I searched for this and was unable to find the informations, so please forgive me if this is posted some place else.
In regard to the spectrum analyser in the area to that far left (bass range) I always see this represented as a straight line usually until around 200 hz. then showing more detail. I see this reprecented even with sounds not having much sub bass. Can anybody explain this to me?
Many Thanks!
Dan
The spectrum analyser probably uses fast fourier transform (FFT) to analyse sound into its frequency components. This method has much lower resolution in low frequencies than it has in the high frequencies. If you set the “frequency scaling” mode to “linear” you see the data quasi plotted with the actual resolution - very detailed up high, but the bass is only a very small range. The logarithmic mode rescales the data so it is more useful for musical purposes.
You can gain bass resolution by setting the “block size” to a higher value. Then the analyser won’t react as fast and take more cpu power. But you will see the bass frequency peaks now - they are rather wide, but you will be able to distinguish frequencies because of the raised resolution.
The spectrum analyser probably uses fast fourier transform (FFT) to analyse sound into its frequency components. This method has much lower resolution in low frequencies than it has in the high frequencies. If you set the “frequency scaling” mode to “linear” you see the data quasi plotted with the actual resolution - very detailed up high, but the bass is only a very small range. The logarithmic mode rescales the data so it is more useful for musical purposes.
You can gain bass resolution by setting the “block size” to a higher value. Then the analyser won’t react as fast and take more cpu power. But you will see the bass frequency peaks now - they are rather wide, but you will be able to distinguish frequencies because of the raised resolution.
Thank you, Very informative!
Well - right click the spectrogram, there you can set the size.
I presume it is the number of samples that are regularly fed into a fft to create the frequency graph after some processing. The more samples, the better (more detailed) the result, but it is also blurred over time. I have running at maximum value all the time.