Some output of the algo for those who want to see :
the beginnings of a beautiful era for audio dsp
-
- KVRist
- 92 posts since 26 Sep, 2005 from France
You do not have the required permissions to view the files attached to this post.
-
- KVRist
- 92 posts since 26 Sep, 2005 from France
And more :
You do not have the required permissions to view the files attached to this post.
-
- KVRer
- 21 posts since 6 Jul, 2016
Awesome, thank you.Chaotikmind wrote:Some output of the algo for those who want to see :
-
- KVRist
- 362 posts since 1 Apr, 2009 from Hannover, Germany
-
- KVRian
- 1379 posts since 26 Apr, 2004 from UK
Which algorithm?Chaotikmind wrote:Some output of the algo for those who want to see :
-
- KVRist
- 92 posts since 26 Sep, 2005 from France
The one explained by the OP ?
i didn't try to do more complicated things like using spherical coords, but seems promising too
i didn't try to do more complicated things like using spherical coords, but seems promising too
Last edited by Chaotikmind on Tue Dec 20, 2016 8:22 pm, edited 1 time in total.
-
Smashed Transistors Smashed Transistors https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=339459 - KVRist
- 142 posts since 10 Oct, 2014
Sorry not a native english speaker what "OP" stands for ?
The waveforms really look like what you get with self phase modulation of a sine oscillator (with direct feedback or with squared feedback).
The waveforms really look like what you get with self phase modulation of a sine oscillator (with direct feedback or with squared feedback).
-
- KVRian
- 1379 posts since 26 Apr, 2004 from UK
Original poster.
OK, I understand now what you mean. It would have been nice to also have the "circle" that you were using
OK, I understand now what you mean. It would have been nice to also have the "circle" that you were using
-
- KVRist
- 92 posts since 26 Sep, 2005 from France
i'll make a video with sound as soon as i have finished my work and have some time.
-
- KVRAF
- 2256 posts since 29 May, 2012
It's a Saṃsāra circle, as far as I could understand from what xoxos has written. Generally it is thought to be undesirable, but those who say it is undesirable appear to prefer nonexistence to it.
Returning back to dsp, changing "walking speed" along the circle mentioned is FM modulation, and changing the distance from the origin is AM modulation, but I could not understand what was the advantage of having an analogy concept that unifies the two, other than the fact that it has some usefulness in understanding things that may otherwise be counter intuitive. For example the instantanous phase is the integral of instantanous frequency as far as FM modulation is concerned, the way distance in general is the integral of speed. If this concept shall be a new era in dsp, I guess dsp is now considered to be dead, waiting to be resurrected for the next iteration of the Saṃsāra.
I have seen a few interesting things in the articles linked though (that spatial modulation concept), but I could not see their relevance to the topic, as spatial hearing cues are more complex than that, but perhaps the paper was about a simplified approach, I didn't really read carefully. A musical instrument that changes location is an interesting concept, do we really want that?
Returning back to dsp, changing "walking speed" along the circle mentioned is FM modulation, and changing the distance from the origin is AM modulation, but I could not understand what was the advantage of having an analogy concept that unifies the two, other than the fact that it has some usefulness in understanding things that may otherwise be counter intuitive. For example the instantanous phase is the integral of instantanous frequency as far as FM modulation is concerned, the way distance in general is the integral of speed. If this concept shall be a new era in dsp, I guess dsp is now considered to be dead, waiting to be resurrected for the next iteration of the Saṃsāra.
I have seen a few interesting things in the articles linked though (that spatial modulation concept), but I could not see their relevance to the topic, as spatial hearing cues are more complex than that, but perhaps the paper was about a simplified approach, I didn't really read carefully. A musical instrument that changes location is an interesting concept, do we really want that?
~stratum~
-
- Banned
- Topic Starter
- 12367 posts since 30 Apr, 2002 from i might peeramid
entertaining but not quite magnitude is not used. g'head and manufacture disinterest if you want, the proof is in the meaningfulness of parameterisation and ease, potential of signal modulation. a 2d form can be transformed in more ways than a 1d form. as chaotikmind will attest, eg. tilting your ellipse tilts the spectral energy, increasing high freqs and reducing low freqs. all the params are immediately meaningful to the novice in terms of both signal energy and geometry. otherwise, correct, there's no rule that states replies have to be either wholly pertinent, derivative or honest in intent or free of intent to mislead or obscure
there is no antialiasing (let's write that AA) but if you use efficient methods, the algorithm requires about the same computation power as a phase distortion implementation. (most efficient form = ~1.1% of my >decade ancient single core) no branching, one math function.
my schedule ahead is of course complexed by hols, the only thing i could add to the above images would be the accompanying orbits. before i talk more about this, there's more to explore.. i was rotating part of it "in the other direction" yesterday and had negative frequencies reflected back into the spectum from the low end.. i don't see that happen much.
(also considered duplicating, transforming and then subtracting/cancelling to eg. generate sets of partials) - my main interest is in synth sax, clar so mucking with higher freqs. and just seeing where it goes.
say you turn your circle into a rounded n-point star by modulating the radius by phase, no difference if we're only using the angle if it's in the center, but move it off center, harmonics... loads and loads to do.
magnitude is not used. g'head and manufacture disinterest if you want, the proof is in the meaningfulness of parameterisation and ease, potential of signal modulation. a 2d form can be transformed in more ways than a 1d form. as chaotikmind will attest, eg. tilting your ellipse tilts the spectral energy, increasing high freqs and reducing low freqs. all the params are immediately meaningful to the novice in terms of both signal energy and geometry. otherwise, correct, there's no rule that states replies have to be either wholly pertinent, derivative or honest in intent or free of intent to mislead or obscure there is no antialiasing (let's write that AA) but if you use efficient methods, the algorithm requires about the same computation power as a phase distortion implementation. (most efficient form = ~1.1% of my >decade ancient single core) no branching, one math function.
my schedule ahead is of course complexed by hols, the only thing i could add to the above images would be the accompanying orbits. before i talk more about this, there's more to explore.. i was rotating part of it "in the other direction" yesterday and had negative frequencies reflected back into the spectum from the low end.. i don't see that happen much.
(also considered duplicating, transforming and then subtracting/cancelling to eg. generate sets of partials) - my main interest is in synth sax, clar so mucking with higher freqs. and just seeing where it goes.
say you turn your circle into a rounded n-point star by modulating the radius by phase, no difference if we're only using the angle if it's in the center, but move it off center, harmonics... loads and loads to do.
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
-
- Banned
- Topic Starter
- 12367 posts since 30 Apr, 2002 from i might peeramid
ty, my development system is an aging netbook that runs opengl 1.4.. pretty advanced for me, but i've never "hit that point" where i can make screen capture videos without them being super crap. on the plus side, ebay is full of opportunities to replace it, if i need to, for around $30Chaotikmind wrote:i'll make a video with sound as soon as i have finished my work and have some time.
really thumbs up society, thanks for all the value.
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
-
- Banned
- Topic Starter
- 12367 posts since 30 Apr, 2002 from i might peeramid
those waveforms are produced by moving the region entirely from the center (> 1.0 radius). i'm finding 0.9 radius displacement to be sufficient for bright spectra.
regarding the theory of this oscillator, it solidified when i made the animated ephemeris posted in another thread. watching planets turn retrograde seemed like it would have interesting spectral potentials. so i see it as elementary and fundamental, as it conveniently makes the classic waveforms (sine and square exactly, saw and triangle approached), has a nice pulsewidth function, brightness. the user sees, squish the circle, move the circle, tilt the circle. simple.
no time this evening but had a thought,
hilbert transform uses quadrature oscs. so it should be possible to frequency shift the oscillator, expanding or compressing the spectrum, and transpose it back to the fundamental. not that that makes for greatly playable pitched instruments but should be a nice basic osc feature for weirdness. i'll be looking for ways to increase rippling/patterning in harmonics, all of that together ought to be useful.
it really discretises the notion, if you want to create more complexity, add a dimension, work there, then reduce back down.
regarding the theory of this oscillator, it solidified when i made the animated ephemeris posted in another thread. watching planets turn retrograde seemed like it would have interesting spectral potentials. so i see it as elementary and fundamental, as it conveniently makes the classic waveforms (sine and square exactly, saw and triangle approached), has a nice pulsewidth function, brightness. the user sees, squish the circle, move the circle, tilt the circle. simple.
no time this evening but had a thought,
hilbert transform uses quadrature oscs. so it should be possible to frequency shift the oscillator, expanding or compressing the spectrum, and transpose it back to the fundamental. not that that makes for greatly playable pitched instruments but should be a nice basic osc feature for weirdness. i'll be looking for ways to increase rippling/patterning in harmonics, all of that together ought to be useful.
it really discretises the notion, if you want to create more complexity, add a dimension, work there, then reduce back down.
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
-
- KVRAF
- 2256 posts since 29 May, 2012
Basically in truth I couldn't understand what you have done and what was special about it, and the articles that other contributors have linked seemed to be about something else. You know, if you deform the circle, it can also be a waveshaper, one tool for everything? That's fine, but we could do FM before, we could do AM before, and we could do waveshaping before, so, if you tell a bit more, we would be better educated. If it is a way presenting these ideas to the user so that they would be intuitive to the users, that may be interesting.magnitude is not used. g'head and manufacture disinterest if you want, the proof is in the meaningfulness of parameterisation and ease, potential of signal modulation. a 2d form can be transformed in more ways than a 1d form. as chaotikmind will attest, eg. tilting your ellipse tilts the spectral energy, increasing high freqs and reducing low freqs. all the params are immediately meaningful to the novice in terms of both signal energy and geometry. otherwise, correct, there's no rule that states replies have to be either wholly pertinent, derivative or honest in intent or free of intent to mislead or obscure
~stratum~