, select MIDI source with the little female plug.
This version has 2 new filter types, formant(vowel) and comb.



INSTRUMENTS->

Dream Sequence:-
Polyphonic,up to 32 pattern and 128 step MIDI sequencer.
Copying and pasting of patterns is supported. There are 2 modes- 'edit' and 'run'.
In edit mode it plays only the current pattern. In run it loops the patterns sequentially
from start and end points which are set by the user.
Graphics by Galaxpel.
Timing seems not to be sample accurate, however it is well within a usable precision.

Drum Roll:-
Drum synth. Hopefully a full drum kit can be coaxed out of the 2 oscillators.
Will come back with an update including the full presets, but for now it only has kick, clap and
cymbal. The vector control's colours can be customized to taste.
It has the requisite multi-point spline envelopes for pitch,amp and filter needed in drum stnthesis.
There are also 4 effects to colour ur drum.

Entrancer:-
Instrument originally made for the KVR Developer Challenge 2009.
It has 375 wavetables, particularly focused on the 303, however it is polyphonic and has 3 osc.
I used wavetables from Muon Tau, IL TS-404, Venom VB303,Roland TB303, Cheetah MS-6 analog among others.
It's a labour of love and damnably frustrating to develop too. There is one singular feature-
the ability to sample the output audiio stream back int a wavetable. You can generate original
soundforms quite easily, just select start and end points, and press the little arrow on whichever osc
slot u want to use when in user(U) mode.
Other nice stuff it does-
porta while polyphonic
unison
aliasing and anti-aliased oscs
PW & PWM
hard sync
draw form with splines
ASM optimized envelopes
18 filter types, filter per osc
4, 3 mode mod sources with ~40 destinations
sequencer with customizable arp and multiple patterns
6 effects
RM & FM

Sampler:-
Simple yet versatile sampler, designed to load bar loops or picks, although it will accept tracks.
It has decent scope for changing the original samples with unison, porta, cut mod and gate/volume mod.
I just got the formant filter working, sounds cool whe modulated over whatever audio material.

I can't post the screens here coz of the size limitation,
but you can see them on KVR, here->

http://www.kvraudio.com/forum/viewtopic.php?t=393487

njoy!
Cheers, Nick

[trogluddite]

Got me thinking weather a Ruby "Midi to Poly" is feasible/possible/makes sense?

Well not exactly Ruby to Poly, I don't think, as the 'MIDI to Poly' prim' is the only way to create "white" poly voices. But a Ruby to "many mono" can be done - i.e. a fixed amount of polyphony as mono/mono4 'voices' instead of the 'white' poly where the voices are allocated dynamically. It can even be done using 'green' - I made one in SM a few years back.

It makes building a synth a bit more tricky, as you need to duplicate a lot of modules, but it does offer a way to get around a few of the limitations that the default Poly voice assignment has - e.g. you can use "free running" oscillators to avoid the usual "stage 0" oscillator reset at note on, and coding your own envelopes is more simple.

I've already ported my "MIDI to multi-mono" module to Ruby, but hadn't posted yet because I've never got around to making a decent example synth - but I'll stick it in a thread shortly so you can take a look.

@nix - Cheers man! It's really generous of you to post such a cool (and BIG) collection like this - the amount of work they must have taken you is incredible! I'm looking forward to find some time to fire up Reaper and have a play.

[billv]

Thanks Nix....heaps of great work here man....
Thanks for sharing.
I'm getting the hang of Entrancer more, the modulation matrix is making more sense now,
and I love this '?' button in the Osc section...WTF!...
Shifter was cool...
Sync was good.....but i did miss a "record on play' function....

Timing seems not to be sample accurate, however it is well within a usable precision.

S##t Nix...I think that's the whole "timing" story in one line.....
After finishing the timing section a few months ago, i go back regularly for the odd
timing test with X11.....fails everytime....I think were stuck with 'usable precision'.....

[nix]

Deffo it needs record on play- will do that this week-ta
I hope a couple of them want to hang in ur VST folder 8D

The '?' is random, built by Angular Momentum, and tailored by me.
I'll send you the module this week too. maybe there is scope for weighting it more to sine,tri etc.

Yeah this usable precision seems to be what we have-
one test I still have to do is load up the new VST example and play exactly the same sound with a host player,
and see if there is phasing. I am doing sample accurate things with Ruby,
but just not in audio threads. I have successfully made a recorder and Shifter uses 'Mono to graph' and a Ruby tick.
If it wasn't sample accurate I would be having helluva discontinuities in my final forms.
Certainly I am using no dynamic green connections in DreamSequence, only v for the body, and a MIDI out of a ruby codebox at the end.

nway-gotta do the test on our stock 'sample accurate' device.
btw, Orion's timing is just freakish good, and Usine is ~no better than us

[nix]

Random osc form generator fixed in Entrancer->
http://www.sendspace.com/pro/dl/j67w53
Press the ? button for a new form,
it will save with the project-
and no more hangy shenanigans when using this.

There was a change in the way strings are being handled.
My workaround was now obsolete anyway, so another bug bytes the dust

Bill- WTF indeed-FS3 so broke my neat string preset system.
the '?' osc mode now does what it is meant to

[billv]

load up the new VST example and play exactly the same sound with a host player,
and see if there is phasing

I did a bit before and failed with polarity testing..can't wait till somebody get's it right..

ScreenShot174.png (9.6 KiB) Viewed 31010 times

i did get one successful polarity test.... but in this particular test, I needed failure for the test
to be a success,and because the test was succsessfull, the end result was a failure..

I am doing sample accurate things with Ruby, but just not in audio threads

Yeh..i know what you mean.... sample accuracy is there....but proving tricky to maintain...
We could blame ourselves, (dodgy designs or whatever)but Trog's had a dig as well and come
up short, and he's no dummy yeh!..So the point is...
maybe give it about another year.....we can always contact malc if the frustration
is too much..

Usine is ~no better than us

I stopped using Usine straight away when i couldn't get a result in timing...
I found it nowhere near SM/FS....I spoke to Oliver about it, he said I could get the result
by bringing the samples in "Best block size" right down, but doing that turns the audio to shit
so became very pointless. Shame though, lots of things i like about Usine...
The hollyhock update looked interesting ...but not for XP users....

Random osc form generator fixed in Entrancer->

I'm still going through the rest of the pack....don't be in a hurry for a reply...

[nix]

Here is an update for the synth->
http://www.sendspace.com/pro/dl/j67w53
Should load quicker and be more stable now