Log inRegister an accountBrowse CSDbHelp & documentationFacts & StatisticsThe forumsAvailable RSS-feeds on CSDbSupport CSDb Commodore 64 Scene Database
You are not logged in - nap
CSDb User Forums


Forums > C64 Composing > ReSid default settings in different players / emulators
2019-10-16 12:53
Flotsam

Registered: Jan 2004
Posts: 84
ReSid default settings in different players / emulators

Hi all, I started working on a filter heavy song targeting PAL / 8580 in GoatTracker. Exporting it to a .sid and playing it in VLC player sounds almost identical to how it sounds in GT. In Vice 3.1 the filters sound too bright and the default setting for filter bias is +500. In GTK3Vice 3.3 on the other hand the sound is SUPER muddy and it's no surprise as the default filter bias there is -3000(anyone know why such radical setting?).

So, what should I use as a reference? It seems the 8580 filters in emulation domain are more unpredictable than on real machines, which is odd.


Edit: It seems that the version of GTK3Vice that I just downloaded doesn't respect the filter bias setting at all. The sound remains the same no matter which value the bias slider is set to. :-/
2019-10-16 14:16
Zirias

Registered: Jan 2014
Posts: 48
Quoting Flotsam
GTK3Vice 3.3 on the other hand the sound is SUPER muddy and it's no surprise as the default filter bias there is -3000(anyone know why such radical setting?).

I was wondering the same thing .. set it to 500 here which sounds somewhat like what I expect.

Quoting Flotsam
So, what should I use as a reference?

Real hardware of course ;)
Or maybe jsidplay2 which is, at least for my ears, pretty close to real hardware.

Quoting Flotsam
Edit: It seems that the version of GTK3Vice that I just downloaded doesn't respect the filter bias setting at all. The sound remains the same no matter which value the bias slider is set to. :-/

I never had something like that. Did you make sure you use the correct SID model? Maybe you also need to save settings and restart, which might depend on the OS you're running on ... btw, Vice in the same version sounds very different here when running on my FreeBSD 12 desktop as compared to my work notebook with Windows 10.

And a word about goattracker: It almost always sounds a bit different. I suspect this might have to do with timing differences (among other factors)...
2019-10-16 15:08
Flotsam

Registered: Jan 2004
Posts: 84
I was doing tests specifically on default settings of different emulators and players (including Vice 2.4, 3.1, 3.3), but yeah, I did check that I have the correct sid emulation in 3.3. I also tried manual saving, saving on exit, restarting, but none changed the sound in anyway. Also the UI slider behaves strangely, the minimum skip is 1000, so I can only choose values -3000, -2000, -1000, etc. I'll try manually changing the config file later, but right now getting 3.3. to work is not my first priority.

I googled some more and found a thread where someone quotes LMan saying that -2500 would be a good setting for filter bias: Vice reSID filter bias
To my ears, -2000 is about right (tested in Vice 2.4 & 3.1).

JSIDPLay2 like all the others make use of ReSID, so I believe that considering 8580 the only difference is what settings are used in each player / emulator.

BTW, using hardware is not an option... my C128D has a revision of 6581 that differs wildly from any other C64/C128 I've ever owned or heard anywhere. That's why some of my songs from the mid 90's sound like crap. :)
2019-10-16 15:59
Zirias

Registered: Jan 2014
Posts: 48
Quoting Flotsam
JSIDPLay2 like all the others make use of ReSID, so I believe that considering 8580 the only difference is what settings are used in each player / emulator.

AFAIK, resid-fp is used by default in jsidplay2, so this sounds a bit different.
2019-10-16 16:00
McMeatLoaf

Registered: Jan 2005
Posts: 108
Quote: I was doing tests specifically on default settings of different emulators and players (including Vice 2.4, 3.1, 3.3), but yeah, I did check that I have the correct sid emulation in 3.3. I also tried manual saving, saving on exit, restarting, but none changed the sound in anyway. Also the UI slider behaves strangely, the minimum skip is 1000, so I can only choose values -3000, -2000, -1000, etc. I'll try manually changing the config file later, but right now getting 3.3. to work is not my first priority.

I googled some more and found a thread where someone quotes LMan saying that -2500 would be a good setting for filter bias: Vice reSID filter bias
To my ears, -2000 is about right (tested in Vice 2.4 & 3.1).

JSIDPLay2 like all the others make use of ReSID, so I believe that considering 8580 the only difference is what settings are used in each player / emulator.

BTW, using hardware is not an option... my C128D has a revision of 6581 that differs wildly from any other C64/C128 I've ever owned or heard anywhere. That's why some of my songs from the mid 90's sound like crap. :)


https://anorien.csc.warwick.ac.uk/mirrors/oakvalley/soasc/hvsc/..
https://anorien.csc.warwick.ac.uk/mirrors/oakvalley/soasc/hvsc/..
The old SOASC recordings, done on 6581, R4 and R3 (not R2 as in the filename).
Guessing the lower recording (R3) is closer to your C128D :)
There was a reSID 1.0-fork of sidplay2w a few years ago where you could change the 6581 filter bias (0-100) on the fly. (for the recordings above: R3 at ~22, R4 at ~82)
2019-10-16 22:15
spider-j

Registered: Oct 2004
Posts: 498
The most accurate 8580 emulations today are:

- current ReSID in recent VICE versions which are compiled with the (imho "experimental") *new* 8580 filters option deactivated

- current (lib)sidplayfp; Sidplayfp V2.0.1

As I wrote in my comment in the latter: do not use older versions (i.e. VICE 2.x or whatever libsidplayfp version was used here Sidplay/w 2.5) if you want ringmod and sync work properly.

P.S.: but do yourself a favor and get a real machine with 8580 anyway ;-)
2019-10-17 08:18
Flotsam

Registered: Jan 2004
Posts: 84
spider-j, thanks for the headsup. With default options the sidplayfp-2.0.1 sounds close enough to GoatTracker, which is a relief. The biggest difference in my current song is white noise, which on lower frequencies sounds quite different. Filter bias is slightly lower, I'd say about 500 units (what ever the unit is). I'll try fiddle with GoatTracker options to get the filter closer to sidplayfp.

I believe the windows version of GTK3Vice I downloaded yesterday must be broken, will see later if there's a fix for that. I think the filter bias is stuck at lowest possible setting.

McMeatLoaf, yeah, my chunky D has the R3. I'll sell it as soon as I've saved all my old shit from the disks. Maybe I'll get a 8580 machine once I've gotten rid of the D.

Thanks for all the input. Now back to GoatTracker! :)
2019-10-17 21:44
Compyx

Registered: Jan 2005
Posts: 631
I just checked GTK3VICE-3.3 (trunk), and indeed the ReSID sliders are broken. Someone (*cough*) probably broke them when adding HardSID support to the UI.

I'll add a ticket and hopefully get around to fixing it this weekend.

Thanks for the heads up.


Edit: bug report is here: https://sourceforge.net/p/vice-emu/bugs/1155/
Feel free to add information.
2019-10-22 18:45
Compyx

Registered: Jan 2005
Posts: 631
ReSID filter settings sliders work properly now. Meaning they no longer make huge steps but can be altered with a step of 1.
Using the mouse one can make large changes, using the keyboard (left/right) one can make small changes. If required I can add spinboxes to allow total control, but see the below:

Keep in mind the 'old' 8580 filter emulation (same as 2.4.x) won't respond to the sliders, for that you'll need the new filters.
So compile trunk with --enable-new8580filter to enable the updated 8580 emulation with all its "farting" sound galore and perhaps suggest some improvements. Preferably not "this sucks", but rather "set bias/gain/passband to ..." or perhaps even patches? (I know, too much).
2019-10-22 20:22
Flotsam

Registered: Jan 2004
Posts: 84
Compyx, I only have one song at hand that I know well enough to trust my ears and that I can compare to a recording from a real 8580R5, so this is very limited, but I would say the correct filter bias is somewhere between 4000 and 4500, which is faaaaar away from the current default setting of -3000. Must say, with filter bias set to 4500, which I slightly prefer over 4000, the sound with it's subtle distortion and authentic analog-digital feel is really sweet. Much appreciated!
2019-10-27 16:51
Compyx

Registered: Jan 2005
Posts: 631
Thanks for the info. Is this song you're referring to in the HVSC, or is it something private/unpublished?
 
... 1 post hidden. Click here to view all posts....
 
Previous - 1 | 2 - Next
RefreshSubscribe to this thread:

You need to be logged in to post in the forum.

Search the forum:
Search   for   in  
All times are CET.
Search CSDb
Advanced
Users Online
REBEL 1/HF
Grue/Extend
Fred/Channel 4
Digger/Elysium
Guests online: 111
Top Demos
1 Next Level  (9.7)
2 13:37  (9.7)
3 Mojo  (9.7)
4 Coma Light 13  (9.6)
5 Edge of Disgrace  (9.6)
6 What Is The Matrix 2  (9.6)
7 The Demo Coder  (9.6)
8 Uncensored  (9.6)
9 Comaland 100%  (9.6)
10 Wonderland XIV  (9.6)
Top onefile Demos
1 Layers  (9.6)
2 No Listen  (9.6)
3 Party Elk 2  (9.6)
4 Cubic Dream  (9.6)
5 Copper Booze  (9.6)
6 Rainbow Connection  (9.5)
7 Dawnfall V1.1  (9.5)
8 Onscreen 5k  (9.5)
9 Morph  (9.5)
10 Libertongo  (9.5)
Top Groups
1 Performers  (9.3)
2 Booze Design  (9.3)
3 Oxyron  (9.3)
4 Triad  (9.3)
5 Censor Design  (9.3)
Top Graphicians
1 Mirage  (9.8)
2 Archmage  (9.7)
3 Pal  (9.6)
4 Carrion  (9.6)
5 Sulevi  (9.6)

Home - Disclaimer
Copyright © No Name 2001-2024
Page generated in: 0.053 sec.