Image Map Image Map
Page 8 of 8 FirstFirst ... 45678
Results 71 to 80 of 80

Thread: What I'm working on 2, Sound

  1. #71

    Default

    Quote Originally Posted by Scali View Post
    Could you give a short recording and/or binary of that?
    I have difficultly imagining what this would sound like.
    Imagine no longer... I just got it to where I'm willing to share it, but not willing to fully document yet. You guys have been good to me despite my being something of an ass, so here's a early preview:

    http://www.deathshadow.com/downloads/musicdmo.zip

    It SHOULD have IMF support too, no idea if that works. To hear the 1.5 bit speaker, either run it with no sound card, or override autodetection with /speaker

    You can also /arpeggio to hear what that was gonna be like... some systems that sounds better, some it sounds worse.

    Tossed a readme.txt in there with command line switches and some thoughts/info. As always though you can just /?

    /debug can tell you some cute stuff like how much memory it wastes...

    the "1.5 bit" is a bit of a mixed bag depending on hardware. My PC/XT clone it sounds great... my Tandy 1000SX it's... noisy. My K62/450 (new middle machine after AM5x86-150 went tits up) it sounds... eh, I'd rather not talk about it... PS/2 Model 30-286 it sounds HORRIBLE (worst piezo EVER)... oddly enough DOSBOX seems to handle it quite nicely (probably the best sounding of them?!?)

    Which is interesting for alternating between on, off, and 44khz at 2.4khz source mixing. The concept alone is... different.

    Was looking at that 2.4khz output routine and thinking I can make that do covox and/or disney with little to no changes apart from what port I'm shoving data at and how I add the voices together.

    Quote Originally Posted by Trixter View Post
    It was meant to simulate a PC speaker, and like the speaker, needed to sustain the tone infinitely once fired and forgotten.
    Which is what sending a note-on does until you set note-off, that's why you set sustain, NOT release. That patch holds the release high with no sustain, which makes very little sense... in a way I'm surprised it even holds... infinite lack of release, that's... strange. It's not like with PC speaker you don't tell it to stop playing... that's what voice on/off messages are for.

    ADSR... you'd think if you want a note to sustain you'd set the sustain, not the release... with both at the same level.

    The patch values I listed above SHOULD be functionally identical though really assuming infinite decay is the same as skipping the sustain entirely. Really though you don't want the note to stop playing, don't release/note off.
    From time to time the accessibility of a website must be refreshed with the blood of owners and designers. It is its natural manure.
    CUTCODEDOWN.COM

  2. #72

    Default

    Just used my little "let's pretend a Teensy 3.2 is a digital oscilloscope" on the output from the K6-2/450 to try and figure out why it sounds so bad... Interesting thing that...

    The speaker port implementation is WRONG... like ... REALLY wrong.

    The "on" value, is the same as off. So on both platforms instead of 0, 50%, and full on, I'm getting 0 and 50%... Anything that would bit-bang the port manually for PWM without using the PIT generated output would in fact output nothing on that machine. In only outputs when connected to the PIT directly.

    I'm wondering if my PS2 Model 30-286 is equally incorrect... but again I suspect that piezo buzzer they stuck in there is the real culprit. Be fun if it was BOTH problems... Question is, do I want to take it apart enough to figure out where to hook up to read the output from it?
    From time to time the accessibility of a website must be refreshed with the blood of owners and designers. It is its natural manure.
    CUTCODEDOWN.COM

  3. #73
    Join Date
    Aug 2006
    Location
    Chicagoland, Illinois, USA
    Posts
    4,419
    Blog Entries
    1

    Default

    On an IBM PC 5160: I tried /arpeggio to start and it played, but then I found my keyboard locked up once returned to DOS. I tried again with /adlib and it did the same (keyboard locked up once back at DOS). I tried with /speaker and it did NOT lock up the keyboard on return to DOS, but the output was somewhat garbled and muted; I'm assuming your comments prior to this post show you need to work on that a little more.

    I can test on a real IMFC and CMS once these issues are worked out. Would you prefer bug reports here, or to the email you provided in the readme.txt?
    Offering a bounty for:
    - Documentation and original distribution disks for: Panasonic Sr. Partner, Corona PPC-400, Zenith Z-160 series
    - Music Construction Set, IBM Music Feature edition (has red sticker on front stating IBM Music Feature)
    - Any very old/ugly IBM joystick (such as the Franklin JS-123)

  4. #74

    Default

    Quote Originally Posted by Trixter View Post
    On an IBM PC 5160: I tried /arpeggio to start and it played, but then I found my keyboard locked up once returned to DOS. I tried again with /adlib and it did the same (keyboard locked up once back at DOS).
    I didn't have keyboard lockups here until I tried it on a 1000HX, it locks up there. Then I noticed it does it on the PC-7000 too -- but NOT the 1000SX, my PC/XT clone, or any newer system.

    The keyboard lockouts seem to be oddly inconsistent across platforms, having a hard time tracking it down... doesn't help that it doesn't lockup the keyboard EVER in DOSBox, and evaluating problems like that on the real hardware is a pain in the ass.

    But since I hook the keyboard interrupt, it's probably not releasing properly or the original pointer is getting screwed up...

    Though I'm wondering if my release order is related... release the keyboard ISR before the timer?
    Last edited by deathshadow; April 12th, 2017 at 10:34 PM.
    From time to time the accessibility of a website must be refreshed with the blood of owners and designers. It is its natural manure.
    CUTCODEDOWN.COM

  5. #75
    Join Date
    Dec 2014
    Location
    The Netherlands
    Posts
    1,631

    Default

    Quote Originally Posted by deathshadow View Post
    Imagine no longer... I just got it to where I'm willing to share it, but not willing to fully document yet. You guys have been good to me despite my being something of an ass, so here's a early preview:
    Cool!
    I've tried it on my 'music' machine, which has an SN76489 and SB Pro 2.0 connected (and a Roland Juno-D on the SB's midi port).
    The PC speaker is piezo, but it seemed to work somewhat okay. There was just some fluctuation in the volume, if that makes sense (might be the same thing Trixter describes with 'muted'). Some parts seemed very soft, others very loud. Other than that it sounded like a nice rendition of the PacMan tune.
    When I tried the AdLib mode, I had the same issue as Trixter: music played, keyboard locked up afterward (didn't happen with speaker).

    I also tried it with my DOSBox build with quick-and-dirty IMFC emulation, and that one produced sound as well. So it may just work on real hardware Sadly I don't have a real IMFC. If I ever find one, I'll buy it

  6. #76

    Default

    Duh, I'm an idiot.

    I was storing the value of 0x61 from the start, but of course the keyboard is ALSO handled through port 0x61 -- so I HAVE to read it and and it 0xFC every time before output. There's no choice on that. That high bit (keyboard clear) could be biting me... as could things like the PS/2 implementation of setting the keyboard clock.

    Also explains why it is inconsistent across platforms, not everyone implements the PPI or its later replacements the same way. Not even all PC/XT clones do it the same way as IBM did. (see tandy 1k's where even the EX/HX are different from SX and no bloody A, B, C, or D...)
    From time to time the accessibility of a website must be refreshed with the blood of owners and designers. It is its natural manure.
    CUTCODEDOWN.COM

  7. #77
    Join Date
    Aug 2006
    Location
    Chicagoland, Illinois, USA
    Posts
    4,419
    Blog Entries
    1

    Default

    This may help you code for different PPI implementations (see section 7.5).
    Offering a bounty for:
    - Documentation and original distribution disks for: Panasonic Sr. Partner, Corona PPC-400, Zenith Z-160 series
    - Music Construction Set, IBM Music Feature edition (has red sticker on front stating IBM Music Feature)
    - Any very old/ugly IBM joystick (such as the Franklin JS-123)

  8. #78

    Default

    ... and that's part of why I HATE when things take the underscore off of links and use some colour that at a glance is either invisible or not different enough from black to even be noticed...

    Since I sat here for five days not wanting to be a rude ass and post "section 7.5 of WHAT?!?"

    all because I didn't realize / couldn't see that "this" was a link.

    Time for ANOTHER addition to the horde of user.css I have to use to even make vBulleshit based forums usable.
    From time to time the accessibility of a website must be refreshed with the blood of owners and designers. It is its natural manure.
    CUTCODEDOWN.COM

  9. #79
    Join Date
    Aug 2006
    Location
    Chicagoland, Illinois, USA
    Posts
    4,419
    Blog Entries
    1

    Default

    Offering a bounty for:
    - Documentation and original distribution disks for: Panasonic Sr. Partner, Corona PPC-400, Zenith Z-160 series
    - Music Construction Set, IBM Music Feature edition (has red sticker on front stating IBM Music Feature)
    - Any very old/ugly IBM joystick (such as the Franklin JS-123)

  10. #80

    Default

    I'm in vivaldi, I'm running Stylish, so I'm more than aware of what that article is talking about, hence the term 'user.css' which is what Stylish and pretty much everything else that article mentioned DOES and HOW they do it.

    ... along with forcing all the fonts to ACCESSIBLE sizes and metrics -- since declaring font sizes in pixels is inept halfwit bullshit proving that the people developing for vBull have zero business coding the front-end of a damned thing... going hand in hand with the "semantic markup, what's that?!?" ineptitude.

    aka functionality that was built into Opera before they told their entire loyal userbase to go **** themselves up the **** with their shoddy inept slapping of the big red "O" over Chrome any-old-way.
    From time to time the accessibility of a website must be refreshed with the blood of owners and designers. It is its natural manure.
    CUTCODEDOWN.COM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •