Observation and musing

Comments and questions about Chordhopper.
Post Reply
Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Observation and musing

Post by Sleeper » Mon Dec 08, 2008 4:30 pm

So, I was playing Chordhopper the other day, and when the F chord came up, I thought, "What the? I've heard that before. That's 'Goodbye Yellow Brick Road'." (I probably hadn't listened to that song in at least weeks)

I looked it up, and, yeah, the first and last chord of the introduction is F.



This leads me to some musing. I think the thing about learning absolute pitch as a child is, you then have that reinforced throughout your entire life. You hear F's or G's or C's constantly, and always know what they are. On the other hand, my brain may have somehow stored that some song starts with an F major chord, but most of the time when I hear songs I don't know what I'm hearing, and I had no way of connecting that F major chord to anything else.

Or, to put it another way, if there is something to the melody trigger method, then every single song with a note to someone with AP could have could be a melody trigger for that note.

Now to make an analogy to color. What if we had eyes that were perfectly capable of seeing in color, but somehow we lived in a hypothetical society where everything was black and white. You see everything in black and white for decades. Finally, someone shows you some blue and red cards and tests you with them. First, you immediately think, "Whoa, that's different!" You can also immediately tell red from blue.

Then the test is over and you go back out to the world, and everything is back to black and white, and the memories of red and blue fade. Then one day, in the street, you see a red fire hydrant. You immediately think back to the test and you remember that what you're seeing is a color, and you think it might be red or blue, but you can't really remember which it was. And maybe it's a little bit darker or lighter red than what you saw, and you're not even sure if it was one of the colors you saw during the test. And if you saw a full color painting, you'd probably be totally baffled.

In real life, though, everything we see is in color, and our brains automatically categorize the colors of everything we see.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Wed Dec 10, 2008 11:56 pm

I'm only up to 9 chords, but here are my specific observations so far:

I was surprised that the "theory" behind the game is that CEG and BDG is supposed to "bring out the G". That might have been the effect at first, but that only helps distinguish between the flower and the others. What I noticed is the B jumps out at me for the blue swirl.

When the basketball unlocked, I started listening based on "height" -- the basketball was the lowest, and the swirl was low, too. The only problem was that if other chords played, and then it went back to those two, I only knew they were low -- I couldn't judge how low, exactly. So, that strategy eventually failed.

By now I hear the basketball absolutely -- it either sounds like "Goodbye Yellowbrick Road", like I said, or the A jumps out and it sounds like the three notes played individually in the mid-game. I also hear the swirl absolutely based mostly on the B, sometimes the G. When it's B, it might remind me of the C melody word from APA -- "C, E, G, B, C, D, C" -- except starting in the middle, on the B.

But sometimes individual notes jump out at me, and sometimes they don't. Even though I hear the B and G in the swirl, I'm not so sure I hear the D. This is giving me some trouble with the fish -- the highest note for that chord seems the "loudest", and I can't always identify it.

Also, I'm not sure any individual notes are jumping out at me for the C major chords. I identify the C majors relatively. Since the whole game is sort of like a chord progression, when it sounds like "resolution", I pick C major. I might pick the flower chord based on relative clues sometimes, too (or I'll think, "It doesn't sound like resolution -- but it doesn't sound like G, either"), though I might be associating it more with the basketball, which might give me better luck at viewing it absolutely eventually.


On a side note, I'm not sure how we're supposed to be playing these games, but I've been focusing more on Chordhopper than APA. APA seems pretty difficult to me. It's probably because individual notes only sometimes pop out at me when played simultaneously. Sometimes I'll hear the C and just KNOW. And then I'm always right. But sometimes it doesn't pop out at me, and when that happens, the C might be there, or it might not.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Re: Observation and musing

Post by Sleeper » Thu Jan 08, 2009 11:31 pm

I guess this should have gone in the observation thread. Ah, well.
Sleeper wrote:So, I was playing Chordhopper the other day, and when the F chord came up, I thought, "What the? I've heard that before. That's 'Goodbye Yellow Brick Road'." (I probably hadn't listened to that song in at least weeks)

I looked it up, and, yeah, the first and last chord of the introduction is F.
I've had this experience several other times. I heard a D on the guitar and thought, "Wait, I know that note. That's the first note of ... uh, some Soundgarden song." I thought the melody, but couldn't remember the name. I went through some titles, and remembered that it was "The Day I Tried to Live." I looked it up, and it was correct.

The thing about that example was that I hadn't heard that song in at least a year, probably. Yet it still stuck with me.

Another weird moment was in APA, which I heard a random sound that did NOT contain the target, but it reminded me of another song. I checked it, and that was right, too (sort of -- the song had a chord, and part of it was in the APA sound)

Just about every time this happens, when I can check it, I turn out to be correct.

So, I was thinking absolute pitch had something to do with the ability to memorize tones. I think I must have some capacity for long term memory of tones, though it is not now organized in any coherent way (ie., I remember one tone as "The Day I Tried to Live", rather than remembering all D's as D's.) And the recall of that memory is random, not voluntary.

But just like mentioned on the site, being able to name tones is not absolute pitch. I've tried thinking about notes in songs as I hear them, and, nope, no success. Part of that may be that a song could be in any random key, and I'm not familiar with all 12 notes (I think if I had to play a note naming game, I'd be okay with C, E, G, maybe all the 'white keys', but I don't have many associations with the black keys yet). But SOME of the notes must be some I probably could name. It does seem to be a different skill, to name notes in isolation, and to name them in any context.

(As for non-musical contexts, I've heard people identify the note of an electric toothbrush. I was skeptical of that, because it seems like my electric toothbrush doesn't play a note -- it hums all over the place changing pitch when I brush. But once I turned it on when it wasn't in my mouth -- and it sounded like a C. I checked it with a chromatic tuner -- and the reading went back and forth between a C and an F. So, maybe partial credit on that one.)


APA still seems difficult, but I can try to describe what I'm doing with it now. First, when I play, I do start from the first level and see how far I can go. One reason I do that is because I noticed that eggs are much more likely to contain the target than not, and if the chick stops part way in the road, I might be able to get it fully across just by marking the other eggs as having the target -- regardless of whether I can hear the target there or not. This trick let me get farther than I think I should be able to (especially since you only need to get 20 or so chickens across total, no matter how many games it takes, to advance), so I reset.

The difficulty curve seems sort of steep. With level 2, you'd often get one egg that had the target by itself, and some others that would have it in harmony with other notes. It's easy to identify the one that has the target by itself (that's just note naming, not absolute pitch). Then I'd go back and forth between that egg and the others, seeing if they sounded like the target was matching up with something in there.

For later levels, I try the same thing, except that I probably won't get the target by itself. But my strategy is to try to remember the target, and play it back in my mind. So, I do the same comparison, target and unknown egg, and see if it seemed like the target I was remembering was harmonizing with anything in the egg.

I think that's the basis of the "C-ness" of a note anyway. I think if I had to give it a technical definition, it'd be something like, "A note has [an idealized] C-ness if it harmonizes like an octave with a C [any C, even real, even with a specific timbre, duration, and amplitude]."

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Tue Apr 07, 2009 3:23 pm

So, there seems to be a question of, What good does this training do if it doesn't result in full AP?

I think one of the first benefits I've noticed is that it results in improved note bends (say, on a guitar).

Thinking in terms of "bending up a full step" requires a height estimation, which might be inaccurate. And while relative-thinking like "going up a full step" might normally work while playing or improvising, that's partially because the frets are there to make estimates exact. Not true while bending.

I think, say, bending an A until it becomes a B is much easier, and more accurate.

Why does it work without full AP?

You're in control of the situation. You know your instrument's timbre. You know the note you're trying to hit (you're not guessing from any of 12 possibilities). And you're probably not needing to identify the note in weird harmonic contexts.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Mon Apr 27, 2009 9:08 am

Up to the first Bdim in Chordhopper.

Sometimes the superfly rounds can feel kind of stressful (since you can't make a mistake). However, I honestly feel as though I am not using relative pitch when I play this game. In fact, as a consequence of it being stressful, I sometimes pay so much attention to each chord, individually, that I don't compare it to the previous chord at all. This is especially weird when the same chord plays twice in a row. Rather than immediately thinking, "Oh, that's just the same chord", I'll think, "Hmm, G7, high D ... oh, wasn't that the last one I did?"

Some notes seem more distinct than others.

Most distinct would be B.

C is coming along. I just started hearing the high C in the root position Fmaj as the same note as in first inversion Cmaj. It kinda sounds the same as the C in the EAC Amin chord. (Previously, I had been identifying that Fmaj on the basis of the entire chord sound)

Worst is probably F. I think the problem is that I had been able to identify the Fmaj chords some other way, and with the G7's I heard the added F as added dissonance. That, however, was NOT good enough, really. Dissonance is sort of subjective, and the more you hear a sound, the more you might get used to it, and then it sounds less dissonant. So, I still sometimes have trouble telling Gmaj's from G7's (maybe an error every couple of games).

I don't have much of a problem with the game taking a long time. (Having to play so many chords to earn a fly, having to earn so many flies to earn a superfly, etc). Considering my rate of development, it makes sense to me to go slowly. I'm hoping that as more chords are added, I'll need to (and will) be able to identify more distinct notes in the chords. (since, for example, Gmaj, G7, and Bdim all have certain overlapping notes)

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Mon Apr 27, 2009 1:59 pm

The way I play Chordhopper is that I only play a few sessions a day, and usually only two games at most in a session. I figured that would most closely mimic the Eguchi method of sticking with a chord for two weeks.

I think what I might try in the future is playing a game of "note naming" before I play Chordhopper. I had been avoiding that in the past. I figured if I first started with Chordhopper, then the first chord I named would be "pure". It would have to be named through at least some (if weak) AP.

But who cares about that. And today I played a game of Chordhopper after a note naming game, and many more notes in chords were jumping out at me. Usually I only get into that mental state rarely, so if I could encourage that, it'd be great.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Tue May 26, 2009 4:36 pm

I was trying the pitch paths method. I had gotten to the point of basically 100% identification of the "white note" tones, across a few octaves. Of course, this was notes played in isolation. Still, it was time to move on to the black note tones.

During this time, I completely stopped using ETC, since in Chordhopper I'd only be tested on white notes. I wanted to keep the two sets distinct, for now.

After about three weeks of using the melody triggers for the black tones, I was basically identifying them at at-chance levels. I guess I got F# every time it came up, but everything else was a jumble. The method worked perfectly for the white tones but failed for the black.

I tried using the "widget" on the front page for alternate melodies. (Every accidental melody is different than the pitch paths melody) It seems to be working. I might be at about 90% for identification of black tones in isolation now.

So, what's the problem? I'm not sure but I can make some guesses. The pitch paths melodies do seem famous and catchy, but maybe they're still not ideal for some reason. The C# and Eb melodies both start by playing the target note, then going down a minor second. That might have caused me to confuse those two notes. Also, the pitch paths melodies are purely melodies, with no harmony, while the "widget" is in full harmony. Assuming the melody method works because adult brains need more context to wrap elemental ideas in, it makes sense that more context is helpful. Also, some of the melodies in pitch paths are almost nothing as just short single note lines. Beethoven's Moonlight Sonata is really famous, but reduced to a three note arpeggio played three times in a row, it's hardly anything.

The other possibility is that the more melody triggers, the better. Chris mentioned sometimes that no melody would trigger when he'd hear a tone. I think the more triggers available, the better the odds of one of them triggering. Oddly, while the pitch path melodies did not trigger when I used them alone, sometimes they trigger now after I've used the other melodies.

That also relates to learning C by synthesizing different experiences of C. F# in "Ode to Joy" might seem happy. In the widget melody, it might seem sad. In reality, it's neither. It's F#.

It's like, what is red? Well, maybe it's hot. Like fire. Or, maybe it's romantic, like Valentines. Or disturbing, like blood. Or pretty, like a sunset. Really, red can trigger any of those associations. I would say that calling the main subjective experience of red as "hot" isn't really correct.

I kind of think that if that widget was expanded to maybe have 5 trigger tunes for each note that it could be a worthwhile component to ETC.

It seems like the melody method is one of the most promising now, because A) it's natural (I've experienced notes triggering songs I haven't heard in months, and later verified that the absolute pitch was correct) B) it works and C) it's amazingly accurate.

For C, even if I was given a starting tone, and then had to name notes in the key of C across a three octave range, using relative pitch it would be extremely difficult to be completely accurate. Sometimes I would have to estimate, I'm sure. There's even a comment somewhere in the archives about most people with RP only judging intervals less than a perfect fifth as a distinct thing that can be judged instinctually; larger intervals are judged intellectually. But with the melody method, it's possible to get them all.


Of course, it's not complete. It does not lead to being able to identifying notes in music or harmony. It does, however, seem to be the best starting point.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Sat May 30, 2009 7:14 pm

I've decided to declare myself having passed the "black tones alone" thing, so I can now go back to white tone stuff, including Chordhopper.

I haven't played that game in about a month. How much has faded? Not much. I got a fly the first time I tried it again.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Wed Jan 27, 2010 8:08 pm

I noticed today that if you search (ctrl f) in Firefox and type in something that's not on the page, the error chime is an "A".

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Tue Feb 02, 2010 8:54 am

I noticed that the beep my car makes when it starts up is a B.

Sleeper
Posts: 69
Joined: Fri Dec 05, 2008 10:30 am

Post by Sleeper » Thu Mar 11, 2010 6:40 pm

I happened to hear "Hungry Like the Wolf" on the radio on the way home today. I did some chores when I got home, and a little later grabbed the guitar and almost immediately played the song. Not just the right melody, but also in the right key. I didn't know the (letter) name of the key, just to hear it, but when I played it it sounded right. I later verified the key was right, but that was just for the sake of objectivity; I had little doubt.

Post Reply