Chroma Buckets

Thoughts and responses regarding the research at acousticlearning.com.
TS
Posts: 168
Joined: Sun May 07, 2006 4:58 am

Re: Initial Observations

Postby TS » Sat Nov 24, 2012 3:03 pm

zacxpacx wrote:TS how did you set up the random octaves? Both with the chroma buckets and the target tone.


The range is 4 octaves, which means the program selects a random number between 3 and 6 and passes that as the octave of the note to the JFugue library which in turn passes it to the underlying Java MIDI-synthesiser.

The program selects a random octave for the note, and then the timbre by selecting a random soundbank between 0 and 95. The octave and timbre for the target tone are generated once when you click "Next" and then the play button always repeats the same settings, whereas the chroma bucket buttons generate new random octave and random timbre on every click.

zacxpacx wrote:One suggested change to TS: Right now the next tone button changes the tone, but I have to click play tone to hear it. Could you set it up so the next tone button also plays the new tone once? That would help a lot if I want to practice with speed.


Done. You can use the same download link in the previous post to get the new version.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Sat Nov 24, 2012 4:01 pm

You should try the game TS. It's hard for me to say how effective it will actually be. After playing the game for awhile, it's cool to be able to walk around and pick out Cs, Es, and Abs from the environment. I don't know if the wearing off of this effect has to do with a memory problem or there being too many sounds that aren't Cs, Es, or Abs. Time and practice will tell I guess.

When a random octave is selected (from 3-6), is the bottom note automatically a C? Will the C be the bottom note of any random octave?

I was thinking to combat the slight height perception I'm having I could randomize octaves, changing the bottom notes between C, E, and Ab every once in awhile. If that's not possible, expanding the number of octaves to 5 or 6 might also help.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Sat Nov 24, 2012 11:35 pm

Also TS, how hard would it be to create the D, F#, Bb version? I'd like to get started on that.

I have a lot of theories right now -- I feel bad posting about them though because I don't want to become a David Lucas Burge. So as a fair warning, this next statement is based purely on my experience with chroma buckets and has not been verified with any other players or scientific studies. It is all 100% speculation.

When I first begin chroma buckets, I have to get the sound of each bucket in my ear before I start getting on a role. Once I'm on a role I make very few mistakes (usually the height-related ones I mentioned in a previous post, with high-Cs and low-Abs). After playing the game for a long time (30-60min), I get off and for a time can hear Cs, Es, and Abs in my environment. Any music I hear is automatically heard in relative pitch, but if I focus I can pick out bucket notes.

Here's the speculation: I only have 3 chroma (not even sure if it's the chroma that's sticking in my mind) that I recognize. As soon as I leave the game, isolated chroma sounds stick out. With music, however, I may hear a C but it immediately moves to a note I don't know and relative pitch kicks in -- making sense of what I'm hearing.

Speculative solution: get to the point where I can play with all 12 tones.

There's no way of confirming anything without more experimenting and practice though. TS, if the D, F#, and Bb version isn't too much trouble, then could I also get the entire 6 whole tone scale? The whole scale will definitely have to be expanded to 5 or 6 octaves or the bottom note of each octave will have to change in maybe two or three versions of the game.

TS
Posts: 168
Joined: Sun May 07, 2006 4:58 am

Postby TS » Sun Nov 25, 2012 2:38 am

zacxpacx wrote:Also TS, how hard would it be to create the D, F#, Bb version? I'd like to get started on that.


I made a new version that has 4 different scales.
I also expanded the octave range for the random buttons, so now it's between 2-7.
You can find the new version from the same download link.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Sun Nov 25, 2012 11:41 am

TS I downloaded the game from the download link but it looks the same as the first C, E, Ab. Is there some toggle function I'm overlooking or is it really just the same game?

Also, it would help to have the black key spellings be: C#, Eb, F#, Ab, Bb. Those are the most common musical spellings.

The progression I had in mind for chroma buckets:
1. C, E, Ab
2. D, F#, Bb
3. C, D, E, F#, Ab, Bb (whole tone)

4. C#, F, A
5. Eb, G, B
6. C#, Eb, F, G, A, B (whole tone)

7. C, C#, D, Eb, E, F, F#, G, Ab, A, Bb, B (chromatic)
8. C, D, E, F, G, A, B (C maj/A min)
9. G, A, B, C, D, E, F# (G maj/E min)
10. F, G, A, Bb, C, D, E (F maj/D min)

I view it as having 3 distinct sections. The first 2 sections build up to whole tone scales and the 3rd section begins with the chromatic and then explores common key signatures. Right now I'm only focused on the first section -- building up to the first whole tone scale.

TS
Posts: 168
Joined: Sun May 07, 2006 4:58 am

Postby TS » Sun Nov 25, 2012 2:09 pm

zacxpacx wrote:TS I downloaded the game from the download link but it looks the same as the first C, E, Ab. Is there some toggle function I'm overlooking or is it really just the same game?


You should see tabs on the top part of the window that you can use to switch the view.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Sun Nov 25, 2012 3:14 pm

Hmm... I don't see any tabs. It looks exactly the same as the original game.

Also, I'm trying to copy your source code to Eclipse. All I've done so far is create two classes called Main and Panel in a folder (chroma buckets) and copied the Main.java code to my Main class and the Panel.java code to my Panel class, but it doesn't work when I try to run Main.java

I expect there's something I have to do with the lib folder in your source code link?

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Sun Nov 25, 2012 4:19 pm

DAMNIT!

Today I had a really terrible practice session. I performed better than any previous practice session by far, and I was able to go very fast clicking next tone repeatedly. What sucks is that my mind has developed a strategy that is definitely not perfect pitch.

As I was warming up with the chroma bucket game, my mind slowly formed a triad of C, E, and Ab in that order. My mind internalized the individual notes of the triad with C on the bottom, E above, and Ab on top. As I played new tones, the triad would "jump" so the corresponding note matched up.

Example: if a high C was played, the triad would jump up so the bottom note matched the target tone, and I'd know it was a C almost instantaneously.

The issue with this, that first and foremost, is it isn't a perfect pitch judgement. The skill taught isn't what we're after. Also, every time an Ab is played I have a feeling of the E and C below it. And every time a C is played I can feel the E and Ab above it. Same goes for the E.

Once I realized this was happening I got upset and started playing notes faster and faster. I could click the next tone button at a surprisingly fast rate and still recognize the notes. Another way to explain what I experienced: my mind had developed an internal triad and each note I played landed in one of the "buckets" of the triad. The categories weren't chroma, however. The internal triad had height. There was a bottom, middle, and top note. Each new target tone would fall into one of those categories.

I'll have to get back to thinking about how to solve this later. I'm pretty frustrated right now...
Last edited by zacxpacx on Mon Nov 26, 2012 8:50 pm, edited 1 time in total.

TS
Posts: 168
Joined: Sun May 07, 2006 4:58 am

Postby TS » Mon Nov 26, 2012 12:47 am

zacxpacx wrote:Also, I'm trying to copy your source code to Eclipse. All I've done so far is create two classes called Main and Panel in a folder (chroma buckets) and copied the Main.java code to my Main class and the Panel.java code to my Panel class, but it doesn't work when I try to run Main.java

I expect there's something I have to do with the lib folder in your source code link?


The easiest way to get the project in Eclipse is that you go to the downloads page in the github project ( https://github.com/tsiivola/chroma_buckets/downloads ) and click the "Download as zip", then extract the zip file and import the contents to Eclipse (create a new project from existing files, or something like that).

Then in Eclipse you need to right-click the JFugue file under lib and select Buildpath->add to build path, and after that you should be able to run it.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Mon Nov 26, 2012 7:57 pm

Right now running receives an error "does not contain main type".

Also, the import EventQueue has an error next to it reading: The declared package "" does not match the expected package "chroma_buckets-master.src"

Same with the import Dimension in Panel.java

:/

All I really want to do is change a couple things.
Create a D, F#, Bb version.
Change the whole-tone names to C, D, E, F#, Ab, Bb

On a different note: I've decided to continue with the ear training and see what happens. Though what I'm experiencing now is not absolute pitch judgement, I don't really know what it is. Continuing my training I want to see what happens when the octave range extends and more notes are added. Today I just did some more C, E, Ab training. By consciously paying attention to what I think are the chroma, I prevented delayed the formation of the triad ladder described in my DAMNIT! post.

The way I "focus" on chroma. I play a couple rounds of APA before I begin with the chroma buckets. APA helps orient my perception toward chroma sounds. I hesitate to put my chroma experience into words because suggestion can be powerful and chroma should be experienced directly via perceptual learning not conscious effort, but for the sake of this entry I think of chroma as how the pitch of a sound "hits" my ear. All Cs "hit" my ear in the same way. Higher Cs hit my ear with a smaller "fist" than lower Cs. And volume determines how much force is in each "hit".

When I shift over to chroma buckets from APA, I focus on how each of the pitches "hits" my ear and try to remember the feeling as the game progresses. This strategy helped delay the construction of the mental triad ladder.

TS
Posts: 168
Joined: Sun May 07, 2006 4:58 am

Postby TS » Tue Nov 27, 2012 1:25 am

zacxpacx wrote:Right now running receives an error "does not contain main type".

Also, the import EventQueue has an error next to it reading: The declared package "" does not match the expected package "chroma_buckets-master.src"

Same with the import Dimension in Panel.java


I tried to import it to Eclipse myself and it seems that it's a little more complicated than I thought, but fortunately not very complicated.

I think all you need to do is right click the main project folder and select Build path->Remove from Build Path, and then right click the "src" folder and select Build path->Use as source folder.

Nikolaus
Posts: 53
Joined: Mon Nov 10, 2008 4:39 pm
Location: Dallas

Postby Nikolaus » Tue Nov 27, 2012 5:26 am

sounds like you're developing a feeling for the scale-degrees. the solution is to just play a 1 4 5 1 cadence in a random key (a key you don't know) before testing yourself on each tone.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Tue Nov 27, 2012 9:19 pm

Nikolaus wrote:sounds like you're developing a feeling for the scale-degrees. the solution is to just play a 1 4 5 1 cadence in a random key (a key you don't know) before testing yourself on each tone.


Today I created the D, F#, Bb version of the game and started playing with it. I picked it up much more quickly than I did the C, E, Ab version so I'm definitely improving at whatever skill I'm developing. To start integrating the two, I played each set up three for 2-3 minutes before switching the play the other, going back and forth. Right now it seems like my mind has to re-orient itself before I start doing well. It takes about 2-4 notes and then I get back on track. If I click all the buckets before beginning to hear their sounds then I can do well from the beginning.

I'm going to try to make it all the way to the chromatic scale before I consider using the 1 4 5 1 cadence Nik. I want to figure out what the skill I'm developing really is. Once I become proficient at the chromatic scale, I'm going to try recognizing environmental tones and notes in slow pieces.

If I dead end there then I'll start using the 1 4 5 1 cadence while working with the chromatic scale. I don't know how hard it will be to move from the two groupings of 3 I have to the whole tone scale, but that's bound to take at least a couple days.

Seem like a fair plan? I don't know if working all the way up to the chromatic scale will somehow effect my using a 1 4 5 1 cadence to mix up my scale-degree feelings. It shouldn't, should it?

Nikolaus
Posts: 53
Joined: Mon Nov 10, 2008 4:39 pm
Location: Dallas

Postby Nikolaus » Wed Nov 28, 2012 7:31 am

establishing random key centers will bring to light your inability to differentiate between two tones that are placed a semitone apart. one can easily enough differentiate between tones that are a whole tone apart, but not a semitone. if you manage to get 100 percent accuracy (PP or RP) with the chromatic scale while also establishing random key centers in between each test question (and in random timbres and octaves to boot), then i will pat you on the back and ask you how you did it, and you will probably tell me that it was just lots and lots of practice, and lots of mistakes. of course, one could also cheat by using MTs, which is why i'd recommend what you also have the user designate what scale-degree feeling each pitch is as well (in addition to its absolute placement). as it stands now, i can only do the above drill in one timbre and the two center octaves of my instrument. and what has this led to? i can frequently tell you what key i am in, because as much as the tones individually stick out from one another, the tones in combination stick out from each other even more. but this is only in one timbre, and in other timbres i am at a loss.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Wed Nov 28, 2012 4:08 pm

Does the random note played after the 1 4 5 1 cadence have to be a completely random tone (from the chromatic scale)?

If it does, then I have yet to work up the chromatic scale in chroma buckets. If the 1 4 5 1 cadence can be used for any collection of pitches, then I could go ahead and give it a shot now. It would just be a problem of implementation. Can anyone program, or does anyone know of a program, a random 1 4 5 1 cadence player -- random in key, octave, and timbre?

Such a program is beyond me, and I can't play a 1 4 5 1 cadence myself on my keyboard. I'm not a pianist so I'd have to search out each note of the chords. Also, I'd know which key the cadence was in.


Return to “Your thoughts about music”

Who is online

Users browsing this forum: No registered users and 3 guests