E-stim waveform A/B testing

This is the place for all suggestions, releases and feedback regarding Milovana Hardware efforts.
Post Reply
diglet
Explorer At Heart
Explorer At Heart
Posts: 390
Joined: Sun Dec 11, 2022 5:43 pm

E-stim waveform A/B testing

Post by diglet »

Recently, I've added A/B test functionality to Restim. This allows you to quickly compare two different signals. To access this functionality go to the device selection screen, select threephase, then select A/B test waveform.

I'd like to collect data on how the different signal generation parameters result in different sensations in order to develop better scripting tools. I would like to ask the community to aid in the research, you can help if you have a stereostim box. I came up with a couple of things that can be tested, but feel free to come up with your own tests if you feel I missed anything.

In all tests, we want to figure out how the sensation changes while controlling for the subjective intensity difference between the two signals. So, adjust the A/B volume until both signals have the same subjective intensity.

Download latest version of Restim here: https://github.com/diglet48/restim/releases

experiment 1: Pulse frequency at low pulse widths. At which frequency does the signal become smooth?.

Background: pulse based boxes and Restim in pulse-based mode sends short pulses with a frequency from 1 to 100hz. Literature says that if pulse frequency is low, (<30-40hz) pulses are felt as independent events. If pulse frequency is high (>60-80hz), pulses are felt as continuous event. Between those frequencies a phenomena occurs that's called 'flutter'.

We would like to know at which point the pulse frequency becomes high enough that the signal becomes fully smooth and discerning between two signals with different pulse frequencies is no longer possible.

Test setup: carrier frequency 1000hz (or other value of your choosing), pulse width 4. Set pulse frequency to different values for A and B (for example 30hz vs 40hz). Adjust the A/B volume until both signals have the same intensity. Can you still distinguish between the A and B pulse?


My results:
40hz vs 50hz: yes
50hz vs 60hz: unsure, extremely small difference.
60hz vs 80hz: no difference
Edger477's results:
I can definitely feel difference between 60 and 80, but it does become very subtle. Adjusting intensity to like 94% for 80 makes it so that it is feels the same intensity wise but there is small difference in texture.

Also, I love to use signals above 100hz pulse freq, with narrow pulses, small differences are hard to notice but such high freq pulses feel more and more like pressure, arousing.
experiment 2: Pulse width at high pulse frequencies.

Background: in the previous experiment we identified the pulse frequency at which the signal becomes smooth. We would like to know how, and if, the sensation of the signal changes when the pulse width (Restim definition) is increased when pulse frequency is in the 'smooth' region.

Test setup: carrier frequency 1000hz (or other value of your choosing), set the pulse frequency to approximately the value you found in experiment 1 (if you don't know, set 60hz). Select different A/B values for pulse width and pulse rise time, taking care to keep duty cycle below 100%. Adjust the A/B volume until both sigals have the same intensity. Can you still distinguish between the A and B pulse?

My results:
50hz, 2 rise time, 4 vs 18 pulse width: feels identical
50hz, 9 rise time, 4 vs 18 pulse width: unsure, extremely small difference.
50hz, 2 vs 9 rise time, 18 pulse width: feels identical
experiment 3: Pulse width at low pulse frequencies.

Background: At lower pulse frequencies, there definitely is a difference between short and wide pulses. We would like to know how the sensation changes as the pulse shape changes at low frequencies.

According to literature, the subjective frequency is determined by the period of silence between the pulses, not the pulse frequency. We would expect to see wider pulses to have higher subjective frequency.

Test setup: carrier frequency 1000hz (or other value of your choosing), set the pulse frequency to 30hz. Select different A/B values for pulse width and pulse rise time, taking care to keep duty cycle below 100%. Adjust the A/B volume until both signals have the same intensity. Can you still distinguish between the A and B pulse?

My results:
30hz, 5 rise time, 10 vs 4 pulse width: Small difference, unsure how it feels
30hz, 2 rise time, 10 vs 4 pulse width: Big difference. A feels more gentle, more random as if there are multiple layers of movement.
30hz, 10 rise time, 20 vs 4 pulse width: A feels like higher frequency. Can pinpoint location of sensation more accurately.
30hz, 2 rise time, 20 vs 4 pulse width: A feels like higher frequency, like electric toothbrush.
30hz 20 wide versus 41hz 4 wide, 9 rise time: When signal is static, very difficult to tell the difference. When signal is moving fast (circle pattern), A feels more complex, and random, like 10hz rubbing. Very interesting sensation.

It seems that making the pulse wider increases the subjective frequency. Comparing a wide pulse with a slightly higher frequency narrow pulse allows us to compare two signals with the same subjective frequency. For a static signal the difference between two such signals becomes very hard to notice. But when the signal is moving the two signals feel completely different.
experiment 4: carrier frequency..

Background: In theory, different carrier frequencies target different nerve fibre diameters, therefore changing the carrier frequency should change the type of sensation.

Test setup: Try comparing 500hz to 1000hz or 2000hz and write down your findings. You might need a stronger power supply (more than 12V) to test frequencies above 1000hz.

My results:
Not tested yet
picofarad
Explorer
Explorer
Posts: 59
Joined: Sun Jul 16, 2023 7:56 pm
Gender: Male
I am a: Switch

Re: E-stim waveform A/B testing

Post by picofarad »

This should be a fun test. I will try to do it this weekend and then report my results.
I have a stereostim amp and built a box with a set of transformers and just a set of hi-power resistor in series.
PossibleSpirit5998
Explorer
Explorer
Posts: 13
Joined: Mon May 03, 2021 4:52 pm

Re: E-stim waveform A/B testing

Post by PossibleSpirit5998 »

I'm not sure if I test correctly.

with the coyote 2 box. so had to select 3phase.


A)

40 50 Y
60 50 Y
60 80 Y


B) Feels all the same? (used 60hz)

C) cant really say anything feels much different A to B
Some rows themselves feel different one to the next, but cant feel much left right different.

D) you cant even select more than 1000 in carrier in A/B test.
But yeah feels different if I select different frequencies

So frequencies are easy to distinguish. The other stuff not.
Electro
Explorer At Heart
Explorer At Heart
Posts: 562
Joined: Thu Feb 13, 2020 9:45 am

Re: E-stim waveform A/B testing

Post by Electro »

PossibleSpirit5998 wrote: Sat Jul 20, 2024 6:23 pm I'm not sure if I test correctly.

with the coyote 2 box. so had to select 3phase.


A)

40 50 Y
60 50 Y
60 80 Y


B) Feels all the same? (used 60hz)

C) cant really say anything feels much different A to B
Some rows themselves feel different one to the next, but cant feel much left right different.

D) you cant even select more than 1000 in carrier in A/B test.
But yeah feels different if I select different frequencies

So frequencies are easy to distinguish. The other stuff not.
If you aren't using xtoys.app to "see the audio", you are wasting your time trying to use the official DG Lab app. Be sure continuous mode is turned on and not pulsed mode. For everything else, if you are trying to have carrier frequency manipulating the pulse frequency of the Coyote, set a pulse frequency script as the carrier frequency using the ranges that would match the correct pulse frequency of the pulse frequency file(Xtoys has a default range that can be changed) because pulse speed is what Coyote frequency is. The amplitude should move left and right with the 'circle' but the 'map to edge' function might make more sense for this device. Apparently the Bluetooth reaction time is faster with the Coyote 3.0, so maybe this would work reasonably, but the original Coyote(2.0) is too slow for anything but the slowest funscripts to work well for me but there might be some slower stuff that could do the job otherwise it might just get jumpy. Watching what xtoys is showing in its graphical display will help.
diglet
Explorer At Heart
Explorer At Heart
Posts: 390
Joined: Sun Dec 11, 2022 5:43 pm

Re: E-stim waveform A/B testing

Post by diglet »

To clarify, I'm interested in the results from stereostim boxes. All the other boxes mangle the signal in some way, very interesting stuff but not really what I'm looking for here.
picofarad
Explorer
Explorer
Posts: 59
Joined: Sun Jul 16, 2023 7:56 pm
Gender: Male
I am a: Switch

Re: E-stim waveform A/B testing

Post by picofarad »

As best I could,I ran through these tests.
Here is a link to my hardware setup: viewtopic.php?p=361126#p361126

40hz vs 50hz: not much difference
50hz vs 60hz: no real difference
60hz vs 80hz: very slight difference
40 v 90 can feel space at 40, buzzy smooth at 90
100 and 120 were good

50hz, 2 rise time, 4 vs 18 pulse width: 4 feels ever so slightly more intense
50hz, 9 rise time, 4 vs 18 pulse width: unsure, extremely small difference.
50hz, 2 vs 9 rise time, 18 pulse width: feels identical

30hz, 5 rise time, 10 vs 4 pulse width: not much difference10 could be a punishment signal with at ad more volume.

30hz, 2 rise time, 10 vs 4 pulse width: difference10 could be a punishment signal with at ad more volume.
.
30hz, 10 rise time, 20 vs 4 pulse width: A feels like higher frequency. Can pinpoint location of sensation more accurately. 20 feels like a good pounding. 4 is a bit smoother

30hz, 2 rise time, 20 vs 4 pulse width: A feels like higher frequency, like electric toothbrush. 20 feels way stronger

30hz 20 wide versus 41hz 4 wide, 9 rise time: When signal is static, very difficult to tell the difference. When signal is moving fast (circle pattern), A feels more complex, and random, like 10hz rubbing. Very interesting sensation. Agree

500hz to 1000hz 1000 feels smoother to me.
N.B., I could not set the frequency above 1000Hz
diglet
Explorer At Heart
Explorer At Heart
Posts: 390
Joined: Sun Dec 11, 2022 5:43 pm

Re: E-stim waveform A/B testing

Post by diglet »

Thanks for testing! It seems that you can discern slightly higher frequency signals than me, the rest seems quite similar.

To enable frequencies above 1000hz, you can change the safety limits immediately after selecting the waveform type.
picofarad
Explorer
Explorer
Posts: 59
Joined: Sun Jul 16, 2023 7:56 pm
Gender: Male
I am a: Switch

Re: E-stim waveform A/B testing

Post by picofarad »

No problem. I will change the frequency now that I understand what to do and run a comparison for higher than 1000Hz.
For what it's worth I wanted to mention electrode type and placement. Using a triphase with A on a metal Sound, B on a bronze clamp attached around the sack, and Common is a stainless-steel butt electrode using just the internal and not the base. The common has the largest surface area so little to no sensation. I use Spectra 360 Electrode Gel on all electrodes.
diglet
Explorer At Heart
Explorer At Heart
Posts: 390
Joined: Sun Dec 11, 2022 5:43 pm

Re: E-stim waveform A/B testing

Post by diglet »

I tried to test frequencies, but I got strange results. All tests in single-channel, no threephase.

Electrode config 1: 2 pads at the base of shaft, left and right.
At a frequency of 1000hz, pulse frequency 60, it feels like a gentle hum at the bottom of glans. As the frequency increases to 6000hz, the sensation moves to the top of the glans, becomes smoother and feels more sharp (smaller area).

Electrode config 2: 2 pads at the base of shaft (combined into 1 electrode), glans loop. Same results.

Electrode config 3: Glans loop, and double ring around shaft below balls. Here the sensation stays in the same spot. 1000hz definitely feels more rumbly and 6000hz feels a bit more smooth, pressure, squeezing, pulling back the foreskin. 2500hz was bit of a mix between the two sensations.

Electrode config 4: 2 pads at the base of shaft (combined), double ring around shaft below balls. Very variable results, poking the electrodes or meat with a pen a little bit drastically moves the sensation. The only consistent factor is that 1000hz feels more rumbly and 6000hz more sharp and focused.

Electrode config 5: Glans loop, and loop around middle of shaft. Very similar to 1/2, rumble at bottom of glans at 1000hz, sharp pain top glans at 6000hz, starts at lower volume levels.

Electrode config 6: Loop around middle of shaft, double ring around shaft below balls. Quite similar to 4, very variable sensations when moving electrodes or meat a little.

Conclusions:

In all tests 1000hz feels more buzz, rumbling, larger area. 6000hz feels more smooth, constant pressure, sometimes sharp pain, smaller area.

If current crosses over the dorsal nerve, higher frequencies seem more likely to generate the sharp pain at top of glans often associated with direct dorsal nerve stimulation. We discussed on discord, and we think this might be caused by the current taking a different path through the body when frequency is higher (more likely to hit dorsal nerve). But we are not sure why.

We might have to test on a different body part that doesn't have as many big nerve bundles close to the surface.


For anyone wondering about safety, 6000hz is safe when using low duty cycle signals. Very unsafe when using continuous signals.
Corndogville
Explorer
Explorer
Posts: 45
Joined: Wed May 05, 2021 10:56 pm

Re: E-stim waveform A/B testing

Post by Corndogville »

Why did this thread die out?

I have a question. As someone who's used to audio files (and never goes over 1000hz) should I be using continuous? I've been using pulse because I am on a stereo box, but wondering if maybe I'm missing out on something. FOMO lololol
User avatar
edger477
Experimentor
Experimentor
Posts: 1114
Joined: Mon Nov 29, 2021 8:24 pm
Location: underfloor

Re: E-stim waveform A/B testing

Post by edger477 »

Corndogville wrote: Thu Dec 11, 2025 8:21 pm Why did this thread die out?

should I be using continuous? ... I am on a stereo box
No, pulsing is for stereo box. Continous kinda works on ET312. Pulsing is way to go on stereostim.
My estim creations: https://mega.nz/folder/73pxmBBQ#X6ylDzRafzTt9wanZ0dacw
And in E-Stim Index: viewtopic.php?t=27090

Try creating your own estims with my restim script generator!
Spoiler: show
You can also thank me with crypto: https://trocador.app/anonpay?ticker_to= ... e+a+coffee
Post Reply