cheapo statistics 1.04 [effect] -------------------------------------------------------------- Usage: - connect it as normal - right mouse click on machine box -> "Show analysis..." - right mouse click on machine box -> "Reset analysis" Not too hard to use is it ;) -------------------------------------------------------------- Statistics information: "Statistical data collected" Tells how much data machine has analysed in ticks, normal time format [h:mm:ss.ms] and even actual count of samples. Statistical data is gathered only when signal passes thru the machine. Buzz does not use unneeded machines so the value here is probably not the same as the playing time of the tune. You can use in sequence editor to disable signal analysis. (=bypass) If the incoming signal is stereo following values are displayed for "left" and "right" channels separately, otherwise only one set of values is displayed under "Mono signal". "Max sample": The value of the biggest positive sample in the signal. Shown attributes: Sample value and relative power in percents and decibels (relative to +32767.0). "Max value location": The time stamp when "Max sample" was detected. Shown attributes: Ticks, normal time format and sample count. "Min sample": The value of the biggest negative sample in the signal. Shown attributes: Sample value and relative power in percents and decibels (relative to -32768.0). "Min value location": The time stamp when "Min sample" was detected. Shown attributes: Ticks, normal time format and sample count. Min and Max sample values are "Peak" values of the signal. These tell the maximum amplitude of the song. "RMS" is a much better way analyse the perceived loudness of the sound. "DC Offset (average)": The sum of all the analysed samples divided by the number of the samples. This figure tells you if the signal is not centered around the zero axis. Having a bad dc offset limits the available "space" for the amplitude of the signal. Please check the "cheapo dc" for more information and how to correct it. Shown attributes: Sample value and relative power in percents and decibels (relative to +32768.0). "RMS Power": The "root-mean-square" value for the analysed signal [rms=sqrt((a[0]*a[0]+ ... +a[n-1]*a[n-1])/n)]. RMS value is a more meaningful way to express signal loudness then the "Peak" values. The higher the value the louder the song is in general. You can increase the value by adding a compressor and a gain to the main signal. RMS Shown attributes: RMS value and relative power in percents and decibels (relative to +32768.0). "Zero crossings": I'm mot really sure about what this is useful for, but you can find out the frequency of a sinewave with it (for example) =) A zero crossing happens when the signal values move from the positive values to negative values (or the other way round). Shown attributes: Number of zero crossings and the frequency of the possible sinewave. (Yes I have really no clue about this) "Max change": New in 1.03. Max change compares the value of the current sample to the value of the previous sample ( code snippet: change=abs(buffer[c]-buffer[c-1]) ). This value is usually below the peak values and the only real use for it (which I've figured out) is that it tells you if there has been a HUGE amplitude change in the sound (which might cause clicks). Shown attributes: Maximum change value and relative power in percents and decibels (relative to +32768.0). A change of 32767 (from 0 to 32767) is 0dB and 65534 (from 32767 -> -32767) is +6.0dB. "Volume slider suggestions based on max and min levels": This section calculates the proper values for the master output and the next machine's input level slider if the signal exceeds the limits (-32768 and +32767). Some machines handle signal that has value outside the normal limits, but some don't. This might cause clipping to a loud signal that goes through the machine. Geonik's saturator is one machine that clips the signal between -32768 and 32767. The master output clips the signal too, so setting the proper value is important in preserving the signal as much as possible. -------------------------------------------------------------- History: 1.04 Added new statistics item: "Average power" 1.03 Noticed an error when suggesting volumes when the signal is really loud: the machine would suggest values that are not in the range of the volume faders. "Please set master volume to 21211 [129.35%]." Added new statistics item: "Max change", "Max change location" 1.02 Got a little guilty consciousness for not obeying the WM_READ mode. now bypasses machine. 1.01 Fixed zero crossings miscalculation and added some more info. Thanks to Tamas for reporting. 1.0 First release -------------------------------------------------------------- oh.. almost forgot: YOU NEED BUZZ1.2 TO USE THIS EFFECT Mikko Apo [http://iki.fi/apo/]