FPGASID vs. SID8580 vs. SID6581

So, here it is. The FPGASID has arrived and you should know what's coming up now:

Yes, exactly. A comparison, a test, audio examples, a lot of audio to listen to and some of text to read.

First of all, I have to mention that I listened some examples already available on the FPGASID page. And I was sad because while the audio sounds good the examples are well-known game tunes from the 80s. As if there was no life after 1990. As if there are no hundreds of thousands of great C64 music out there.
I understand that many people grew up with playing those good old games, listening to the awesome soundtracks. And only a small percentage of them are aware of the fact that demoscene does exist and also the chiptune scene keeps the C64 music alive.
Yes, old game tunes are gold, their composers were the pioneers back then.
But there are many more pioneers who experienced with the SID, created music, exploited many features and discovered undiscovered things in the last 25+ years too...
These people should be noted or at least mentioned as well, so as usual, my test is not about old game music but "new" demoscene tracks and composers.

During testing, I was thinking what to post about the FPGASID. It's actually hard to make a proper blogpost about it, it's not easy to collect and record music and arrange them to a nice post.

Especially hard to do it, because this replacement of the original SID is so good that it's very hard to find tunes that break the emulation.

Waveform and filter reproduction in 8580 mode is so good that majority of the test music were simply identical on both FPGASID and SID8580. Ringmod, sync and many other tricks and effects are almost perfect at all times.

The "filter bias" setting in 6581 mode is a great feature to simulate various 6581 revisions and it's very easy to find a satisfying setting for filter strength. To me "bias 0" sounded the best, I used this value for recording all 6581 music in this post.

Let's bring up my favorite test tracks and see how they sound.

• Left channel: FPGASID
• Right channel: SID8580
• no volume changes, no additional effects were used
• recorded directly from C64 output, without using additional cables of the FPGASID, so it acts like a single SID
  

01. Jeff - Analogue
This is a music written on 6581 SID. However, I recorded it in 8580 mode and on 8580 SID for this test. The reason is filter bias. Everyone can finetune it on the FPGASID to match their original SID6581 sound. To me, it wasn't that important here, the shape of the waveforms and the sound is more important.

Obvious observation 1: the music is almost completely mono. Which means, the sound is almost identical on both chips.

Obvious observation 2: there's minor difference in the filter and obviously, in the noise waveform due to it's random nature.



02. Jammer - Kraut Pleaser
Jammer was kind to share his experience with me about a recently discovered (possible) ADSR bug in the SID. He mentioned that the following music is broken on the real SID but plays fine on emulation.

Listen closely, the tune is mostly mono due to the identical waveforms. However, a few notes are different on the real SID because of the ADSR bug.
I believe, because this bug is not documented, any emulator would miss it and play the music as FPGASID. No idea if this is something that's easy to fix but this might be the reason for the upcoming tracks to break too.



03. Jammer - Quickie Veto
Same applies to this music as well, ADSR bug and almost 100% mono playback.

04. Necropolo - Cadmium
The good old 8bit car engine sound. Lot of trickery with ringmod and waveform sync. Metal to the ped... SID metal at its best.

Minor differences here and there. Overall, it's a great playback, those minor differences wouldn't be noted if this is not a direct comparison.



05. LMan - My Life
This is a recent music from X2018 demoparty, based on samples but also using SID sounds.

Obviously, the sampled sounds are identical. However, there's a SID instrument that plays the melody, its release seems to be different.



06. Vincenzo - Freefall
I'm not saying I'm special but this music of mine is a special case.

I couldn't sync up the two recordings. There are significant waveform and tempo differences, so this audio clip doesn't sound mono at all. I'm not sure what's so different in this music compared to Jammer's trickery. Maybe the extensive use of ringmod or... no idea. This issue needs more time to find out what's going on.

On a sidenote, I might have also discovered a bug in SID-Wizard's exporter. This music sounds different when played back from SID-Wizard and played back as a C64 executable music.

There are obvious differences, not only in the sound but in the visual representation of the wave files too. This is actually a one-case scenario, so far I couldn't find another music that breaks the FPGASID this much. And to be honest this is not even a big issue because the FPGASID playback sounds good on its own as well.



Conclusion:
The FPGASID is one hell of a SID replacement. With some very minor adjustments it can be the current best option for SID enthusiasts. Emulation of 6581 filter is extensive, and easy to change its bias. Configuration tool is easy to use and straight forward.

Congratulations to Andi6510 and his team for achieving this great result.

I didn't test Stereo SID and Pseudo Stereo settings. I simply had no patience to setup the additional wires. It's actually one thing I should mention, the clamps are bigger than the holes between CPU pins and it's hard to connect.

There's a bit of noise in 8580 mode that doesn't exist in 6581. Hopefully this can be fixed with firmware update.

Also, it's an expensive replacement. Yes, it's great. It sounds great, it's feature rich. But... still. I'm happy to have one but I would think A LOT to get another for this same price.


FPGASID webpage: http://www.fpgasid.de/
FPGASID at Kryoflux: https://webstore.kryoflux.com/catalog/product_info.php?products_id=63


Instead of showing more individual music examples, have a listen to a mix I compiled. All were recorded with FPGASID, set to either 6581 (filter bias 0) or 8580 mode depending on SID model required by each tune.
Have fun.




Track list:
00:00 Swallow - Fantasmolityc Tune 2 (8580)
03:19 Danko - Wonderland X Intro (6581)
04:35 Jammer - Kraut Pleaser (8580)
05:34 Vincenzo - Freefall (8580)
08:57 Jeff - Analogue (6581)
10:52 Vincent Merken - Kkan (8580)
13:34 GRG - Ambient (6581)
14:41 Wiklund - Euro Trash (8580)
16:48 Metal - Phat Phunk (6581)
17:36 Drax - Neurosis (6581)
19:42 Chubrock - Man On Fire (8580)
21:24 DOS - Chubby Tune (8580)
23:58 LMan - Could Be Cool (8580)
27:16 Linus - Cauldron II (8580)

C64 C - new look with old board

Ever wondered how to identify the inside of a C64 by just looking at it from the outside?
I can highly recommend Ilesj's article to start with because it gives you nice examples with photos and explains the differences well.

I recently get my hands on a C64 C, it looks nice from outside, has just slight yellowing. Keys are stiff, on/off switch isn't lose so I was sure it will just simply work when turned on.

Oh yeah, it did! Nice red LED, old-style keys, C64 C is written on the back. I quickly loaded my favorite test track (Jammer - Club Style) to hear how the SID sings.

I immediately realized it's a 6581 because of the filtered sounds and the fact the track is kinda broken. I also made a guess that it's a 6581 R4.

Hint: read my previous articles about comparing different SID revisions, search for "C64 SID shootout" in this blog. Cheers.

Opened the case, removed the shiny alu-cover aaaaaand...!

It's an "old" board, Assy No. 250466. Youngest of the old boards. And have a closer look on the SID:

It's a 6581 R4AR aka. R4 Advanced Resonance. My guess was close enough, right?



.....



Today's music: BitJam Podcast #195 - Manifesto - Dane's Mix From X2014

C64 SID Shootout part 5. - SID 6581 vs. SwinSID Ultimate (in 6581 mode)

It seemed to be a good idea to give another round to the SwinSID Ultimate because, did you know it has 6581 mode as well? Yeah, this little gem is capable of emulating both 6581 and 8581 revisions. However, its main mode is 8580 - as CodeKiller explained it to me.
Also, 6581 mode is not complete yet as there were not enough memory to emulate filter distortion and some other bits as well. Let's see (hear) how accurate is SwinSID Ultimate's (SSU from now on) 6581 emulation.

Nope, I won't normalize the recorded audio files. Yes, I know the loudness level is different on 6581 and SSU. Still, in my opinion the true comparison is WhatYouSeeIsWhatYouHearWithoutAnyModificationHearingTrueOutput. That's it. You have volume control on your audio equipment, use it to compensate loudness diferences.

The following lines needed to switch the SSU to 6581 mode:
POKE 54272+29,ASC("S")
POKE 54272+30,ASC("E")
POKE 54272+31,ASC("6")
Reset C64.

To switch back to 8580 mode just simply write "8" instead of "6" into the third line. It's quite simple, isn't it?

I used Breadbin with the most loved SID 6581, the 1485 so the expectation for SSU is VERY high. Ladies and gentlemen, make your bets. Based on previous tests, what do you think? How SSU can meet the expectation?

Ready? 3... 2... 1...

Chris Hülsbeck - Hollywood Poker Pro main - SID 6581 1485

Good ol' Hollywood Poker with the dance animation scene..! (and more but that doesn't matter right now)


Chris Hülsbeck - Hollywood Poker Pro main - SwinSID Ultimate - 6581 mode

I'm going to repeat myself: impressive! The original SID is a bit brighter and the filter seems a bit more opened but SSU is great as well. Sounds like differences between 6581 revisions.


MSK - Land Of Illusion - 6581 1485

It's a multispeed track. I didn't dig into the code to check it but it's clearly audible, especially on some arpeggiated notes. Nice track, nice sound trickery.


MSK - Land Of Illusion - SwinSID Ultimate - 6581 mode

Wow, nice. Flawless multispeed reproduction, the only noticeable difference is the filter. Again, it's a minor thing, SSU's filter doesn't have distortion and it seems a bit more closed or deeper.


Drax - Drudgery - 6581 1485

This track is something special. Not just because it's beautiful and the bassline is extraordinary, but it also has nice filter variations between bass and arpeggios.


Drax - Drudgery - SwinSID Ultimate - 6581 mode

SSU did it again, same result as before: waveforms reproduced well, filter is a bit different but has that mojo. Actually, I like the deeper filter in this track more than in the 1485 version.


Jeff - Blowing - 6581 1485

This. Is. One. Of. The. Best. C64. Tracks. Ever. (personal taste of course, don't take it too seriously) Perfect dance music, awesome sounds, great melodies.


Jeff - Blowing - SwinSID Ultimate - 6581 mode

Again, question of taste matters. Which one do you prefer? The original with a bit more opened filter or SSU that is a bit deeper-boomier?


And now, here comes my favorite 6581-breaker - that is actually a music written on a 8580 SID. Guess how SSU delivers a 8580 track in 6581 mode. Will it break as it would on a real 6581? Or will it play perfectly, because this is an emulation?


Jammer - Club Stylier - SwinSID Ultimate - 6581 mode

Aaaaawyeah, it breaks as it should! :) Honestly, I expected it would play nicely but it didn't. It makes sense, SSU is switched to 6581 mode and a real 6581 breaks this track, so SSU did it as well.


.....


I wanted to write something nice as conclusion but I realized it's not necessary. The SwinSID Ultimate is the current best SID alternative if you want to replace your broken SID in the real C64. It doesn't matter which revision you prefer, SSU's performance is outstanding either in 6581 or 8580 mode.

Blu - Close to perfection

*I renamed my "Blue" C64 to "Blu". It's better now, right? Right??

.....

It's so wonderful how many options we have to ruin those good old computers. For example, expansion cards, extra RAM, extra CPU, etc. This time I decided to ruin Blu with an additional SID.

It's not really a new method to expand the sound capabilities of a C64 with another SID. DualSID, so called. Use 2 similar SIDs, set them to different memory address and that's it. If the software, for example a music editor supports this, you will have 6 independent SID channels which is great. It means more sounds, more opportunity to create great music on the '64.

The "Pseudio-Stereo SID" by Tynemouth Software is a bit different. The idea was to create a Stereo C64 using 2 SIDs, where the SIDs are set to the same memory address AND they play the same audio output AND they can be either 6581 or 8580 SIDs.

This way the C64 has still 3 channels but they are doubled and have separated audio outputs as well
I ordered my PCB for a C64C (SID 8580) with additional slot for a 6581.

You know me, I like to compare the SID revisions to each other and this would be a great way to listen to some music and record it simultaneously - or just simply compare the audio differences.

So, it has arrived. Nice and clean PCB, well organized and made. I wanted to use RCA connectors as they are easy to connect to any interface or mixer. I decide to drill holes on the left side of the case as there is enough room for them.

One RCA connector for 8580 (right side) and another for 6581 (left side). Maybe they are a bit too close to each other but well, it works fine and is not _that_ ugly.

The 2 best sounding SIDs (in my opinion) were plugged into the slots, 8580 2288 and 6581 1485. This 8580 is significantly louder but that can be compensated on my mixer easily.

I tested it with a 8580 3991 and that was on kinda the same loudness level as the 6581. Conclusion, there _is_ difference between 8580 models as well.

Anyway, I will keep 2288 and 1485 because I like them.

Blu is close to perfection. I still need two things to solve:

1. Blue keycaps, replacing the original white keys (Kickstarter is over, we are waiting for shipping)
2. Additional electronics to protect SID outputs (SID output is very sensitive and can break easily)

.....



Today's music: Chiptune Extravaganza @ Revision 2016 demoparty

C64 SID shootout part 3.

Page has been updated with a nice music player, shouldn't any playback issue appear anymore \o/

. .. ... .... ....... .... ... .. .

We used a music composed on 6581 SID for the previous test, so today we're gonna break the SID 6581 with a 8580 track. But first, let's start with a music that sounds kinda okayish on all SID.

Stellan 'Dane' Andersson - Break the Cycle. Played on a 8580 R5 first - as this supposed to be the SID revision the music was made on - then with 6581 revisions:

Clip 01. Dane - Break the Cycle - 8580 R5

It's a beautiful music, isn't it? Great instruments, nice melodies, a perfect track.


Clip 02. Dane - Break the Cycle - 6581 R2

Well, it obviously sounds different, especially the filter on the bass. Still, it's quite enjoyable.


Clip 03. Dane - Break the Cycle - 6581 R3

Thaaaaat background noise... 6581 R3 is definitely the worst SID version...


Clip 04. Dane - Break the Cycle - 6581 R4

Okayish, okayish, but...


Clip 05. Dane - Break the Cycle - 6581 R4AR

Better than R3 and R4 but not as good as R2 in my opinion.

. .. ... .... ....... .... ... .. .

Let's go deeper and listen to another track, Jammer - Mr. Marvellous, composed on 8580 and broken on 6581.

Clip 01. Jammer - Mr. Marvellous - 8580 R5

Great, great track, awesome instruments and speech-a-like sounds. Jammer did a very good job with designing these sounds and pushing the SID into its limits.


Clip 02. Jammer - Mr. Marvellous - 6581 R2

Whoah... this is... like a completely different music. Poor 6581... Some instruments sound completely broken.


Clip 03. Jammer - Mr. Marvellous - 6581 R3

:( :( :( the intro is completely broken :( :( :( the instruments and melodies are broken :( :( :( and the background noise is ugly :( :( :(


Clip 04. Jammer - Mr. Marvellous - 6581 R4

Same as the previous but at least the bass in the middle part is beefy.


Clip 05. Jammer - Mr. Marvellous - 6581 R4AR

I can hear the intro part! At least a bit. The bass is nice, overall it sounds better than the other 6581's but it has nothing to do with the original 8580 version.

. .. ... .... ....... .... ... .. .

Simon says let's go even deeper and break the cycle, I mean break the 6581! Jammer has some more awesome 8580 tracks, I'm curious how Club Style would sound.

Clip 01. Jammer - Club Style - 8580 R5

I just can't believe it how the SID is capable of doing these kind of sounds! Really awesome sound design, well done Jammer. The sidechain effect, the filter cutoff changes, the echoing lead. Just wow.


Clip 02. Jammer - Club Style - 6581 R2

Sounds like a completely different music. It's not a wonder, it's the difference between SID revisions. Jump to the next SID.


Clip 03. Jammer - Club Style - 6581 R3

I shall write something about this version but have no words to describe what I hear. Oh, did I mention the background noise of the R3?


Clip 04. Jammer - Club Style - 6581 R4

At least some of the instruments sound unique but the sidechain effect is still broken.


Clip 05. Jammer - Club Style - 6581 R4AR

Surprisingly, this version starts nice, keeps the okayish sounds and goes broken in the middle part. The distorted instruments at 2:00 sound kinda cool.

. .. ... .... ....... .... ... .. .

It's time to go mad - if you are still reading and listening. The following tracks is by Jammer as well, called Club Stylier (no, it's not the same as the previous one). Have a listen to the original 8580 version then to the 6581's without further comments. Thank you for reading this looong long post, I hope you enjoyed it.
See you next time.


Clip 01. Jammer - Club Stylier - 8580 R5

Clip 02. Jammer - Club Stylier - 6581 R2

Clip 03. Jammer - Club Stylier - 6581 R3

Clip 04. Jammer - Club Stylier - 6581 R4

Clip 05. Jammer - Club Stylier - 6581 R4AR

C64 SID shootout part 2.

Page has been updated with a nice music player, shouldn't any playback issue appear anymore \o/

. .. ... .... ....... .... ... .. .

Thank you for all the feedback! Looks like the previous post is a big success so I decided to add some more info and details about the shootout.

The recording chain was 3 C64, connected one-by-one to the input of a RME Babyface soundcard with RCA cable. Recorded with default loudness settings, all input faders were on 0dB. Obviously, there's not much to set on the C64, so the signal simply flew to the input of the soundcard.

Recording software was Adobe Audition 3.0.1 (the best Audition version of all :) ).

In case you were wondering, yes, all of the recordings are clipping a bit - this means the signals from the C64 were too hot but not _that_ much. I wanted to keep the original signal so I didn't limit/compress/lower the volume.

Loudness level is different for each SID, the recorded waveforms are also quite different:

I edited the recordings and chopped them to smaller parts to show the significant differences between the SID revisions. Play the audio previews and listen closely. Each preview plays the same part/melody and I switch between SIDs at even beats.

The order is obviously R2-R3-R4-R4AR-R5 :)

Clip 01:



The second and third part (R3 and R4) sound like the SID is pushed to its limit and it's hard to generate the flute-a-like sound.


Clip 02:



Here, the most significant difference is the snare drum instrument. It sounds quite different in each version.


Clip 03:




Different filtering on the bass. The most remarkable is the last one, which is the 8580 R5 version. The sound is more opened (higher filter cutoff and wider LFO) so the *wow*wow*wow* sounds different a lot.


Clip 04:



Again, the modulation on the filtered bass sounds different. Very. Different. Also, did you notice the loud background noise during the second section? That's the noise of the R3 SID.


Side-note #1:
I'm not 100% sure about the SID revisions I have. According to http://sid.kubarth.com/ my 6581 R2 1485 might not be R2 but R3. But still, it sounds so different to the real R3 1585 that I tend to believe it's a real R2.

Side-note #2:
Please follow this download link if you would like to download the audio files in .OGG format. Let me know if you would like to listen to the uncompressed .WAV files too.

Side-note #3:
Why .OGG compressed audio files? Because they sound better than .MP3's with the same compression rate and need less disk space. Also, for this test the compression doesn't take much information about frequencies, the difference between SID revisions is still well-audible.

C64 SID shootout

Page has been updated with a nice music player, shouldn't any playback issue appear anymore \o/

. .. ... .... ....... .... ... .. .

Have you heard about Stone Oakvalley? He is the man behind http://www.6581-8580.com aka. SOASC aka. Stone Oakvalley's Authentic SID collection:

"The SOASC= project is an automated recording technique invented by me (Stone Oakvalley) in order to mass-record music from the legendary Commodore 64 and its original SID sound chips (6581 and 8580), including variations of 6581, such as R2,R3 and R4!"

He gave me the inspiration to make a SID shootout and record 1 (one) music with different SID revisions. I was lucky enough to have my hands on 5 different SID, basically I have one SID from each available revision.

Why did I do it?
Because there is a difference in each SID revision's sound - sometimes it's just a slight difference but it's there - and I was curious how one of my favorite music would sound on each.

Brief SID history, copied from Wikipedia:

• 6581 R1 - Prototype, only appeared on the CES machines and development prototypes, has a datecode of 4981 to 0882 or so. Has the full 12 bit filter cutoff range. An unknown number were produced, probably between 50 and 100 chips. All are ceramic packages.

• 6581 R2 - Will say "6581" only on the package. Filter cutoff range was reduced to 11 bits and the MSB bit disconnected/forced permanently on, but is still on the die.The filter is leaky at some ranges and they tend to run hotter than other sid revisions. Made from 1182 until at least 1483. First 10 weeks or so of chips have ceramic packages (these usually appear on engineering prototypes but a few are on sold machines), the rest have plastic packages.
  
• 6581 R3 - Will say "6581" only, "6581 R3" or "6581 CBM" on the package. Had a minor change to the protection/buffering of the input pins. No changes were made to the filter section. Made from before 2083 until 1386 or so. The 6581R3 since around the week 47 of 1985 made in the Philliphines use the HMOS HC-30 degree silicon though the manufacturing process remained NMOS.
  
• 6581 R4 - Will say "6581 R4" on the package. Silicon grade changed to HMOS-II "HC-30" grade, though the manufacturing process for the chip remained NMOS. Produced from 4985 until at least 2590.
  
• 6581 R4 AR - Will say "6581 R4 AR" on the package. Minor adjustment to the silicon grade, no die change from R4. Produced from around 1986 (week 22) until at least the year 1992.
  
• 6582 - Will say "6582" on the package. Typically produced around the year 1986 in Hong Kong.
  
• 6582 A - Will say "6582A" (or "6582 A") on the package. Typically produced around the years 1989, 1990 and 1992 in the Philippines.
  
• 8580 R5 - Will say "8580R5" on the package. Produced from the years 1986 to 1993 in the Philippines, Hong Kong and in the US.

Are you still with me? I hope so because we shall jump into the most interesting part now.

So, based on the WIKI article and the fact that different revisions have different properties, the different SID revisions must sound different. And they do, of course :)

Let the competition begin! The music is "Jeff - Anal'ouge" (HVSC/MUSICIANS/J/Jeff/Anal_ogue.sid), which was composed on a 6581 - as Jeff confirmed he used a C64 with 6581 SID for most of his music.

The computers are Breadbin, Aussie and Blue:

The SIDs are:

01. 6581 R2 - 1485 (Breadbin's original SID)



02. 6581 R3 - 1585 (Aussie's original SID)



03. 6581 R4 - 3387 14 (put into Breadbin)



04. 6581 R4 AR - 4486 14 (put into Breadbin)



05. 8580 R5 - 3991 25 (Blue's original SID)

Conclusion:
This track sounds different on each SID revision :) Even the C64's own noise is different and not equally loud.

My favorite is the 6581 R2, it's rough and dirty and the bass is awesome. However, I like the 8580 R5 as well, it's clean (just listen to the flute-alike instrument at the start) and the bass is softer but somehow this makes the music more ear-friendly to me.

The other revisions sound like they have glitches, the snare-clap is a bit muffled (filtered?), bass is different but not bad at all. The most interesting difference is the already mentioned flute-alike instrument which is kinda distorted/modulated/filtered in a different way on each SID revision.

Which revision is better? Can't tell as it's a question of taste. Pick your favorite.