I doubt they actually increase the quality of perception, they just make you feel like you can see more clearly. That would be an interesting study to do, though, whether people on psychedelics can see things differently in an externally measurable way. A different sort of study has been performed that measured that people who had used psychedelics in the past were less able to see strobe lights at certain frequencies than people who had never used them, with people who suffered from HPPD being even more impaired.
So they are essentially like a "sharpen" image filters.
Sharpening filters don't actually sharpen an image, you would need deconvolution for that and it is really hard to do and require accurate knowledge of the process that made the image blurry. And even the best algorithms can't recover information that wasn't here in the first place.
What sharpening filters do is that they increase contrast locally, so instead of having light grey -> dark grey, you have light grey -> white -> black -> dark grey. The sharp "white -> black" transition gives off the illusion that the image is more detailed than it really is. In fact, since it is generally a type of convolution filter, you are actually destroying data. Overdo it and you get a noisy mess.
Thats exactly right, except at least you don't get people walking around saying the sharpen filter "shows you reality like it really is" or "elevates your state of consciousness."
One thing I noticed one time on mushrooms was that certain scenes in video became absolutely incomprehensible. Modern codecs save space by basically "smearing" pixels into areas of similar color. On mushrooms, I was only able to perceive this smearing, and so a scene of a character walking down a hall just looked like an amorphous blob floating across a pool of paint and leaving a wake of color turbulence behind it. I could only tell what was going on by really concentrating and inferring from context. I can't remember if pausing made any difference.
I doubt it, but I wonder if it would be possible to perceive a merely audibly compressed audio stream as being completely garbled.
Don't blame the video codec. You can observe similar artifacts on non video stimuli. Psychedelics reveal how the brain works, and the brain probably compresses information similar to a video codec.
I completely agree that psychedelics make the functioning of the brain (and I would argue the whole nervous system) more obvious. Given that they cause subnetworks that don't normally communicate, to communicate, and vice versa, I interpreted it as two steps in the visual pipeline talking directly to each other when they would normally have to go through a separate network that processes motion. Or perhaps signals from an unrelated network filtering into the pipeline and messing with the interpretation of the visual input.
Obviously video codecs should be optimized for sober psychovisuals. It would be wasteful and impractical to do otherwise.
I think discrete pipelines is the wrong paradigm. I surmise what is happening is the incoming information is being changed to the frequency domain for much the same reason that many visual and audio filters are implemented in the frequency domain. It's just a convenient basis for signal processing. It also so happens that compression codecs work largely by also converting to frequency space and then throwing out all the terms with negligible amplitude. So obviously when you poke at the part of the brain that does visual signal processing, you get something that looks a lot like a Photoshop filter or a compression artifact.
Unlike hearing, I don't think the visual system works in the frequency domain. Common psychedelic visual hallucinations are better explained as either transformations in the space domain, or stages of a pipeline being reconnected in odd ways. As an example of the former, a sharpen filter can be implemented as a simple convolution, which would be compatible with crosstalk in the neurons processing the "pixels" of the visual input. Of the latter, fractals are a common theme in hallucinations; they can be explained as feedback loops in a pipeline, where the signal ends up making several loops through a few processing stages. The sharpen hallucination might also be caused like this. Perhaps the brain already applies a sharpen filter and what happens is that it's applied multiple times to the signal.
I think you might have missed out on learning about the 2d Fourier transform and the meaning of "frequency domain". The standing wave patterns of the feedback loops are exactly the wave patterns you want to decompose the image data into. The waves have a characteristic frequency inversely proportional to the length of the loop. One can apply a decomposition into a wave basis just as easily on the space domain as one does on the time domain. The convolution in the original domain just becomes multiplication in the frequency domain. Sharpen and blur filters are literally just high pass and low pass filters applied to the 2d Fourier transform.
I highly doubt there is one big pipeline to which the original sensory inputs go through some fixed filters and then return back to the main start point as feedback. It's not exactly like pointing your camera at the projector, or holding the microphone too close to the speaker. Rather, there's many possible paths for neural inputs to loop back, and none of these loops is the dominant "pipeline". It's more like a microphone is near a whole wall of speakers, each with its own characteristic delay and amplification factor. The repeating spatial patterns of whatever length directly translate to a standing wave in neural loops with the right amount of loop length.
The transform to a convenient wave basis doesn't need to be implemented as a layer in a pipeline because the loops themselves are the transform.
A transform into frequencies (in some weird basis) is almost an unavoidable consequence of connecting neurons at random. Just like any old metal sheet is a bell if you hit it with a hammer, any old clump of neurons is going to have standing waves of activations that resonate with some sensory input pattern.
If you randomly poke at some neurons, possibly altering their delay time or adding/disabling connections, statistically the longer the loop the more likely it is to be affected. This view also nicely explains the other effects of LSD. Your sense of time is wrong for the exact same reason your perception is distorted: the timing of all the loops have been slightly detuned.
What strikes me as unconvincing about interpreting vision as operating on the frequency domain is that anyone can look as a sound wave's spectrogram and understand what it's representing, even if they're totally unfamiliar with graphs in general. I've looked at spectrograms and noticed sounds that I was missing, then looked for them at the right times and frequencies and found them. You may not be able to understand the sound from a spectrogram (e.g. understand words, or melody), but you can understand what it's representing.
If you look at the 2D FFT of a bitmap, the result is completely incomprehensible. You could look at two videos and their corresponding 2D spectrograms side-by-side and not know which corresponds to which.
Apparently there have been no documented cases of schizophrenia found in people who've been blind since birth. It would be interesting if someone explored the relationships between the effects of psychedelics and those with another kind of blindness, Aphantasia and those without.
I share your doubt, but one thing I dare say is that they make you more likely to spot things in your peripheral vision that would otherwise go unnoticed. Things stand out more, small differences are enhanced.
Also looking at a paper I could see the fibers it was made up of, not sure if that was real though.
They do apply a subjective sharpening (literally like an image processing sharpen filter) to the entire field of vision, and they also enhance pattern recognition, and so details that might be more difficult to see stand out. But on the other hand, they also make it more difficult to concentrate, and they also make you see patterns that don't actually exist. So it's difficult for me to imagine anyone using them for immediately practical purposes successfully.
I think it operates more on the attention mechanism, as well as exploring different (and often unbeknownst) search space in the processing (the long-term effect of which can be biological in ways we don't understand).
And if taken regularly and had a certain training for a prolong peroid of time, they should be able to "increase the quality" of perception the same way when you play tennis regularly and with enough time you will develop the motor skills and can finally "feel" and effortlessly hit the ball at the center of the racket while swinging your hand in a way the racket travels straight.
Or the same way one becomes less tone-deaf or more sensitive to sounds, etc, (of course, it may go the other direction too - but I think the computational cost is quite different (acquiring vs losing) - still, it's definitely possible, eg, for one to 'unlearn' their mother tongue, etc).
Do note that throwing a basketball is not a purely, or even mainly, perceptual task. But I'm amazed you were able to do that. My coordination gets shot to shit when I'm on psychs.
