IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Detecting the noise floor of a file?
MikeFord
post Apr 16 2013, 02:27
Post #1





Group: Members
Posts: 132
Joined: 22-May 04
Member No.: 14241



Sometimes noise is signal, a recording of the wind or waves crashing on rocks, but I wonder if there wouldn't be some merit to a tool that measured the content of a file with no harmonic relationship to the predominant signal?

A quick and dirty estimate might come from looking at some average level of the quietest parts that are not artificially set to zero or some dither value. Between notes after decay ends, there might not be a lot of such spots, but they might have some common level that could be tracked.

I'm not sure what use knowing the noise floor might be, but I think it might be interesting.
Go to the top of the page
+Quote Post
saratoga
post Apr 16 2013, 04:47
Post #2





Group: Members
Posts: 5116
Joined: 2-September 02
Member No.: 3264



Are you familiar with lossywav? This is essentially what it does.
Go to the top of the page
+Quote Post
MikeFord
post Apr 16 2013, 12:35
Post #3





Group: Members
Posts: 132
Joined: 22-May 04
Member No.: 14241



QUOTE (saratoga @ Apr 15 2013, 20:47) *
Are you familiar with lossywav? This is essentially what it does.

Thanks, not til now, just looked it up. I haven't read enough yet to see if it generates any kind of numerical assessment, or if that would require adding on something. I'll be studying it once I am awake again, and post again if I have questions.
Go to the top of the page
+Quote Post
saratoga
post Apr 16 2013, 16:41
Post #4





Group: Members
Posts: 5116
Joined: 2-September 02
Member No.: 3264



If you want the raw data, you can simply process a file and compare the difference between the original and the processed version. Thats essentially a very good estimate of the noise power per sample.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Apr 23 2013, 12:16
Post #5





Group: Members
Posts: 4247
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (MikeFord @ Apr 15 2013, 21:27) *
Sometimes noise is signal, a recording of the wind or waves crashing on rocks, but I wonder if there wouldn't be some merit to a tool that measured the content of a file with no harmonic relationship to the predominant signal?


You've got a serious inherent problem. The relationship between fundamentals and overtones produced by many musical instruments aren't exact integer relationships.

I usually judge the noise in a file by the amplitude of the file when there is no music playing. However many realk world noise sources such as HVAC equipment that include large amplitude signals with harmonic relationships.

So you've got a conundrum with using the presence or absence of harmonic relationships as a guide. Their presence is not reliably indicative of the signal being music, and their absense is not reliably indicative of the signal being noise.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 23rd November 2014 - 01:02