| |
Steppe
Registered: Jan 2002
Posts: 1510 |
Release id #151275 : Wonderland XIII
In the production notes Bob writes:
"There is an hidden feature in the demo to determine the sample playback quality on your SID chip. Hold down fire on joystick in port two while the demo is starting and you will get a border color. The color means:
Green = Best (8580)
Yellow = Good (8580)
Red = Bad (8580)
Gray = Unknown (8580)
Dark gray = Bad (6581)
Black = Awful (6581)"
I'm using Sidfx with a 6581 and 8580 and no matter which model I select, I always get the "black / awful" screen and this shows in the sound quality, really noisy...
To double check, I tried the demo on my C128D-CR (metal case with 8580) and I get a yellow border and the sound quality is really a lot better, a whole different league! Cheers for making the demo work on my old metal 128!
Bob, can you explain that border coulor check? What does it do and why doesn't it work properly with the Sidfx? |
|
| |
Steppe
Registered: Jan 2002
Posts: 1510 |
Is it related to the Sidfx at all? Can some of you Sidfx users try this on your machines? |
| |
CreaMD
Registered: Dec 2001
Posts: 3035 |
Steppe try on plain C64 before you ask... ;-) |
| |
Steppe
Registered: Jan 2002
Posts: 1510 |
Roman, you probably misunderstood. I'm not saying there's something wrong with WL13, but it's the hardware extension Sidfx that's causing unwanted trouble. Of course I'm using it in a real C64 and also compared the strange result with an "untouched" machine, my 128D.
The Sidfx guys meanwhile cleared it up a bit. A SID detection routine will always detect the SID in socket 1, no matter which position the SID select switch is in. So I swapped the SIDs (#1: 8580 and #2 6581R3) and I'm now getting a light gray border with brutally improved sound quality in the end part. So you can consider the riddle solved.
Still, would be nice to know some technical stuff behind these various quality levels that are possible, depending on the chip type and probably other frame conditions. I mean, an 8580 is an 8580, but there are three (four) levels of quality possible. What is it that gives me green, yellow, red, or in my case, gray? |
| |
Swallow
Registered: Aug 2013
Posts: 4 |
The routine first does a model detection and then checks the ability to use combined waveforms for sample playback on your chip.
It is interesting that you get an unknown configuration. Do you have the ability to test another detection routine if I send one to you? |
| |
Steppe
Registered: Jan 2002
Posts: 1510 |
Sure, always happy to help out Swallow. |
| |
Frantic
Registered: Mar 2003
Posts: 1627 |
Quote:Still, would be nice to know some technical stuff behind these various quality levels that are possible, depending on the chip type and probably other frame conditions. I mean, an 8580 is an 8580, but there are three (four) levels of quality possible. What is it that gives me green, yellow, red, or in my case, gray?
Yes, this made me curious as well. Swallow? Different 8580 chips play combined waveforms in different ways, or what? |
| |
Swallow
Registered: Aug 2013
Posts: 4 |
The routine uses different combined waveforms, eg. combinations of pulse, sawtooth and triangle, to play the sample. The SID spec states that if more than one waveform is selected, the output will be an logical and between the selected waveforms but this is not true. Different chips generate different output when combining waveforms, even chips with the same model. The resolution of the samples depend on how many combined waveforms that produce useful output on your chip. |
| |
Frantic
Registered: Mar 2003
Posts: 1627 |
Quote:even chips with the same model
A-ha! Thanks for that information. I wasn't aware of that. I thought 8580 was 8580... so to speak. (In contrast to all the revisions of 6581). |
| |
lft
Registered: Jul 2007
Posts: 369 |
Isn't that also temperature dependent? Meaning that you have to assume that the machine has been on for a while. Or is that a myth? |
| |
ruk
Registered: Jan 2012
Posts: 43 |
New parameter for VICE
x64sc -temperature 37.0
:D |
| |
CreaMD
Registered: Dec 2001
Posts: 3035 |
Quote: Roman, you probably misunderstood. I'm not saying there's something wrong with WL13, but it's the hardware extension Sidfx that's causing unwanted trouble. Of course I'm using it in a real C64 and also compared the strange result with an "untouched" machine, my 128D.
The Sidfx guys meanwhile cleared it up a bit. A SID detection routine will always detect the SID in socket 1, no matter which position the SID select switch is in. So I swapped the SIDs (#1: 8580 and #2 6581R3) and I'm now getting a light gray border with brutally improved sound quality in the end part. So you can consider the riddle solved.
Still, would be nice to know some technical stuff behind these various quality levels that are possible, depending on the chip type and probably other frame conditions. I mean, an 8580 is an 8580, but there are three (four) levels of quality possible. What is it that gives me green, yellow, red, or in my case, gray?
Sorry I thought you only tried with sidfx machine installed. Anyway if you switched addressing mode with config tool, Wonderland's detection routine would ignore it and still check sid in socket I? that's really interesting. |
| |
Frantic
Registered: Mar 2003
Posts: 1627 |
@CreaMD: I think you misunderstood. The detection routine in Wonderland XIII does what every detection routine would do when the SIDFX was configured like it was in this case. See Lotus' explanation in the SIDFX thread for how to set it up to allow the SID chip to be properly detected. |
| |
Swallow
Registered: Aug 2013
Posts: 4 |
@lft: I have seen indications that it is probably not a myth but it is kind of hard to prove. ;) |
| |
Steppe
Registered: Jan 2002
Posts: 1510 |
Swallow, release your SID test and have it record in certain time intervals, so it should be easy to spot if stats are changing over time and with rising temperatures. |
| |
Xiny6581
Registered: Feb 2004
Posts: 72 |
I can give this a try on my two SIDFX which are configured a bit different with different SID models.
1. 6581R4AR + 8580R5 (using correct setup to get each chip autodetected depending on what model is switched)
2. 8580R5 + 8580R5.
If this is any help :) |
| |
Xiny6581
Registered: Feb 2004
Posts: 72 |
Quote: I can give this a try on my two SIDFX which are configured a bit different with different SID models.
1. 6581R4AR + 8580R5 (using correct setup to get each chip autodetected depending on what model is switched)
2. 8580R5 + 8580R5.
If this is any help :)
Okay so it turned out that the machine with 6581 and 8580 gave me a yellow screen (running in 8580 mode)
The other machine with two 8580 have me both green and yellow (SID 1 and SID 2).
Kind of cool and I could hear some differences, actually.
But no major that gave me a "WOOHOO I just sh.. my pants" :) |
| |
Swallow
Registered: Aug 2013
Posts: 4 |
Haha, no the difference is not within the sh.. your pants range, that is why I called them Best and Good. ;) |
| |
Oswald
Registered: Apr 2002
Posts: 5020 |
someone can test this ?
https://twitter.com/polyp2000/status/803869755396517888
"anyone know whats up with the flashing squares of color on the scrolling #c64 #robot #lady in this demo? C64reloaded + U1541 II+ & 6581 #SID"
is it reloaded or 1541u ? |
| |
Grue
Registered: Dec 2001
Posts: 146 |
Looks like color ram issues I had with faulty power switch on my c64.
Faulty switch caused noise on +5v voltage line which caused color ram to fail randomly. For example I remember lots of color ram problems on impossible mission elevator scrolling. |
