Spirantho's Adventures with SID Chips

What this page is...

SID chips come in many forms... there are different revisions, different manufacturing processes, different factories in different countries... There is no hard and fast rule as to what is "correct" in a SID chip.
Using X-Out as a test case, which is renowned for its heavy use of filters, this webpage shows some of the different kinds of SID chips that are available! It does not intend to demonstrate what a faulty SID chip sounds like (though some are included so that you know what to look for when testing SID chips for faulty filters.

"Fake" SID chips

There are many SID chips on the market that are not fully working. SID chips should only be bought from a verified seller who offers a full guarantee, and who is willing to test fully the SID chips they sell (I myself sell some on my shop). Trying to buy one cheap will almost certainly result in losing money as the vast majority of the SID chips available (especially from China) are faulty.
It is not known where the faulty SID chips have come from - some may be factory rejects, but if this were the case they should all be in batches with the same date-code and factory, and this is unusual. It is more likely that the faulty SID chips either came from faulty Commodore 64s (in which case there should be a mix of working and faulty SID chips), or that they came from working Commodore 64s, but were not stored properly and so were damaged - SID chips are very sensitive to static electricity and must be looked after.
A common misconception is that a chip that does not have its original markings much be fake. This is not always the case - in China it is normal to clean and remark chips before selling them. A remarked chip may be completely genuine - or it may be a completely different chip. The only sure way to test is to try it.
Buying SID chips is a harsh lesson in caveat emptor - Buyer Beware.

The Test System

All these recordings are done using the monitor output of a "breadbin" C64, recorded using Audacity on a standard Windows PC. They are saved as FLACs for maximum accuracy.

Significant points in the music stream

The recorded clips comprise the first 3 minutes of the introductory video of X-Out. This uses filters particularly at three points:
  • Directly after the initial fade-in. The filter is used for the speech "Get Ready For Cross Out". If this speech is not present, your filter is defective. However, speech being present does not necessarily indicate a good filter.
  • At 1:16, the tune line is made by a heavily distorted vibrato synth instrument. If the filter is bad, this sounds more like a staccato flute.
  • At 2:37 the melody is again carried by a distorted waveform. This sounds sharper when the filters are working correctly.


    • Below are three different SID chips: the first is a working revision 2 6581 chip made in Korea in week 45, 1982.
    Good speech waveform

    Next is a CSG 6581R4AR made in week 5, 1991, in Hong Kong. The speech is present, but much quieter than it should be.
    Silent speech waveform

    Lastly is a Chinese remarked SID chip, claiming to be a CSG 6581R4AR made in week 32, 1990. There is no manufacturer stamp on this so it's impossible to say from where it originated, but it's likely to be from Hong Kong.
    Distorted speech waveform


    Following are some more SID chip recordings of the same music:
    SIDBench MOS 6581 (r2?) Week 45 1982 Korea AM3827?? -
    SIDBench MOS 6581R3 Week 42 1985 Korea AH391858 -
    SIDBench MOS 6581 (r3?) Week 5 1986 Hong Kong HH0218?? -
    SIDBench MOS 6581R3 Week 10 1986 Philippines IH084820 SIDBench shows distortion on the Saw+Pulse filter.
    SIDBench MOS 6581R4 Week 16 1986 Hong Kong HH131110 Voice very quiet. "Staccato flute" at 1:16. SIDBench shows distortion on Saw+Pulse filter.
    SIDBench MOS 6581R4 Week 34 1986 Philippines IH332103 Voice very quiet. "Staccato flute" at 1:16.
    - MOS 6581R4 Week 33 1986 Philippines BH302113 -
    - MOS 6581R4 Week 33 1986 Philippines BH302113 -
    SIDBench MOS 6581R4 Week 33 1986 Philippines IH312139 -
    SIDBench MOS 6581R4 Week 33 1986 Philippines IH312142 Major distortion in the Saw+Pulse filter.
    SIDBench MOS 6581R4 Week 35 1986 Philippines IH342122 No problems apparent in SIDBench, but has a "staccato flute" in X-Out.
    - MOS 6581R4 Week 36 1986 Philippines IH352107 -
    SIDBench MOS 6581R4 Week 6 1987 Philippines IH03510 SidBench shows slight distortion in the Saw+Pulse filter.
    - MOS 6581R4AR Week 30 1986 Philippines BH282139 -
    SIDBench MOS 6581R4AR Week 30 1986 Philippines ?????? SIDBench shows slight distortion in Saw+Pulse filter.
    - MOS 6581R4AR Week 35 1986 Hong Kong HH332126 -
    - MOS 6581R4AR Week 35 1986 Hong Kong HH342160 Filters seem to be basically working (speech is good) but the second stage sounds rather too much like a staccato flute.
    SIDBench CSG 6581R4AR Week 5 1991 Hong Kong HH022127 -
    - CSG 6581R4AR Week 5 1991 Hong Kong HH022127 -
    SIDBench MOS 6581R4AR Week 44 1986 Hong Kong None Remarked. I don't know when or where this chip was actually made but it sounds very distinctive, with a very strong filter.
    - MOS 6581R4AR Week 44 1986 Hong Kong None Remarked.
    SIDBench MOS 6581R4AR Week 44 1986 Hong Kong None Remarked.
    - CSG 6581R4AR Allegedly Week 32 1990 None None Missing its filter, it seems.
    SIDBench CSG 6581R4AR Allegedly Week 32 1990 None None Good but shows distortion in SIDBench.