emkay Posted August 18, 2004 Share Posted August 18, 2004 Well ... It's really nice to see some great demos on the A8... All demos from QuaST 04 are great, but they all have some lacking issues... 1. They are too big for a 64k machine. So the demos are either 320k big or 1K (4K) intros. Even 128K is to big,when someone wants to show, what a A8 can do. My suggestion for the coders: Try to search for creating demos that exactly fit into a 64K machine... 2. The sound is nothing comparable to any other system at the demo scene. Actually... the main song of "Reditus" is unlistenable out of tune... it seems cheap automated MOD converting isn't the way Heck... I wish, the softsynth could do while the demo is running, but this will ever be impossible, when you want relieable sounds... So I've appended a Zip with Testsongs and Conversions from me past... you may still know it, or it is overall new for you. The tunes are built sounding almost different from standard POKEY sounds. Please take note, I am no Musician in profession, and you might do them better.... so please keep it up So while the tunes may not fit into your taste of music, those are the styles, POKEY HAS to sound for Demos... to be recognized in the Demoscene. Quote Link to comment Share on other sites More sharing options...
Cybernoid Posted August 18, 2004 Share Posted August 18, 2004 Thanks for these emkay! These sounds awesome with the current tools (ie - RMT). You are right, we need to be doing somthing differently for sound creation. Do not misunderstand me. I think that RMT and the SID player are great. I also I think it would be great if we had a tool that took just a raw sound track file, analyzed it, then used all 4 voices to recreate it as closely as possible. I think some of the problem now is the destructive interference between the voices and the limited 4bit PCM. The destructive interference is caused by the output from each voice competing for the speaker, and the 4bit PCM would cause quantization errors. Both cause unwanted effects. Now, maybe those with more POKEY coding experience than myself understand these limitations and can compensate. However, I think a tool that looked at a simple 16bit mono wave file, then output something to control all 4 voices to recreate the waveform would be very useful. I do not want to see just a 4bit PCM approximation either, but something that uses PCM (volumn only) and other frequencies/volumns to approximate the signal. Maybe there could be controls for how much CPU time to use either. I know this is contrary a little bit against the "tracker" mentality, but would be fun to play with... Time to break out Matlab... Just thinking out loud here. ~C Quote Link to comment Share on other sites More sharing options...
emkay Posted August 19, 2004 Author Share Posted August 19, 2004 Thanks for these emkay! These sounds awesome with the current tools (ie - RMT). You are right, we need to be doing somthing differently for sound creation. Do not misunderstand me. I think that RMT and the SID player are great. Same here.... Without RMT I would never have thought of creating this stuff. But, you know the issues. I think some of the problem now is the destructive interference between the voices and the limited 4bit PCM. The destructive interference is caused by the output from each voice competing for the speaker, and the 4bit PCM would cause quantization errors. Both cause unwanted effects. There are already existing tunes with combined generator-sounding and samples for the "main voice" (Blood fighter, Fluid Kha, ...) and it works. But the samples are not really used as a "softsynth"; they are simply played as samples. As the Sid-player does great, with included two 16 bit pokey channels the cpu usage would be reduced. In addition it would be possible to manipulate the generators in realtime. The main problem is, that when creating music inside an emulation, the timing has to be 100% accurate. And... may the "Lord above us" send a fully functional POKEY emulation I do not want to see just a 4bit PCM approximation either, but something that uses PCM (volumn only) and other frequencies/volumns to approximate the signal. Maybe there could be controls for how much CPU time to use either. actually... the controls inside a game or a demo still have to be set to 1 x VBI when you want to "move" some graphics .-) Quote Link to comment Share on other sites More sharing options...
Jaskier Posted August 19, 2004 Share Posted August 19, 2004 And... may the "Lord above us" send a fully functional POKEY emulation What in the actual emulation doesn't work correctly? Except that the pokey is updated every scanline. Quote Link to comment Share on other sites More sharing options...
emkay Posted August 20, 2004 Author Share Posted August 20, 2004 And... may the "Lord above us" send a fully functional POKEY emulation What in the actual emulation doesn't work correctly? Except that the pokey is updated every scanline. Just create a tune in RMT... Play it on the "old" 800 win, it will nearly sound the same. Play it on the "new" 800 win, it will sound more like the original hardware. Play it on the SAP Player 1, 79MHz is mostly not played, played wrong, or the player "makes a gosub without return" ... it hangs. 1,79MHz with the generators "2" and "8" seems to be played correctly at least. Playing the tune in the Deli-Player so every "out of standard" sounding will become a "new face" Making a comparision of the RMT used Libraries, several generators do not sound similar to the real hardware. The 1,79MHz filter is still buggy and generators "C" and "E" are partially wrong. And ...( I am not shure if it is in the emulation or in the RMT runtime... ) , normally POKEY is resetting registers when starting from volume level "0", so you can remove unwanted flanger effects or "not played" notes. This seems not to work as it should have to work. It seems, when playing a note with volume level "0" the song is really playing nothing instead of resetting registers. Actually the 800 win+ 4 emulation is sounding very close to the real hardware, but the "color" isn't the same. And the handling with the registers still has to be optimized which is the most important issue. Quote Link to comment Share on other sites More sharing options...
Cybernoid Posted August 20, 2004 Share Posted August 20, 2004 I have a question a little bit off topic, but... How does the audio output work if you set one voice to volume only and to volume 4. Then, you set another voice to some frequency and to a volume of 5. Do the two average together to get a wave that is instead of being 5, 0, 5, 0, 5, 0.... (voice 2 only) is rather 4.5, 2, 4.5, 2, 4.5, 2? where 4.5 = (5+4)/2 and 2 = (4+0)/2 Just thinking about Pokey... ~C Quote Link to comment Share on other sites More sharing options...
Saint (Swiety) Posted August 21, 2004 Share Posted August 21, 2004 At sid player i use volume table where center is at 8 value - volume 4 is 6 and 10 (at square wave) -> (8+2,8-2)! Quote Link to comment Share on other sites More sharing options...
Like us on Facebook
Follow us on Twitter
AtariAge on Instagram
Watch us on YouTube
Chat on Discord
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.