Article 499 search instead | all articles | select by topic ------------------------------------------------------------------------ select article number: ------------------------------------------------------------------------ ------------------------------------------------------------------------ \nFrom root Mon Jul 6 07:13:42 1998 Received: from uhura.concentric.net by oa.ptloma.edu (AIX 4.1/UCB 5.64/4.03) id AA15972; Mon, 6 Jul 1998 07:13:41 -0700 From: Cdkaiser@concentric.net Received: from galileo.cris.com (galileo [206.173.119.84]) by uhura.concentric.net (8.8.8/(98/05/18 5.10)) id KAA21895; Mon, 6 Jul 1998 10:10:09 -0400 (EDT) [1-800-745-2747 The Concentric Network] Errors-To: Received: by galileo.cris.com (8.8.5) id KAA17139; Mon, 6 Jul 1998 10:10:09 -04 Date: Mon, 6 Jul 1998 10:10:09 -0400 (EDT) Message-Id: <199807061410.KAA17139@galileo.cris.com> To: ckaiser@oa.ptloma.edu Subject: Q: F1 fuse replacement / SID silence Status: RO >From: John Iannetta <76703.4244@CompuServe.COM> >Subject: Re: Q: F1 fuse replacement / SID silence >References: Organization: CompuServe, Inc. (1-800-689-0736) >Message-ID: >Newsgroups: comp.sys.cbm >Date: Mon, 06 Jul 1998 06:24:34 -0400 >Path: news!global-news-master!newsfeed.concentric.net!ais.net!jamie!WCG!arl >Lines: 60 >Xref: news comp.sys.cbm:97759 Nate Dannenberg said (replying to my question about why he considers the 8580 to be inferior to the 6581 for playing digitized sound), "Because on some models of the 8580, there is NO output when trying to play a $D418 digi using a routine normally written for the 6581. Why? Because the 681 has a small voltage leak inside it, which happens to leak through the volume control, giving it the ability to play digis. In the 8580, the leak was eliminated (or so they intended).". I've seen you say similar things several times, and I think that I see where the dichotomy lies. Please tell me whether I'm on the right track. I was using a C-64C (with an 8580 SID) when I first experimented with digitized sound. I fed SID register $18 with an offset voltage from all three voices, to play back digitized sound. In discussing my record/playback system with me, Gary Farmaner mentioned that a well-known programmer was playing digitized sound files by using a defect in the 6581, and that the 8580 didn't play the files very well. I suspect that others may have used his method, and that that is what you had in mind. Nate also said, "The software patch involves creating a voltage leak with voice 1's pulse output function. You will find there are programs that support this patch, both of mine (Sound Studio and Modplay 64/128) included.". That pretty much confirms my suspicion. There are three methods that I have tried for playing digitized sound data; I use all three voices, tripling the audio output. Considering voice 1, $F0 is first written to register $06 (using all three methods). In the first method, $01 is written to $04, all other registers are "don't care". That sets the gate bit, with no waveform selected. After the sustain level is quickly reached, a negative offset will be fed to register $18. The second method writes $41 to $04, and $FF to $02 and $03; all other registers are "don't care". That too feeds a negative offset to register $18. The third method writes $41 to $04, and $00 to $02 and $03; all other registers are "don't care". That feeds a positive offset to register $18. Ideally, all three methods produce the same audio output. An 8580 should approach that. The 6581's that I have checked seem to have incorrect biases in the register $24 circuitry, giving unexpected results. And there is considerable variation from device to device. My best results have been with the third method. Nate concluded, '> "ELLAFITZ.BIN" was written on a C-64C using an 8580. 'Where can I get this program? (And for that matter, what is it?)'. It is a C-64 program that plays 19.6 seconds of Ella Fitzgerald singing "Goodnight my Love" with Benny Goodman's orchestra (recorded November 5, 1936). The file was in comp.binaries.cbm, not too long ago. I have sent it to you, UUencoded. But since I use CompuServe for E-mail, it is fairly likely that you will get a corrupted version of it. You might even wind up getting Bil Herd's GIF of Jack Tramiel. My apologies if the file doesn't get through intact. -- When backing up your hard drive, shift into reverse gear S M O O T H L Y. John >Path: news!global-news-master!newsfeed.concentric.net!nntp-out.monmouth.com >From: Nate Dannenberg >Newsgroups: comp.sys.cbm >Subject: Re: Q: F1 fuse replacement / SID silence >Date: Mon, 6 Jul 1998 08:14:13 -0500 >Organization: SouthWind Internet Access, Inc. >Lines: 50 >Message-ID: References: NNTP-Posting-Host: onyx.southwind.net >Mime-Version: 1.0 >Content-Type: TEXT/PLAIN; charset=ISO-8859-1 >Content-Transfer-Encoding: 8BIT >In-Reply-To: >Xref: news comp.sys.cbm:97766 > sound files by using a defect in the 6581, and that the 8580 didn't play t > files very well. I suspect that others may have used his method, and that > is what you had in mind. Correct. For me, it came about by plugging a few numbers into $D418, and noticing that it made my speakers pop. I recall asking on the Fido-Net "CBM" conference about this phenomenon, to which I got a number of replies about the digi capability of the 64. Once I learned what type of data it needed (raw, unsigned, 4-bit), it was a matter of time before I wrote a digiplayer, which because Sound Studio. I recall Russell Prater on that group telling me about the 8580's deficiency in this department, and the fix mentioned below, which I've since incorporated into Sound Studio and Modplay 64/128, as well as any future digi apps that I'll write. > Ideally, all three methods produce the same audio output. An 8580 sho > approach that. The 6581's that I have checked seem to have incorrect bias Correct - one of these methods (the second one, IIRC) will cause a 6581 to output a slightly lower volume than if you simply write $00's thru the entire chip before playing. > The file was in comp.binaries.cbm, not too long ago. I have sent it to > you, UUencoded. But since I use CompuServe for E-mail, it is fairly > likely that you will get a corrupted version of it. You might even wind > up getting Bil Herd's GIF of Jack Tramiel. My apologies if the file > doesn't get through intact. I got it, sounds pretty good. I'll give this a run on the stock 128-DCR later as well (this one has 8580's, even tho I said that it has 6581's... that was a tired brain talking :) Interesting - gimme that GIF while your at it :) _/ _/ _/ _/_/ _/ _/_/ _/_/_/ _/_/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/_/_/_/ _/ _/_/_/_/ _/ _/_/ _/ _/ _/ _/ _/_/_/_/_/ _/ _/ _/_/_/ _/ _/_/ _/_/_/ _/_/_/_/_/ | GCS d- s++:++ a-- C++ UB>++ L>+ E- W++ N++ K- w--- M- V- PS | \ PE Y+ PGP- t+ 5 X+ R tv+(++) b+ DI(+) D+ G e+ h+ r- y- / search instead | all articles | back to the top of this article All articles are the intellectual property of the original author and any/all copyrights within this article lie with the original author. All other content and software is copyright ©1998 Cameron Kaiser. All rights reserved. No warranty is expressed or implied concerning the veracity, completeness or bias of data expressed in this or any article.