Page 1 of 1

Tonal Ear Training

PostPosted: Tue Jun 04, 2013 5:55 am
by Mark Bassett
Hey folks,

I've just released an online version of a technical ear training program as part of my PhD. You can find it at

http://www.tonaleartraining.com

Run it with Google Chrome for the smoothest experience.

It's designed as a basic ear training program with several restrictions built in (like the inability to hear the filtered or bypassed sound again) in order to gather useful data on the effectiveness of training method on tone colour discrimination.

I'm presenting the research behind it at the AES 50th Conference of Audio Education in Nashville in July.

http://www.aes.org/conferences/50/

Cheers

Mark.

Re: Tonal Ear Training

PostPosted: Tue Jun 04, 2013 11:38 pm
by Lee
Congratulations on getting to present your work at the AES. I look forward to checking it out. Are you going to check out any studios while your there?

Re: Tonal Ear Training

PostPosted: Wed Jun 05, 2013 11:12 am
by Text_Edifice
Hi Mark,

looks cool but wouldn't work on FFox or Safari (work machine doesn't have Chrome). FFox stalled on the loading audio and there was no sound with Safari. I might be doing something wrong but just thought I'd report back.

Re: Tonal Ear Training

PostPosted: Wed Jun 05, 2013 10:07 pm
by Mark Bassett
Cheers for the feedback.

Yep it's built for chrome, but it does work well on the latest version of firefox I'm told.

Forget mobile devices and IE.

Re: Tonal Ear Training

PostPosted: Wed Jun 05, 2013 10:57 pm
by audioio
Mark wrote:It does work well on the latest version of firefox I'm told.

Yes, it does.

What are you using? HTML5 markup, CSS and JS? One of those spiffy new SPA frameworks, like Angular? When I View Source, Firefox just says Cannot GET (never seen that before). Still, it looks pretty neat.

EDIT
Oh, I see it now. That's a couple of whacks of crazy complex minified JS there!
END EDIT

I found the introduction unclear. It sort of says what will happen, but some more explanation would be good... something like "we will play excerpts from recordings: unfiltered for three seconds, filtered for three seconds, and then with an adjustable slider which you must adjust to match the filtered sound". Or something like that.

The icon for filtered has cogs. This sort of symbol looks more like those one sees for settings (hardware and software), so there's a sort of cognitive dissonance for experienced web users there.

But not only am I an amateur web programmer, I now seem to be an amateur sound guy, because it all went to fast for me and I can't tell much difference between filtered and unfiltered! (I was always bad at frequency discrimination.)

Re: Tonal Ear Training

PostPosted: Wed Jun 05, 2013 11:19 pm
by Damien
id like to hear the samples a little longer, and maybe un filtered, filtered, unfiltered, filtered. or longer samples.

it was hard to actually hear the difference at the start as the tracks had little content in the area that was filtered. i thaught it didnt work until i got to some tracks where the effect was clearly heard. i didnt know what i was listening for.

i think its cool ,and im about to send my employees to the site to have a go.

cheers,

damien.

Re: Tonal Ear Training

PostPosted: Thu Jun 06, 2013 2:43 am
by Mark Bassett
Thanks again for the feedback.

The experiment is intentionally designed to only play the filtered sound once. The inability to return to the filtered sound is key to the design. It allows me to test for what I'm interested in.

Damien wrote:it was hard to actually hear the difference at the start as the tracks had little content in the area that was filtered. i thaught it didnt work until i got to some tracks where the effect was clearly heard. i didnt know what i was listening for.

This is also part of the study. It is as much a data collection tool, actually far more, than it is an 'ear training' program. It will provide data on these very issues though, like in influence of genre, exposure time, and most importantly training method.

audioio wrote:The icon for filtered has cogs.

Yeah I might change this. But I have to wait for now.

audioio wrote:it all went to fast for me and I can't tell much difference between filtered and unfiltered! (I was always bad at frequency discrimination.)

Stick with it and the three seconds will seem like more than enough time.

I should add that I did not code it. David Major, a fellow audio lecturer here in Dubai did. I built the original version in Max/MSP then he took it online.

He recently made this cool little site http://www.doyouknowblue.com/