Skip to main content

Our research group at the International Audio Laboratories Erlangen is working on some topics in music perception. We ran several experiments to find out: What is the number of instruments in polyphonic music humans can estimate correctly?
To compare our internal results with a broader user base, we published a new web-based listening experiment at http://www.audiolab… WICE : Web Instrument Count Experiment[/]="http://www.audiolab… WICE : Web Instrument Count Experiment[/]

We would be very thankful for your participation. The experiment takes ~10 minutes to finish.

Any feedback or questions about the experiment are very welcome!

Topic Tags

Comments

audiokid Mon, 02/18/2013 - 10:28

Good for you Jeff, That's pretty impressive. Thats your acoustic experience showing there!

I got :

Congratulations! You achieved a score of 845

72.86% of all listeners didn't achieve a score as high as yours!
Number of correct items: 7/12

I'm going to try it again. Looks like we can have fun with this.

audiokid Tue, 02/19/2013 - 09:39

Interesting results. Did you look at them? On a previous one I got those right. Pays to not over think.

But an interesting experiment and company.

I think its worth supporting here. Try it.

The International Audio Laboratories Erlangen (AudioLabs) develops technologies for future applications incorporating audio and video components. To date, three professors have been appointed to work on the topics audio coding, audio signal analysis, and perception-based spatial audio signal processing. Planned growth over the coming years will include the addition of further research topics.

audiokid Tue, 02/19/2013 - 09:47

bayerha, post: 400824 wrote: Our research group at the International Audio Laboratories Erlangen is working on some topics in music perception. We ran several experiments to find out: What is the number of instruments in polyphonic music humans can estimate correctly?
To compare our internal results with a broader user base, we published a new web-based listening experiment at [[url=http://[/URL]="http://www.audiolab…"]WICE : Web Instrument Count Experiment[/]="http://www.audiolab…"]WICE : Web Instrument Count Experiment[/]

We would be very thankful for your participation. The experiment takes ~10 minutes to finish.

Any feedback or questions about the experiment are very welcome!

I think its worth supporting here. Try it.

An interesting experiment and company.

The International Audio Laboratories Erlangen (AudioLabs) develops technologies for future applications incorporating audio and video components. To date, three professors have been appointed to work on the topics audio coding, audio signal analysis, and perception-based spatial audio signal processing. Planned growth over the coming years will include the addition of further research topics.

bayerha,

Could we have feedback from you or your company, please?
I'm particularly interested in perception-based spatial audio signal processing.

I would like to know how this music was processed because, to my ears it doesn't have good spacial separation. However, I am 56 so is that me?
My analog summing system ( or my personal approach) would have separated all these instruments much better than inside the box. This is how I hear most music today. Digitally summed music is one big mash-up. Maybe this is why I appreciate and invest in a recording and mixing system that can process and separate audio more than others?

Can you tell us more?

anonymous Tue, 02/19/2013 - 14:06

Some Details

Dear Forum,

thank you all for your interest and your participation. Your input is very helpful and we are very glad, that find our experiment interesting.

audiokid wrote:
An interesting experiment and company.

We are actually an independent research institute at the University of Erlangen/Nürnberg in Germany. We work in close collaboration with the Fraunhofer IIS (Home of MP3) in nearly all fields of audio research.

audiokid wrote:
Could we have feedback from you or your company, please?
I'm particularly interested in perception-based spatial audio signal processing.

Sure. I can try to answer your questions as good as possible. Please keep in mind that more information about the intention and background of the experiment can spoil future participants so I will keep it short here.

audiokid wrote:
I would like to know how this music was processed because, to my ears it doesn't have good spacial separation.

I am wondering that you even hear any spacial separation at all. Because the files are completely mixed in mono. This is totally intended as we want to focus on other cues than spatial ones.

audiokid wrote:
My analog summing system ( or my personal approach) would have separated all these instruments much better than inside the box. This is how I hear most music today. Digitally summed music is one big mash-up. Maybe this is why I appreciate and invest in a recording and mixing system that can process and separate audio more than others?
Can you tell us more?

Techniques and methods that are usually applied by audio engineers to let some instruments better stand out in a mix have not been considered here. The mixing has been done as "raw" as possible. We have only added some basic convolutive reverb to the tracks.

I can post further details when we have finished the experiment and we have published the results if you are interested here.

Fabian

x

User login