Aley Keprt wrote: > Thanks. > I tried it but can't manage to get it work in WinCoupe (I need to test i > there). Whenever I start playback, the message "press symbol to exit" > shortly appears and then it immediately closes (going back to main menu), > like I presses symbol shift. But I don't press anything.
Thanks for giving it a go... It sounds like you are not loading an e-tracker file (there a one or two on the disk). The Midi-tracker file is just a bunch of settings specifying how each instrument in the e-tracker file should be played (program number, velocity, etc.). So to hear anything you have to load both an e-tracker file *and* its corresponding Midi-tracker file before pressing play. > Other MIDI program "MIDI Sequencer" works in WinCoupe, but there are some > other problems with it. I'd like to know whether the emulation isn't > correct, or the original program is ehmm... not-very-good. It > takes over 10 > seconds to redraw whole screen in edit mode, and whenever I start > playback a > very loud "home-less" tone is generated before starting the > actual playback. Yes, I remember MIDI Sequencer being quite slow! A lot of the user interface is in BASIC IIRC (as is my program!) > Also, I think it wouldn't be very clever to use MIDI simulator in DOS, > because it would degrade digital-audio playback quality. Of course, this > doesn't affect using real MIDI devices like MPU401. Btw.. SimCoupé does no internal sequencing at the moment to give correct MIDI _output_ timing (it tries to get the timing as SAM sees it correctly of course!) - it just writes the data straight out to Windows when it gets it. An implementation would be a lot easier than managing the SAA output, but has not been done yet... > Adding MIDI interrupts for demos gives more sense. I'd like to see it. > Please could you tell me where can I get those demos? I can put mine up soon... I might just, err, sort out the scrolly text first.. :-) I'll reply again when it's sorted. I'm not sure if anyone else thought about using MIDI for timing... > Note: > PAL speed is 1/64MHz. MIDI speed is 1/32MHz. It means that 2 bits are > transmitted per line. > Am I right? Yep, that's right... > How many bits are transmitted to complete one byte? If you say > interrupt is > generated 5 lines after issuing MIDI out, it means that one or two > (start/)stop bits are added to each byte. I assume MIDI out interrupts are > generated on line basis, exactly as the line and frame interrupts. Am I > right? Two extra bits are sent, yes. The interrupts are not linked to screen lines though. When an output is made to a MIDI port the transmission starts more-or-less straight away... although the MIDI hardware seems to have a resolution of 32 T-States (apparently at offsets of 12, 44, 76, etc from the beginning of the frame). The interrupt occurs one half-bit (96 T-States) before the transmission completes, and lasts for those last 96 T-States (so, shorter than other interrupts). The TXFMST bit in LPEN is set throughout the transmission, and outputting to the MIDI port is ignored if a byte is already in the process of being sent. Dave. > Aley
