Milovana.com

https://cfs6t08p.github.io/funstim/funstim.html
Hello friends.
I have a few questions about the funscript converter.
What settings did you set there?
which settings make which changes?


Who can explain?

Fade in/out
Sample rate
Frequency (for et312)

Inverted
Double time (Every action is one full up & down motion)
Expand to full range
Fade out during pauses/slow sections
Output amplitude & phase envelopes (FOR TESTING ONLY)

I would be grateful for an explanation and settings for the best results for Estim

Big Thx

elektrowichser wrote: Sat Feb 18, 2023 9:48 am https://cfs6t08p.github.io/funstim/funstim.html
Hello friends.
I have a few questions about the funscript converter.
What settings did you set there?
which settings make which changes?


Who can explain?

Fade in/out
Sample rate
Frequency (for et312)

Inverted
Double time (Every action is one full up & down motion)
Expand to full range
Fade out during pauses/slow sections
Output amplitude & phase envelopes (FOR TESTING ONLY)

I would be grateful for an explanation and settings for the best results for Estim

Big Thx

@edger477 is the author I hope he can give the best explanation. I am just enjoying the tool

BTW you have the new version for testing control over the levels per channel:
https://edger477.github.io/funstim/funstim.html

--
jgrants

jgrants wrote: Sat Feb 18, 2023 10:34 am

elektrowichser wrote: Sat Feb 18, 2023 9:48 am https://cfs6t08p.github.io/funstim/funstim.html
Hello friends.
I have a few questions about the funscript converter.
What settings did you set there?
which settings make which changes?


Who can explain?

Fade in/out
Sample rate
Frequency (for et312)

Inverted
Double time (Every action is one full up & down motion)
Expand to full range
Fade out during pauses/slow sections
Output amplitude & phase envelopes (FOR TESTING ONLY)

I would be grateful for an explanation and settings for the best results for Estim

Big Thx

@edger477 is the author I hope he can give the best explanation. I am just enjoying the tool

BTW you have the new version for testing control over the levels per channel:
https://edger477.github.io/funstim/funstim.html

--
jgrants

wow the new version is very great thx

An exact instruction manual would still be very nice

I am not sure how exactly et312 processes the signals, and how triphase signals go through. Do you notice difference between my version and original cfs6t08p converter?

I can only test it tomorrow.
what settings can you recommend?
on the et312 i use audio3
and trephase

and the elektrode configv on my cock
common on shaft
a+ on cockhead
b+ anal

I don't have et312 so I can't recommend anything, but I expect that its processing will cause normal triphase signals (from cfs6t08p converter) to be "fine" (without too strong signals on common), and if that is true, then my converter will cause the common signal to be too weak. But try and see.

elektrowichser wrote: Sat Feb 18, 2023 11:28 am I can only test it tomorrow.
what settings can you recommend?
on the et312 i use audio3
and trephase

and the elektrode configv on my cock
common on shaft
a+ on cockhead
b+ anal

Try this:
1 sec Fade in/out
DEFAULT values for all else (load the URL in new window to see defaults)

The ET-312 will do a decent job on that output.

The tricky part is the frequencies, since the ET-312 counts zero crossings to resynthesize an output wave and allegedly does not produce output higher than 400 Hz. The default wave of (420,520,620) Hz gives it a lot of extra zero crossings, so I am not sure what it really does with that. You can experiment with single frequencies to see what happens. I would expect frequencies just under 400 to be most predictable, but I haven't tried. Above 400, it either maxes out to 400 or it aliases in some way.

An improvement:
Use Audacity to compute (L+R,L-R)/2.
In other words, open the cfs6t08p output mp3 in Audacity, split the stereo track into two mono tracks (use the dropdown in the track's title), pan both mono tracks to center, set both volumes to -6dB, select the two tracks and do Tracks > Mix > to new track, which will be the new (L+R)/2. Then invert the right mono track (select it and go Effects > Invert) and repeat to get the new (L-R)/2. Pan (L+R)/2 to left and (L-R)/2 to right, mute all other tracks, and play. You can also combine the two new tracks into one stereo track and/or export as mp3.

This SHOULD give your ET-312 much better material to work with, by "baking in" the tri-phase interference into the signal itself, so the ET-312 will just follow the amplitude and it won't matter so much what it does to the frequency/phase, zero crossings, etc. I say "SHOULD" because I have not used my ET-312 since discovering this technique, but extrapolating from my decade-plus use of the ET-312 and my current use of this technique with other devices, I expect this to work more reliably with the cfs6t08p output than straight tri-phase. It should be more robust vs. the frequencies you enter into the tool. Connect the new left channel "(L+R)/2" to A+ on the cock head and the new right channel "(L-R)/2" to B+ on the anal plug.

In general, same goes for the 2B, though its algorithm differs from the ET-312 and I don't know the details. I have actually used this Audacity conversion with the 2B with some success. The 2B really throttles fast-moving waves though, so it is useless on the faster Cock Heroes. But it is GREAT on medium/slow/smooth signals.

About sampling rate:
It is tempting to think you will get a smoother or higher-quality stim with a higher sampling rate, but this is not true. Unless your stim signal has frequencies above 4000 Hz (WHY?), you have nothing to gain by sampling above 8000 Hz. Today's DAC outputs are very high quality and will interpolate your 8000 Hz samples with all the fidelity your dick could possibly need. The only thing to gain by increasing that sampling rate is wasted disk space (more bytes per sec). Or, I guess if your playback system is very rudimentary and can only handle 44.1k or something, that would be the only real reason to change it.

An 8000hz sample rate with a 600hz signal is 13.3 samples per sine wave, or about 6.6 per each half cycle(up and down). Higher frequencies could get ragged for the output and combining frequencies while doing that will look more ugly. Your example of 4000hz signal with 8000hz sample rate is going to give you two points, you aren't going to make a sine wave with two points.

If you don't believe me, create a 1200hz signal with a funstim conversion and zoom in with Audacity until you see the individual samples with 8000hz and you will see you aren't getting a good signal out of it when a round curve is defined by 3 points. I think a 600hz single frequency would be fine for 8000hz sample rate, but anything more and it's worth going up a step. 22.5khz would be plenty for our purposes.

Two points is enough to create a perfect sine wave of half the sampled frequency. See Nyquist-Shannon theorem. VoIP codecs commonly sample at 8 kHz.

While the digital representation will look choppy in a tool like Audacity, measuring the DAC output with an oscilloscope would show a perfect signal.

Electro wrote: Sat Feb 18, 2023 8:36 pm An 8000hz sample rate with a 600hz signal is 13.3 samples per sine wave, or about 6.6 per each half cycle(up and down). Higher frequencies could get ragged for the output and combining frequencies while doing that will look more ugly. Your example of 4000hz signal with 8000hz sample rate is going to give you two points, you aren't going to make a sine wave with two points.

If you don't believe me, create a 1200hz signal with a funstim conversion and zoom in with Audacity until you see the individual samples with 8000hz and you will see you aren't getting a good signal out of it when a round curve is defined by 3 points. I think a 600hz single frequency would be fine for 8000hz sample rate, but anything more and it's worth going up a step. 22.5khz would be plenty for our purposes.

If you have frequencies in the realm of 400, 600, etc., WAY below 4000 Hz, then you are nowhere near needing 8000Hz sampling. The default 8000 is more than adequate for any e-stim signal. It does not matter how ugly the wave form looks. But yes, even if you had a 4000Hz wave, 8000 Hz would be adequate, barely adequate, but ENTIRELY adequate to reconstruct the wave from a properly sync-interpolated DAC.
(CORRECTION: my original note about 3999.9999Hz was wrong--I was thinking of hitting the sampling rate and degenerating to DC. But the above works right up to 4000Hz.)

The 1200 Hz signal is fine at 8000 Hz sampling. That is simply factual in the field of electrical engineering. It flows from examining the Fourier transform of a step-wise sampled signal and observing that an interpolation of sinc(x) reconstructs every original wiggle that can be contained in an input of that limited bandwidth, EVEN IF you have only the two minimum points. It does hinge on having a high-quality DAC, but we all have them nowadays in every device.
(CORRECTION: I wrote "sync" instead of "sinc"! Corrected.)

When you occasionally hear a low-quality waveform sampled at 8000Hz, like some of our speech/phone sounds, it is due to poor filtering as the waveform is sampled, resulting in aliasing. This is actually harder/more expensive to do than the DAC, so our low-end systems do sometimes bring us bad-sounding 8000Hz signals, but that's not the fault of the sampling rate. For computing e-stim waves, none of this is relevant at all.

Ask any electrical engineering prof. This is basic EE 101. It's counterintuitive and I'm not bashing anyone for not automatically knowing it. I'm just saying... this stuff is hard fact. You DO NOT GAIN any quality at all, in any way, by sampling faster than 2x your frequency content. If you're entirely under 4000Hz, then 8000Hz is all you need.

jtcrave69 wrote: Sat Feb 18, 2023 10:12 pm Two points is enough to create a perfect sine wave of half the sampled frequency. See Nyquist-Shannon theorem. VoIP codecs commonly sample at 8 kHz.

While the digital representation will look choppy in a tool like Audacity, measuring the DAC output with an oscilloscope would show a perfect signal.

LOL, you said that a LOT more efficiently than I did!

zebbg69 wrote: Sun Feb 19, 2023 12:14 am
LOL, you said that a LOT more efficiently than I did!

Yes, but I left out the more interesting stuff that you covered. Fourier transforms blew my mind in college.

The reason it's counterintuitive is that we visualize waveforms in the time domain and assume we must recreate them by connecting dots. But the information needed to create waveforms can be reduced, because we know there are limits to what is physically possible. An analog waveform will never have instantaneous jumps or breaks. It moves continuously from one sample to the next. The range of possible frequencies is limited due to filtering - natural or artificial. The Nyquist-Shannon theorem mathematically proved the minimum information necessary to describe a waveform. How it is done is a whole field of study that it would appear you have some experience with.

All this science being applied to zapping our sensitive bits ... We stim on the shoulders of giants!

A better intuitive way to explain the sampling rate question:

When you slap a pool of water, its ripples are always rounded sine waves regardless of how you slap it. Your hand can make the most jagged, choppy motions, but the ripples will always do what water does. Electrical circuits are the same way: they have a bandwidth that limits the speed at which waves can develop.

A proper DAC has an output bandwidth matched to the sampling rate of the input signal, so even with that 4000 Hz signal and only two dinky points per cycle making such an ugly wave, those two points are "slapping the water" at the right time to make those ripples at precisely 4000Hz, and you get a perfect reconstruction.

In more math-y terms, the DAC is a curve fitter. It fits a curve to those points, constrained by the bandwidth. The only curve you can fit to that 4000Hz example, ASSUMING you are limited to 4000Hz, is a perfect 4000Hz sine wave.

elektrowichser wrote: Sat Feb 18, 2023 11:28 am I can only test it tomorrow.
what settings can you recommend?

How was your test? Did you find some good settings?

zebbg69 wrote: Thu Feb 23, 2023 8:04 pm

elektrowichser wrote: Sat Feb 18, 2023 11:28 am I can only test it tomorrow.
what settings can you recommend?

How was your test? Did you find some good settings?

yes
I have the best results with 5 sec fade in out and frequency 777th in threphase. makes the best feelings on my cock