Reporting benchmarks could be interesting

Talk about what you've discovered by using ETC-- and post your high ranks!
Post Reply
Musa
Posts: 51
Joined: Sun Aug 07, 2005 9:06 am
Location: Spain

Reporting benchmarks could be interesting

Post by Musa » Tue Sep 13, 2005 8:41 pm

There are many reports of rank/waves around here, but reporting benchmarks could be also interesting. Don't you think?
My first one with APB is in my "journal".
Cheers.

aruffo
Site Admin
Posts: 1692
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

Post by aruffo » Tue Sep 13, 2005 9:38 pm

With some trepidation, I just completed a benchmark-- not only was I totally cold (no warmup game) but I missed yesterday so I haven't played in two days.

The RED percentage is how many you correctly identified as NOT having C.
The GREEN percentage is how many you correctly identified as HAVING C.
The BLUE percentage is your correct identifications overall.

Initial benchmark, before any training:
RED: 77%
GREEN: 53%
BLUE: 65%

Current benchmark, after 434 waves, currently on Red Colonel:
RED: 90%
GREEN: 57%
BLUE: 73%

So I've gotten considerably better at detecting when there is NOT a C-pitch, but am still at chance levels when it comes to detecting the C.

If I had to offer an explanation, it's probably because in playing the game I've only just started working with multiple octaves, so it's entirely probable that I don't recognize the high and low C pitches when they appear, and the 57% could be not so much "chance levels" but correct identification of the middle C.

That speculation is supported by the 90% for red. If these percentages merely meant that this time I pressed the red button a lot more often, then my score for green would surely have gone down instead of staying the same (because I'd have pressed green less often). But since green stayed at the same %, and because when I was doing the benchmark I know that I was very confident in my not-C choices but not at all certain of off-octave choices, the best explanation is that the 90% red very probably represents improved ability and the 57% very probably represents my lack of familiarity high and low octaves.

I'll have to try again once I've struggled through the different octaves.

miklos
Posts: 41
Joined: Mon Aug 08, 2005 7:46 am
Location: Paris, France !

Post by miklos » Wed Sep 14, 2005 3:24 am

It's been a while since my last benchmark, but I remember I had similar figures : around 90% red and 50% green.

At the time I did that benchmark, I had the same explanation for my green percentage : not enough familiarity with the other octaves of C.

I guess I'll do another benchmark soon and see what comes out.

Musa
Posts: 51
Joined: Sun Aug 07, 2005 9:06 am
Location: Spain

Post by Musa » Wed Sep 14, 2005 6:00 pm

Very interesting. Yes, a good hypothesis. While benchmarking myself, I also felt now and then as if I was missing octaves.

Also, we can have non exclusive hypothesis. After my benchmark of 83% red + 40% green, I played some APB rounds more. Then I repeated the benchmarking but triying to be more prone to click "C". I got 83% and 57%.

Benchmarks of people who are playing with multiple pitches would be welcomed.

Ryan Herr

Post by Ryan Herr » Wed Sep 14, 2005 10:47 pm

I've only played this program a few times, but I agree that it is also harder to hear the C being there than the C not being there.

To give just one example, if the program plays two notes in quick succession, and the second note is a C, it's not usually hard to identify. But, if the program plays two notes in quick succession, and just the first note is a C, and the second note moves away, then musically I don't hear those two notes as a C, I hear it as being whatever note got landed on.

Wow, that was extremely inarticulate of me, but I'm not sure how to better express what I'm saying. Does anyone else follow?

-Ryan.

aruffo
Site Admin
Posts: 1692
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

Post by aruffo » Thu Sep 15, 2005 12:10 am

if the program plays two notes in quick succession, and just the first note is a C, and the second note moves away, then musically I don't hear those two notes as a C, I hear it as being whatever note got landed on.


That's how it's designed to work, all right. That experience will constantly repeat itself as you continue to play. When I began playing APB, I was doing fine recognizing the C as either the first or last tone, but as soon as the C was the medial tone I couldn't pick up on it at all. Now, of course, I welcome those medial tones as child's play, and it's the random groups of intervals that are foxing me. But as I keep going, those too are getting easier, and the octaves are the new challenge.

davemac
Posts: 6
Joined: Thu Aug 18, 2005 8:06 am

benchmarks

Post by davemac » Fri Sep 16, 2005 11:54 am

Benchmarks!

This was done yesterday after playing [currently working on G the highest rank is commodore 4th class]
C- RED 93% G- 93%

GREEN 87% 80%

BLUE 90% 87%

Today having just finished work I tried again on the C pitch with no preparation at all..

C- RED 88% GREEN 77% BLUE 82%

While the above is encouraging I don,t really really understand what it means.I,ve been working on this for a good while now
with some progress, for instance tones seem to have a lot more depth to them than they use to,and most times especially when fresh I can think of an open string on the guitar play it to check, and be right.Sometimes I try to guess what key something is in and this is often proved correct more times perhaps than chance would suggest.On the other hand this is very fragile and if there are lots of key changes I soon get lost.I should add that this was using the old method, common mistakes were often confusing notes with the adjacent one ,for example E with F,why this is I don,t know, perhaps note naming was the problem.? Perhaps what I perceive to be the chroma of a note is not really the chroma at all,the beauty of the new system is that it does not matter anyway, all you have to do is listen, is that right?
For the record I have played interval loader to the highest level but have not played chordfall for sometime as there is a revision due soon I think?
For relative pitch I find pitch coach useful and am now working on ,7th chord inversions open position.What I do is start with only one pitch and then work up with as few errors as possible,you get a fanfare when you get 10 correct answers,then I increase the number ot piches upto the full 12.I think it would be useful for those of us are struggling to hear a tone hidden whithin a chord
Hope this has been of use to someone

Davemac!

aruffo
Site Admin
Posts: 1692
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

Post by aruffo » Fri Sep 16, 2005 3:21 pm

The benchmarks are a measure of how well you detect the presence or absence of a target pitch. They're useful to keep an objective and quantitative record of your progress. The real progress, though, of course, is what you begin to hear and experience outside the game in your regular music making.

Post Reply