Prelude to a DiPT experiment (you don't have to be a DiPT user to help)

I'm planning an informal (not approved by any academic body; being rather just a product of my personal curiosity) series of experiments on the auditory effects of DiPT. To this end, I have written a program to test people's pitch perception...

http://diptology.pbwiki.com/f/DiPTtest0_1.zip

This is a program designed to test people's pitch perception, and attempt to train them, to see if their pitch perception can be improved. It will take about 5 to 10 minutes (or more, depending on how you do), but you'd be welcome to quit the program when you get bored, and send me the incomplete data.

This program is not to be run while under the influence of DiPT... It is just a pilot study to see if people can do the task okay normally: I need to know whether they can before I distribute the first proper DiPT experiment. So you don't need to be a DiPT-user to help out with this: you just need to have vaguely normal hearing and a desire to help further our understanding of this remarkable psychedelic, and/or of the parts of our auditory system that it affects.

You also don't need to have perfect pitch. I don't have perfect pitch, but I can do the task (not perfectly, but well enough). It might help to have some experience of playing music, but it may not be necessary. Anyone with normal hearing can attempt this task.

If anyone's willing to help with this, please download the zip file, extract it, then run the program within it, then send the output file to diptology (at) safe-mail.net. (Out of interest, let me know when you send the output - if you don't mind - what sort of experience of music you have, e.g. listening, playing, studying, etc.)

Please let me know if there are any problems in running this program, or any bugs. In anticipation, many thanks.

(The results of this pilot study, and of future experiments, will be documented at diptology.pbwiki.com, but individual participants' data will not be associated with their identity, unless they wish to disclose their data.)

 

Sweet. I'll give it a try.

I need to either fix the power supply for my speakers or find me a pair of headphones first, though.

 

Brilliant, thank you so much.

Ah, yes, that would probably help.

By the way, if anyone's reading this who doesn't read bluelight, here's a trip report including (at the end of the post) a brief report of the results of my testing of a version of this program during a DiPT experience.

 

Ok, finished the test and emailed you the results. I'm pretty much musically retarded, so my results should show that! Lol.

Hopefully more people will give it a try!

 

Thank you very much, pedaltopedal! I've replied to give you feedback on your results: I'll not document them here, to preserve the confidentiality of your data, unless you want me to.

Indeed, I hope so. A few people over at bluelight have expressed enthusiasm for the idea of this series of experiments, and there's a reasonable handful of DiPT users there (enough for a typical psychophysical experiment anyway; we're generally quite happy with five or six people; although given the moderate difficulty of this task, a few more than that might be needed). So I expect this should be able to fly; or at least glide from tree to tree.

And, as I've said elsewhere, if the worst comes to the worst, I always have one participant I can rely on to listen to beeps while tripping on DiPT (myself), and single-participant experiments are, rightly, quite acceptable in psychophysics, because perceptual phenomena are, as a rule, not terribly influenced by cognitive biases, and tend to be relatively stable across different people. So, whatever happens, we should have some worthwhile results in the end.

 

Cant wait to try this, plan to hear from me tonight when I get off work...ugh work...

 

It's not the most fun experimental task in the world, unless you're very into sine waves as an artform, (later experiments will, I suspect, be more interesting, as I'll probably introduce more complex sounds than beeps), but bless you for your enthusiasm. May your work pass swiftly.

By the way, it has occurred to me that I may be taking the wrong approach in thinking that I need to ensure that participants can produce errors that are mostly clustered around perfect accuracy, with respect to circular pitch-class-space (a circle of semitones, that is: C, C-sharp, D, D-sharp etc).

Given that pitch space is at least two dimensional, perhaps three dimensional, with, for instance pitch-classes that are proximal on the circle of fifths (e.g. C, G, D, A etc) or have shared harmonics seeming similar, as well as pitch-classes that are close, semitone-wise; I'm losing relevant information (treating data points as outliers to be removed) by trying to force a simple circular pitch-space on the data.

So, unless people are very precise in their pitch-perception, there will inevitably be multiple peaks in their error distributions: One at perfect accuracy, one smaller one at a fourth and a fifth above it, etc. The error distribution in semitone-space should be a sum of (circular) normal distributions, with the locations and sizes of the peaks being perhaps unique to each person.

So, as long as people aren't completely lacking in pitch perception, it may be possible to model their pitch-class-space at baseline and then, during the DiPT (or control) experiment, slide that model around in semitone-space until a best fit is found.

I'm having to learn quite a lot of new stuff about pitch perception and about statistics conducted in variable-spaces with unusual topologies. Much more of a learning curve, designing these DiPT experiments, than I anticipated. Which is cool... it's interesting stuff, for me anyway.

 

Did you get rid of the zip file? Cant find it...

 

Woops, sorry! I made a slight tweak to the program, and screwed up the link after reuploading it. This link is now correct: http://diptology.pbwiki.com/f/DiPTtest0_1.zip

 

Sweet thank you, I got bummed there for a minute.

 

I've had messages elsewhere from people suggesting that the instructions in the program aren't entirely clear... Here's some more detailed instructions; please do ask in the thread or by e-mail if you want to take part, but don't understand the instructions...


The experiment is a series of 'trials'. You start each trial by clicking on the 'Continue' button.

This is what happens in each trial:

(1) You hear a series of random beeps. These are just to try to make you forget any previous note you heard (so that each trial is properly separate). You don't need to make any judgment about the random beeps. They have nothing to do with the task.

(2) You then hear a long continuous tone of one pitch. The words 'Give me a B flat' or suchlike appear, at the same time.

(3) Your task is to move the slider around, until the note you can hear is the note you were asked for. So if it said 'Give me a B flat', you move the slider around until you think you can hear a B flat.

(4) When you think you've chosen the right note (or you decide you've got no idea), click 'Confirm Note'.

So that's what the trials are like in the first 12 trials and the last 12 trials of the program.

In the (longer) middle section of the program, there are two extra things that happen, to give you feedback and help you learn, at the end of each trial...

(5) If you got the note right, the program will say 'Correct!'

(6) If you got the note wrong, the program will say 'No, that was a G' (for example), and then it will play the correct note ('B flat' in this example), while saying 'This is a B flat')

 

I took this quote that you posted on bluelight (hope thats ok) and decided to put it up here because I think it could be quite useful for some people. This is essentially what I was trying to during the training phase. I have some limited musical experience, but nowhere near enough to be any good at being able to determine a note just by hearing it. I was trying my best to learn during the training phase, by associated a sensation, feeling, memory or just some thought with each note. It helped, but in reality, I would need far more training and lots and lots of practice.

 

That's very interesting to know! I suggested it just because I do, without trying, have sort of textural or tactile or maybe almost coloured or gustatory associations with different notes. Not to the extent of synaesthesia, by any means; it is perhaps just my mind trying to find words for describing an aspect of hearing that we lack special words for in our language. The differences of pitch quality just don't have such words, so one has to borrow from other senses.

ETA: Also, I'm aware from the psychological literature on memory that this sort of multi-sensory elaboration can help with learning.

Since you mention your performance here, I guess it's okay for me to comment briefly on it as well? [If not, I'll edit this out.] I'm working on various ways of analyzing less precise data, like yours (mainly by taking into account pitch-similarity on the circle-of-fifths); and most of these ways suggest that you mostly either perceive the pitch correctly, or you perceive it as just off from the correct pitch, either by a semitone (on the chromatic scale) in either direction, or by a fifth (adjacent on the circle of fifths) in either direction. Other errors occurred, of course, but they were clustered around the correct pitch and - usually to a lesser extent - other similar sounding pitches. I reckon this would be quite sufficient to model the pitch effects of DiPT. It's conceivable it was a fluke, but unlikely.

So, I don't think you need much more training, basically. (Of course, the more precise your performance, the easier it'll be to analyse, and the clearer the effects will be; but imprecise performance can be compensated for by producing more data (either within one trip, or across several trips), and - with strong enough effects, as I think DiPT's effects are - a fair amount of imprecision can be coped with, anyway.)

 

Yes. Exactly.

It was almost second nature for me to try and make these associations, I just couldn't really see any other way about going about it (for me, anyways). It would be interesting for me to hear how somebody who can identify each pitch went about memorizing them. I wouldn't be surprised if they used a similar approach in the beginning. I'm sure with enough practice it would become second nature.

Oh yes, of course. No problem. You are free to show and discuss my results as you please.

 

I think it may often, for people who are relatively good at identifying pitches, be a product of sufficient exposure to the sounds in the presence of their names, without necessarily any conscious learning.

Different pitches tend to stimulate different neurons: there are neurons semi-specialized for each pitch, iirc, so there's the capacity in everyone to be able to identify them. I suspect that most people would find that different pitches sound different, just in an indescribable way.

Certainly, my (moderately precise) conscious pitch perception wasn't deliberately learnt. I became aware that I could tell what different piano notes sounded like in my teens (after at least a decade of piano playing, since I started at 3), iirc; but I thought it was specific to the piano (perhaps the way the strings are made), and never tested it rigorously. It was only on taking DiPT that I realized my pitch perception was absolute, and not due to physical differences between the different strings of the piano; because now the strings that would normally produce the sound of a C were producing the sound of a B, and then a B flat, and then an A, all sounding just like the actual B, B flat, and A notes normally sounded like.

Perhaps a relevant analogy is wine-tasting. Anyone can tell, if presented with two wines in quick succession if they were the same or different wines, because they will taste different if they are different. But (1) They wouldn't be able to do this if the two wines were only slightly different; and (2) They wouldn't necessarily find it easy to put the tastes or the difference between them into words, or to - on separate occasions - name the wine. An experienced wine-taster (or just an experienced drinker of a wide variety of wines) would have both more precise perception, and more capacity to put words to those percepts. (This wine-tasting analogy isn't mine, by the way, but that of the perceptual learning theorists of the earlier part of the twentieth century. Current thought has neurons retuning themselves so they become more precisely exclusive in which pitches will activate them; as an explanation of this sort of perceptual learning. )

Great, thanks!

 

where can we get the DiPT?

 

That makes a lot sense, invert. You're probably right. During the first phase of the program, I was trying to remember back to when I used to play the alto sax many years ago. I never consciously tried to learn each note back then, but while taking the test there were a few times where I was like "Yea, that's the way I remember this note feeling". Of course, it's been almost a decade since then...

That's a great analogy, btw. I might have to look into this perceptual learning, it peaked my interest a bit.

 

Ah! Yes, I guess experience playing is likely to be a key factor in the extent to which people are already (pre-training) able to do the task. It's the equivalent of staring at a wine label every time you sip the wine.

It is a fascinating field with many intriguing branches, imo. Both the discrimination-type perceptual learning (of the wine-tasting or pitch-tasting sort), and the sort of perceptual learning that involves altering (not merely enhancing the detail of) one's perception to compensate for weirdness (such as wearing prism-goggles that shift one's visual perception off to the right: after a while, things look normal again, then - if you take the goggles off - everything looks like it's over to the left; or such as staring at purple vertical lines and horizontal green lines for a while, and then finding that your perception of the colour of vertical and horizontal lines generally has shifted accordingly).

By the way, I've been reflecting on my personal associations with different pitch-classes, and they do seem to mostly go along the circle of fifths, rather than the chromatic scale (even though a lot of my errors in the task are semitone errors);

thus C, G, and D are - to varying degrees, G more so than the other two - quite definite and heroic sounding, but with C having an added clarity to it, and D perhaps some added warmth;

A, and E, and B are bright, but with increasing depth and fuzziness as you approach B (A is rather cool and watery; B's quite warm and vague, but full);

F sharp, and C sharp are rather misty, but - while F sharp has great warmth, C sharp's rather cooler, almost like a thick cold smoothie;

with G sharp and D sharp, a degree of other-worldliness is retained, but not mistily, and not with the fullness of B, F sharp and C sharp, and D sharp has some harshness to it;

B flat and F are getting quite dry and harsh sounding, with F sounding both drier and clearer than B flat;

which brings me back to the clear, definite sound of C.


So... it's mostly like the further I am away from C on the circle of fifths, the more warm and fuzzy and full notes sound, and the closer I am to C, the clearer, cooler and drier they sound.