home *** CD-ROM | disk | FTP | other *** search
-
- -----------------------------------------
- SCOPEXT by Smack/Infect
- -----------------------------------------
-
- this tiny prog is some kind of a 'sampling oscilloscope'. it uses
- the parallel port sampler, if you don't have one, then this prog
- will be useless for you.
-
- requires: AGA/68020+/OS V37+
-
- the sampling rate is 16kHz so frequencies above 8kHz will not be
- visible. if you say now 'this is not very much' then i wanna tell you:
- 16kHz means that during every pal-frame we receive 320 bytes so it fits
- perfectly on a lores screen. and there's simply not enough cpu power
- on todays amigas to use a higher resolution for this task. (i think ;-)
-
- the prog is controlled via keyboard so try these:
- F1 - lores 20ms display
- F2 - hires 40ms display
- F3 - single dot
- F4 - eight average dots
- F5 - line (*)
- F6 - solid (*)
- F7 - spectrum analyzer
- F9 - show idle cpu time
- F10- freeze display
- ESC- exit
-
- NOTE: the modes marked with (*) need different amounts of cpu time
- depending on the current sample data.
- if the screen borders get red then the plotting routine runs
- too slow. to avoid this you could a) get a powerful cpu and
- some fast ram b) use F2 for 40ms display c) if that doesn't
- help choose a different plotting mode
-
- ABOUT THE SPECTRUM ANALYZER:
-
- it uses a 128-point fourier transformation which seems to be quite accurate.
- each bar represents a band of about 125Hz, the 64 bars go upto 8kHz.
-
- the spectrum analyzer always scans only 256 bytes of the sample,
- this is due to the fft loop, sorry! if your machine is fast enough you
- should run it on 20ms display, so you 'miss' only 64 bytes each frame
- which is 20% of the whole sample. on 40ms display your chances to miss
- short impulses (such as hihats etc.) are 60%...
- i don't know how to fix the problem of these 'time gaps' without coding
- a new fft loop from scratch!
-
-
-
- and remember: this is only a programming experiment. you can't claim
- me for any damages or trouble my prog could cause. it's all your
- fault coz you dared to start it ;-)
-
-
- contact me if you want:
- Michael Henke
- Praetorius-Str.1/205
- 06124 Halle
- Germany
-
-
-
- HISTORY
-
- 170195
- -first release
-
- 190195
- -FFT spectrum analyzer implemented
- NOTE: in this first version there are some 'inaccuracies' in the
- frequency range. i will fix it as soon as i get some equipment to
- calibrate the frequencies. and maybe i'll implement 'acoustic scaling'
- which linearizes the spectrum like the human ear does.
-
- 200195
- -spectrum analyzer calibrated (using Masque's Chicken Dialer!!)
- -peak levels implemented
-
- 240195
- -code optimizing
- -fixed a serious bug in the sampling routine that appeared on the 68030
- but not on the 68020! (strange? it was caused by an unintentional
- instruction overlap in the cpu's pipeline.)
-
- 310195
- -sprite anim built in -> dramatical file length increase, but who cares?!
- the logo can be switched on/off using the HELP key
-