Skip to main content
Topic: Analysis of knowzy samples (Read 4457 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Analysis of knowzy samples

Steely Dan, Aja, Home At Last. CD vs AT LP-2D USB table, L+R and L-R signals. Window length is 300ms, overlap is 66%, Blackman-Harris window.

EDIT: Deleted pic because I looked at the code and it has serious issues. I'll post updated (and more) pics soon.

Analysis of knowzy samples

Reply #1
New pics.
  • Huge bug fix #1: Removed some experimental filtering code I had in place for some other vinyl investigations. Woops.
  • Huge bug fix #2: Plots were being clipped somewhere around 5-10khz but being displayed as if they went all the way to 22k. Woops.
  • This time I've split up the plots into 1-1k and 1k-22k, because the response ranges for both tend to be vastly different, and because different blocking factors work better at different frequencies.
  • Overlap is now 75%, Hanning window, mean of amplitudes of windowed spectra.
  • Window length is 1s for 1-1k plot and 0.1s for 1k-22k plot.
Plots as follows:
  • 1hz-1khz, Pink Floyd "On The Run", CD vs AT2D, complete song - I'm pretty sure the CD is the 1991 Capitol version, so not the 20th anniv version that Jeff has (I think).
  • 1khz-22khz, Pink Floyd "On The Run", CD vs AT2D
  • 1hz-1khz, Steely Dan "Home At Last", CD vs AT2D (snip)
  • 1khz-22khz, Steely Dan "Home At Last", CD vs AT2D (snip)

As usual, black is L+R, red is L-R.


SimplePortal 1.0.0 RC1 © 2008-2020