Somatic Vision

I saw that the smoothness meter in the Dual Drive is computed as a 3 min. average. Thus if the smoothness dropped for a few seconds this will hardly change the average. Can I shorten this interval to get a more real-time feedback? Or maybe there are other games that allow this?

Also you wrote earlier that there is no implementation of BVP as a feedback in the games. However how can I use the BVP graph for feedback?

Hi Mordavid,

What makes you think the smoothness meter is computed as a 3 minute average? If you are using the emWave, or Alive with an emWave and have it set to use the emWave Coherence algorithm, it may be 3 minutes, I don't know. The Alive / Dual Drive algorithm that Somatic Vision designed changes within a few seconds, it depends on various factors but you can expect to see an increase in smoothness within 2 to at most 10 seconds.

Regarding BVP feedback, you can train using the BVP screen, watching the amplitude of the BVP. We are working on a major Alive add-on that gives you new training measurements (BVP and others). These new measurements will work with all of Alive and come with new graph training screens and environments. To use the BVP graph for feedback, relaxation is a vertical expansion of the BVP wave (higher amplitude, greater bottom to top distance). Tension shrinks (vertically) the bvp wave. It will be easier to train with this once we release the new graphing screens and measurement specifically designed to help you work with BVP.

Best,

Ryan

Ryan Deluz wrote: What makes you think the smoothness meter is computed as a 3 minute average?

In the smoothness graph mode I've noticed that I can make the smoothness graph go up or down very quickly while the smoothness meter will change quite slowly (the meter can show above 90% while the real time smoothness dropped to zero and went back to 100%). Is the real-time data used in any of the training options? Or can it be translated to a simple audio feedback?

I think I understand what you mean. On the right side there is a slider bar. It goes from red/orange to green. Green is highly smooth. This bar should be able to move quite quickly. The number under the bar is the session average. So if the smoothness bar has spent half the time all the way to the right, and half the time all the way to the left, this number will say 50% smoothness. The average smoothness is not used in the real-time training. So, for example in Dual Drive the car will go fast, and the audio loud, when the smoothness bar is the right, even if the average smoothness is 5% for the whole session. All (*most) of the feedback is based on the momentary changes in the smoothness bar. The session average is not used for feedback, except for the fact that it is of course visible on screen. This average % doesn't control any visual, auditory or game feedback.

Might it be possible (I understand there may be proprietary issues involved) to share some description of the algorithms involved in the calculations of HeartMath's "Coherence" and Alive's "Smoothness"? & the differences between these? I believe I understand the basic principles involved, but would like to have a better appreciation of the details involved in these functions.

HeartMath calculates coherence, so I can't comment much on the details. Also, Somatic Vision's smoothness algorithm is proprietary and a trade secret.

That said I can say some things about these.

In both cases (coherence & smoothness) an FFT is first used to determine the power of various spectral bands, this means how much of a certain frequency there is. As you breathe in your heart rate increases, as you breathe out it decreases, so if your breathing is smooth and steady, and you don't have much other stress that would cause rises or variations in heart rate, your heart rate will look something like a sine wave (not perfect, but similar) where the frequency is how fast you are breathing. Therefore, doing frequency analysis on the heart rate allows you, for instance, to see if the breathing is smooth and steady and there isn't a lot of other inputs that would make the heart rate wave jerky. If the heart rate moves up and down quickly, that would be a higher frequency wave.

In heart rate variability (HRV) analysis, there is a concept of low frequency, which includes 0.1Hz (10 seconds per cycle). So a 10 second breathing time would be a 0.1Hz wave. LF technically is 0.04 to 0.15Hz. Higher frequencies (HF) indicate additional, normally irregular but quickly changing inputs that indicate some form of stress or less positive state. So its common to use LF / HF as a basic measure of this sort of coherence/smoothness training.

Somatic Vision Smoothness is more of an LF / Other (LF compared to all other frequencies), and not LF / HF. Also, LF isn't exactly 0.04-0.15Hz in the Somatic Vision algorithm. Finally, different algorithms choose a different length of input data. If you take a lot of input data, it takes longer to detect changes as the changes are sort of averaged over the length of time.

Somatic Vision Smoothness has further optimizations designed to immediately (well it's more like a 1-5 second delay) show changes in your state. Coherence is a slower algorithm that uses more data. We have spent a lot of work optimizing our algorithm to attempt to as quickly as possible reflect changes in stress/emotional state, as it is too confusing, in our opinion, if you make an inner change at a certain time, but only see that change on the screen 15 seconds later. Studies have shown that this delay in feedback creates problems, generally people assume whatever they did immediately before the change caused it, so often times if feedback is slow people think that since they just starting doing B, that B helped them, when it was really A that they did earlier that made a difference.

In any case, this is some additional description of Smoothness and Coherence. To summarize, both are based on frequency analysis of the heart rate, and both reward increases in the low frequency (LF) bands that correspond to slow, relaxed breathing and other positive physiological states (such as nervous system balance).