To better understand sensor characteristics, I collected a decent amount of data.

I will first look at general patterns in orgasm data. Second, I will take a closer look at specific recordings.

AS5311 girth sensor

Averaging many different trials allows us to observe common factors.

Observations:

• During orgasm start, there is an unusually strong clench. This corresponds with the contraction of Ischiocavernous muscle observed by Gerstenberg 1990. Girth data seems to be a good proxy for this muscle.
• For orgasm to occur, the sensor readings have to be quite high. The maximum sensor reading often occurs during orgasm. Therefore, the sensor reading can be a proxy for closeness to orgasm.
• The amplitude of the rhythmic contractions generally is quite low compared to max voluntary contraction.

Pressure sensor

Top pane is the girth sensor. Bottom pane is pressure with inflatable buttplug.

Observations:

• At the start of the orgasm, the butt pressure drops significantly. There is a peak in the girth data, which falls about 2 seconds after the butt pressure drops.
• This matches the data from Gerstenberg 1990.
• The average pressure is almost constant for 50 seconds before orgasm. This suggests that pressure readings give no information on closeness to orgasm.
• The amplitude of the rhythmic contractions generally is quite low compared to max voluntary contraction.

EMG sensor (side-to-side)

Data collected with anuform probe with side-to-side contacts.

RMS is the RMS voltage of the sensor, PCA1 is the first principal component calculated from short-time fourier transform. In practice this is almost identical to weighting the frequency bins by freq^2

In the first trace, we see the familiar sight of the butt muscles relaxing about two seconds before the girth data drops and the contractions begin.

In the second trace, there is a much larger gap of about 10 seconds. I will explain why this happens later.

EMG sensor (top-to-bottom)

Data collected with Large Flo (EStim-Systems brand electrode).

Very similar to side-to-side probe. The latency between EMG drop and girth drop is 2.7 seconds.

Pressure + EMG sensor (bulbospongiosus muscle)

Data collected with inflatable buttplug and 2 EMG pads positioned between the butthole and balls. This placement measures the bulbospongiosus muscle.

The main point of interest is the large amplitude of the first few contractions, which are higher than elsewhere in the data trace. The amplitude decreased once I turned off the stimulation. This matches the data from Gerstenberg 1990.

It looks like the correlation between EMG readings and butt pressure is poor, I will show later that it's actually quite good.

Direct comparison between EMG locations

A direct comparison can be made by looking at the raw EMG waveforms with different probe locations (three different traces).

The data indicates the sphincter muscles don't contribute much to the rhythmic contractions. Their amplitude is far below max voluntary contractions.

You can't see it on this graph, but probe EMG data slightly leads the contractions, and BCM EMG data slightly lags.

Longer traces

Data collected with anuform probe with side-to-side contacts.

There is a very pronounced on-off pattern in the EMG data, which I observed in pressure data as well. It appears there are psychological limits to how long one can clench.

Shafik 2006 explains the phenomena as follows:

Under normal physiologic conditions, striated muscles continue contraction for a period of 50–70 seconds, after which time they relax spontaneously. Muscle recontraction may occur after a few seconds. [...] On cavernosus muscle contraction, the ICP increases to above the systolic pressure [and] are apparently transmitted into a high-tension closed cavity. Retention of this high-tension closed cavity [...] may lead to cavernosus tissue ischemia. Therefore, the intermissions in cavernosus muscle contractions allow for alterations in the periods of muscle contraction and relaxation during which the cavernous tissue could be well oxygenated, particularly because the periods of cavernosus muscle contraction are relatively short.

The duration of the 'on time' maxes out at about 2 minutes, but is usually shorter. After this 'on time', there is an 'off time', a period of 1 minute is fairly typical. During this time, involuntary clenching does not occur but orgasm is still possible.

In most traces the butt pressure drops 2-3 seconds before girth drops. But in some traces, the gap is 10-20 seconds. I believe this occurs when orgasm is initiated during the 'off time'.

Correlation between bulbospongiosus muscle and pressure

There is decent correlation between readings of pressure and EMG readings of bulbospongiosus muscle. The following image shows the pressure reading and the predicted pressure from a linear model with only one EMG feature.

As expected, the correlation breaks down heavily during rhythmic contractions.

The correlation with girth data (now shown) is not very convincing. This suggests that the anal sphincters and BCM are quite correlated, but the ICM is capable of doing its own thing.

Conclusions

Before orgasm:

• Pressure, butt EMG and Bulbospongiosus muscle EMG data are all well-correlated.
• Pressure and Girth correlation is only so-so. Measuring both could be valuable.
• All sensors show very pronounced on-off pattern about 1min long each.

Leading up to orgasm:

• Pressure data shows no useful information for determining closeness to orgasm.
• Girth seems to be an indicator of distance to orgasm.

During orgasm:

• Forced relaxation of anal sphincters and BCM.
• Strong contraction of ICM muscles, increase in girth for about 2-4 seconds.
• After ICM muscle relaxes, rhythmic contractions start.
• Amplitude of BCM contractions is higher than max voluntary contraction.
• Amplitude of anal sphincter contractions is medium.

A simple algorithm that uses girth as a proxy for arousal is implemented in Restim.

References

Shafik 2006: The cavernoso-anal reflex: response of the anal sphincters to cavernosus muscles’ stimulation
Gerstenberg 1990: Erection and Ejaculation in Man. Assessment of the Electromyographic Activity of the Bulbocavernosus and lschiocavernosus Muscles