Sign In
Create Account
Follow Us on Twitter
Subscribe on Youtube
3
We are to create an application that when you play a song can recognize if there is a SNARE DRUM SOUND that is present. I want to ask if it's possible to create a database that contains sound. What database are e gonna use? And if it is possible, how can it be done? Also another question how can that sound in a specific song be recognized? Anyway we are using C# as our Programming language.
Hope you can help me.. Our deadline is nearly coming..
Thanks a lot!!
2 replies to this topic
Back to top|
|
|
#2
Posted 03 July 2011 - 12:20 AM
Alexander
-
- Moderators
-
- 4,118 posts
It's Science!
- Location:Vancouver, Eh! Cleverness: 200
I assume you could use an audio processing library, as the snare drum (if simple) will likely have a reproducable frequency and amplitude, however you would require a fair amount of knowledge in that specific library to do something like this.
Be sure to read the updated FAQ! || Health is achieved through the same 10,000 steps.
If a suggested code/method fails, informing us is less important than telling us why or what errors occurred.
If a suggested code/method fails, informing us is less important than telling us why or what errors occurred.
#3
Posted 08 July 2011 - 09:40 AM
sam_coder
-
- Members
-
- 372 posts
Programming Expert
in theory, a database could contain anything at all.
a binary field for example is what you would use if you were storing the actual sound. But I doubt that would be very efficient.
In order to identify a sound, you would have to process the input, and I would imagine generate a sort of hash. or at least I would assume.
The problem of course, is that there would almost have to be an expected difference, so you would be generating a list of sounds, that are similar. Your results would almost look like
95% sure it's this song
70% sure it's this other song
So, it wouldn't make much sense to store all the sounds, as sounds in the database, that would result in pulling all the sounds back every time you compared it.
What you would want is to process the sounds. and have the program put a meaningful hash of some sort on it. And stored in a way, that when the sample you want to compare, could generate a similar hash, and your program could query the database for sounds it processed previously, that sounded similar. It would also have to calculate how similar.
Anyways, I don't know a whole lot on the topic. That's just my two cents. =)
a binary field for example is what you would use if you were storing the actual sound. But I doubt that would be very efficient.
In order to identify a sound, you would have to process the input, and I would imagine generate a sort of hash. or at least I would assume.
The problem of course, is that there would almost have to be an expected difference, so you would be generating a list of sounds, that are similar. Your results would almost look like
95% sure it's this song
70% sure it's this other song
So, it wouldn't make much sense to store all the sounds, as sounds in the database, that would result in pulling all the sounds back every time you compared it.
What you would want is to process the sounds. and have the program put a meaningful hash of some sort on it. And stored in a way, that when the sample you want to compare, could generate a similar hash, and your program could query the database for sounds it processed previously, that sounded similar. It would also have to calculate how similar.
Anyways, I don't know a whole lot on the topic. That's just my two cents. =)
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users
