No$gba v2.4a fixes a newly discoverd adpcm sound bug... hmmmm, my computer is too slow to produce continous sound with HLL coded games... but for people with faster computers... is that right, nobody of you guys has ever noticed adpcm problems since implementing sound in v2.3a?

04 April 2007 - version 2.4a
- nds/help: added nds-lite case/joypad ascii arts, and nds-lite supply pinout
- spam: nocash-lite was posted online - new addresses in email.htm or about-box
- nds/3d: forces instant new gxfifo irq (if any) on attempt to acknowlege it
- nds/sound: fixed output to sample.wav script file (for diagnostics purposes)
- nds/adpcm: emulates nds-rounding-error, uses fast-pre-multiplied adpcm tables
- nds/adpcm: fixed clipping-bug in no$gba, and emulates real nds-clipping-error
- nds/sound/help: added more ima-adpcm notes (rounding-error and clipping-error)
- stat: performance indicator counts frames on nds9 only (=60Hz) (instead 120Hz)
- nds/tsc/debug: displays all tsc channel values in io-map (touchscr, mic, etc.)
- nds/tsc: prevents penirq if disabled, penirq more in sync with tsc-adc values
- nds/3d: drains gxfifo before gxfifo dma (fixes mariokart/downhill slowdowns)
- nds/3d: forces DepthMask=1 on ClearDepth (if it was 0 from translucent/attr)
- nds/firmware: applies different mac addr to each machine with wifi-crc adjust
- nds/firmware: forces def.fmw/wifi header (if firmware.bin missing) (downhill)
- nds/firmware: forces user settings [65h]=FCh=good (if firmware.bin missing)
- nds/firmware: fixed touchscreen/calib screen coords origin 1,1 (instead 0,0)
- nds/dos/3d: translates by w, recurses current viewport x1/y1 and width/height