emkay Posted February 24, 2012 Author Share Posted February 24, 2012 (edited) SAM is a full digital speech prozessor. And this isn't programmed? (As far as I understand, you wouldn't take 'Kennedy Approach', '221b Baker Street' etc. into account, since the voices are based on samples, but what is your problem with SAM?) It is digital using the CPU for producing waveforms , not programmed sound (using POKEY's sound generators) Edited February 24, 2012 by emkay Quote Link to comment Share on other sites More sharing options...
emkay Posted February 24, 2012 Author Share Posted February 24, 2012 http://www.youtube.com/watch?v=GEc8UsVzIU0 The "speech" is clearer than in Wolfenstein Quote Link to comment Share on other sites More sharing options...
emkay Posted February 24, 2012 Author Share Posted February 24, 2012 LOL. I realized that I don't have even one copy of the "Funny step - Roundup" tune.... also the stored places on the Internet got cleared. Quote Link to comment Share on other sites More sharing options...
sack-c0s Posted February 24, 2012 Share Posted February 24, 2012 If we're talking about speech this is probably a good challenge - a very simple music backing but a lot of very mechanical sounding speech. Anyone think they can pull this off (minus the Bill Hicks quotes - that would just eat memory)? I once thought about hacking about a speech synth to try and do it but I never got around to it. How small do you reckon you can get the replayer? Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 24, 2012 Share Posted February 24, 2012 Don't forget Mahoney released all the source for the encoder for Cubase64, which you might get some interesting results with that if you had a clean vocal track to work with.. Or just recorded the lines yourself Quote Link to comment Share on other sites More sharing options...
analmux Posted February 24, 2012 Share Posted February 24, 2012 LOL. I realized that I don't have even one copy of the "Funny step - Roundup" tune.... also the stored places on the Internet got cleared. Quote Link to comment Share on other sites More sharing options...
emkay Posted February 24, 2012 Author Share Posted February 24, 2012 LOL. I realized that I don't have even one copy of the "Funny step - Roundup" tune.... also the stored places on the Internet got cleared. Thanks. Quote Link to comment Share on other sites More sharing options...
emkay Posted February 24, 2012 Author Share Posted February 24, 2012 Hi, hi , hi http://www.youtube.com/watch?v=b5GX4ufbw2c Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 24, 2012 Share Posted February 24, 2012 There's an update to this stuff you might be interested in Quote Link to comment Share on other sites More sharing options...
Mclaneinc Posted February 24, 2012 Share Posted February 24, 2012 http://www.youtube.com/watch?v=9Nz5CJi7gZc Wow, is this what passes as music these days..dear oh dear... Quote Link to comment Share on other sites More sharing options...
sack-c0s Posted February 24, 2012 Share Posted February 24, 2012 It's not great - I was just suggesting it as a technical challenge, rather than 'listen to this - it's brilliant!' Quote Link to comment Share on other sites More sharing options...
emkay Posted February 25, 2012 Author Share Posted February 25, 2012 There's an update to this stuff you might be interested in Another idea comes to my mind. As this is using "channel streaming", it must be possible to play random melodies aswell. Together with gr. 7 (or an even lower mode)an always changing visualizer could be possible. Charmode wouldn't fit well due to the high cycle stealing. It could result in an endless demo. Quote Link to comment Share on other sites More sharing options...
Mclaneinc Posted February 25, 2012 Share Posted February 25, 2012 It's not great - I was just suggesting it as a technical challenge, rather than 'listen to this - it's brilliant!' Yeah I know, wasn't meant as a critique of your musical tastes If someone can recreate that from an Atari I'd be seriously impressed, well as long as it wasn't just a huge ram user, from a stock 800XL or 130xe would be great. Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 25, 2012 Share Posted February 25, 2012 Another idea comes to my mind. As this is using "channel streaming", it must be possible to play random melodies aswell. Together with gr. 7 (or an even lower mode)an always changing visualizer could be possible. Charmode wouldn't fit well due to the high cycle stealing. It could result in an endless demo. That's not using any streaming in the versions there, I was thinking about doing it because the NSF player is so slow, but I haven't actually done that yet.. RED = NSF Player + me shuffling the registers around to play the audio on my virtual channels.. GREEN = Build volume tables BLUE = Synthesise stuff.. Unless you meant something else ? I think this will work reasonably well in char modes actually.. There's a few possible issues, but I think it'll playback fine at 15.6k with a full screen on.. We shall see when I next play with it.. 1 Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 25, 2012 Share Posted February 25, 2012 Now with added chur-cuf-chur-cuff-churr-chufff Now with added skiffleboard.zip 1 Quote Link to comment Share on other sites More sharing options...
Creature XL Posted February 25, 2012 Share Posted February 25, 2012 Andy can you make it run with non-VBXE machines? I didn't try the latest version, but the others work for 1 second then they crash. However, I have to read your description in detail tomorrow, the sound is awesome. I want to try emkay's idea, to use this technique for low notes. BTW, the NES has a sawtooth voice as well, right? This could be simulated with the POKEY-sawtooth "tricks". So you need 1 POKEY channel for puls waves, 1 for noise and 2 for sawtooth. Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 25, 2012 Share Posted February 25, 2012 They will crash because it'll promptly generate 500+ blits over the poor NES sound driver.. It's synthesising it all in software, just all as pulse waves.. I'm not really doing a triangle wave for voice 2, just a quieter square wave.. I'm not going to bother with the triangle since I've got other goals, I think.. It's not meant to be a NES emulation, just a means to play around with this low overhead audio method.. But it's been a revelation in how few interrupts doing the pulse waves actually needs per frame.. I just threw the Pokey in to do the noise channel, since it was annoying me it was missing, and generating noise with the pulse method is too expensive at the moment, though in the future maybe not so.. ANTIC versions of previous.zip 2 Quote Link to comment Share on other sites More sharing options...
emkay Posted February 26, 2012 Author Share Posted February 26, 2012 LOL. So it does exactly the "vice versa" of what is good for POKEY. Playing the high notes through software and the low notes/percussion via hardware. Is this some curse or else? Quote Link to comment Share on other sites More sharing options...
emkay Posted February 26, 2012 Author Share Posted February 26, 2012 BTW, the NES has a sawtooth voice as well, right? This could be simulated with the POKEY-sawtooth "tricks". Using Sawtooth isn't a good choice here. It's rather limited and costs a channel. 2 16 bit channels would do better, if you want to have those "musical conversions" from the NES. Quote Link to comment Share on other sites More sharing options...
miker Posted February 26, 2012 Share Posted February 26, 2012 (edited) @andym00: Hm... cool as there are ANTIC version of these tunes. But they are still not loadable on real Atari. Could you move the code after $2000 so the files will become more DOS-loadable, please? Edited February 26, 2012 by miker Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 26, 2012 Share Posted February 26, 2012 Ah... Forgot about that.. I lowered when I was trying for a bigger NSF and forgot to put it back to something sensible.. Will do that, but it's pub time now (null) Quote Link to comment Share on other sites More sharing options...
miker Posted February 26, 2012 Share Posted February 26, 2012 OK, thanks in advance. Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 26, 2012 Share Posted February 26, 2012 By the way, is $2000 the best start point ? I mean in compatibility terms ? I've always gone for $1800 since that worked fine.. Curious what the best address is to use now (null) Quote Link to comment Share on other sites More sharing options...
miker Posted February 26, 2012 Share Posted February 26, 2012 In general, if the file is meant to be really "DOS loadable", then you shouldn't go under $2000 (of course directly, without any relocating). But for most no-DOS loaders $1800 is also fine. Quote Link to comment Share on other sites More sharing options...
andym00 Posted February 27, 2012 Share Posted February 27, 2012 ANTIC versions, with the a real screen on, doing things, with 15.6K playback rate.. It does need some memory expansion though! Trying to work around the 16K ROM footprint of the the NES music code, and then 16K of RAM needed to capture APU writes (and other things) is just too much of a hole for me too keep working around.. I took the NSF relocator out because it was corrupting some tunes.. I tried some other methods of figuring out if I was looking at code or data but nothing was utterly fool proof.. So.. I resorted to using a 16KB bank to do the capturing.. I didn't want to waste lots of time on something that's not important in the long run.. That aside, something has broken somewhere.. Tunes that once did play have gone a bit wonky here and there, something is corrupting something somewhere.. God.. I hate searching for those kind of bugs! Anyway... You now get the same display as the VBXE, but using ANTIC, so it's mostly mode E and a real world kind of setup.. And note.... Big note.. This is running with wsync stabilised IRQs which means it's burning more time than necessary, but sounds nicer.. It has highlighted that there's no way on earth that you're going to have 15.6K (with it stabilised) output when using mode 2 or 4/5 with the entire badline.. That said, the non-stabilised version doesn't sound too bad actually VBXE be gone.zip 4 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.