dietermoreno
- Dec 30, 2012
- 238
- Joined
- Dec 30, 2012
- Messages
- 238
So I recorded the output of my guitar with no strings being played and recorded this garbage which was roughly tuned into the medium wave band with the pickup selector switch and more finely tuned with both the volume pot on my guitar and the tone pot on my guitar.
Then I used Audacity to equalize the frequency response, attenuate frequencies that are not important for hearing human voices (e.g.: low pass at 500 Hz and high pass at 2000 Hz, with extra attention to attenuation below 100 Hz to suppress mains hum), then the removed garbage allowed more signal to be amplified without clipping, then this showed clipping which was removed with a hard limiter, and then amplify yet again.
This is all in Audacity. So the computer software is actually part of the radio receiver because proffesional audio gear to do all of these step is not cheap (and Audacity is free software).
Unfortunately Audacity does not have "monitor in real time" due to piracy concerns, so the tuning is a broad frequency sweep on the pots once the medium wave band is established with the pick up selector switch.
This is what it sounds like:
So is my explanation on my video description correct:
Then I used Audacity to equalize the frequency response, attenuate frequencies that are not important for hearing human voices (e.g.: low pass at 500 Hz and high pass at 2000 Hz, with extra attention to attenuation below 100 Hz to suppress mains hum), then the removed garbage allowed more signal to be amplified without clipping, then this showed clipping which was removed with a hard limiter, and then amplify yet again.
This is all in Audacity. So the computer software is actually part of the radio receiver because proffesional audio gear to do all of these step is not cheap (and Audacity is free software).
Unfortunately Audacity does not have "monitor in real time" due to piracy concerns, so the tuning is a broad frequency sweep on the pots once the medium wave band is established with the pick up selector switch.
This is what it sounds like:
So is my explanation on my video description correct:
dietermoreno said:I found something interesting when I started to care about SNR values of my shitty guitar recordings that has radio interference bleeding through.
I found that the same EQ, attenuation, and amplification used in Audacity can improve SNR of received demodulated radio signals.
SNR = ((max db level of signal ) - (max db level of noise) ) / ((max db level of noise))
SNR db = 10 log [ ((max db level of signal ) - (max db level of noise) ) / ((max db level of noise))] This equation only works when the SNR is less than 1:1.
In this video:
((-36 db) - (-48 db)) / (-48 db) = ( -36 db + 48 db )/ (-48 db) = 12 db / -48 db
= 0.28 S:N
10log [ 0.28]= 5.5db
Around 5 db SNR is common for AM radio regardless of the receiver used. The only difference with more complicated receiver is less interfering frequencies heard so the signal is more intelligable.
In my previous guitar cover recording videos:
((-2db) - (-48db)) / (-48 db) = +46db / -48 db = 0.95 S:N This is about 4 times greater SNR so about 20 db SNR.
This is standard for FM, but does not take into effect interfering signals.
For comparasion, CD is 97 db SNR which is 70,794 times bigger than the noise, compared to analog radio almost always has less than 1:1 relationship with signal : noise.
So this is why the forumla works for analog radio because must be less than 1:1 because log (0) is undefined and very close to log ( 0) the equation doesn't work.