SID 8580 vs. ArmSID vs. SwinSID Ultimate vs. SwinSID Nano

It was time for a new comparison post for more reasons.

I acquired an ArmSID on the weekend of this year's Revision demoparty and a SwinSID Nano from a friend (thanks, Joe!), so this is a good occasion to compare them while I'm waiting for the release of others like the FPGA SID.

First of all, I wanted to write one post only, but then I realized it's better to split it to two episodes. I'm going to show a detailed example with waveform screenshots and write about my personal opinion. The second part will be about audio examples, recorded with the 4 chips and you can compare them for yourselves and make conclusion.

Some notes before we start:

• the example track composed with SID-Wizard 1.7 on a 8580
• the audio examples are pure recordings, directly from C64's output
• all chips were in 8580 mode
• no volume normalization, no additional effects, just pure output
• nope, no volume normalization, this way you can see/hear the real output of each chip
• really, no volume normalization, you have a volume knob on your audio system, right? use it

So, let's see what we have here: \HVSC\MUSICIANS\V\Vincenzo\SIDrip_Arok_Invit_2017.sid

FYI, the proper title is "SIDrok" because it was composed for Arok Party 2017 invitation and released by SIDRIP Alliance. SID-rip, a-ROK, see? SIDrok. But whatever, this is a minor thingie

I cut pieces from the track to show remarkable differences between each chips. The order of playback is:

• SID 8580
• ArmSID
• SwinSID Ultimate
• SwinSID Nano

First audio example, heavy usage of ringmod and oscillator sync:

Listen to "Example 01 [SIDrok - vincenzo]" on Spreaker.

8580: It doesn't need any explanation.
ArmSID: Output is definitely quieter with a few dB's, and the tail of the sound has somekind of strange flanging effect and it's shorter than the original. Other than that, it sounds good.
SSU: The output is definitely quieter, however, the sound is very good, except that the tail is shorter than the 8580's tail. SSU doesn't have the flanger effect.
Nano: Wow, I didn't expect this to happen. The first half of the sound is okay-ish, the second half is... what? Out of tune. And sounds weird.


Let's hear the second clip, another part of the music:

Listen to "Example 02 [SIDrok - vincenzo]" on Spreaker.

8580: Raspy, beefy, nice filters, dynamic.
ArmSID: A bit muffled filters, however, it sounds beefy enough.
SSU: Wow, definitely different filter sound, it's muffled and the snare became darker.
Nano: Ouch... I mean... ouch.


Third example:

Listen to "Example 03 [SIDrok - vincenzo]" on Spreaker.

8580: Same as the previous clip.
ArmSID: Same as the previous clip.
SSU: Same as the previous clip.
Nano: I would rather not comment this.


Waveforms:

Now the differences are not only audible but visible too. To my eyes, the waveform reproduction is good on both ArmSID and SSU, the Nano however is a bit far from perfect.
They are not perfect emulations. However, as replacement of a burned SID, well, they are good enough and I would be interested to do a blind A -B (C) comparison to test people and myself too. Okay, I can recognize the Nano for sure, but SSU and ArmSID are hard to tell.

Conclusion, before I post the complete tune 4 times:

• Audio output is louder on 8580 than on the rest of the chips.
• ArmSID's output is quiet, noiseless.
• SSU's output has a very low, very quiet buzzing noise.
• Nano's emulation is something. Different. But maybe it's the tune what's so special.
• Hard to choose between ArmSID and SSU, they are both good even if we consider the darker filter output of the SSU.

See you in the next episode with more example tracks.

01. vincenzo - SIDrok [SID 8580]

Listen to "vincenzo - SIDrok [SID 8580]" on Spreaker.


02. vincenzo - SIDrok [ArmSID]

Listen to "vincenzo - SIDrok [ArmSID]" on Spreaker.


03. vincenzo - SIDrok [SwinSID Ultimate]

Listen to "vincenzo - SIDrok [SwinSID Ultimate]" on Spreaker.


04. vincenzo - SIDrok [SwinSID Nano]

Listen to "vincenzo - SIDrok [SwinSID Nano]" on Spreaker.