How can we say something is the "identical code" when they were changing code throughout the entire development process?

All you have to do is a quick search on here to see all the comments about Viper sounding just like a Virus, before the DSP56300 emulators became available. Now we can load a ROM file into a software synth and give it the benefit of the doubt. You can even look on here at the development threads of these DSP56300 emulators where people were saying one version sounded different than another. Code optimization being done to unburden the whopping cpu hits they first had, etc. Simply loading a ROM file is just that, loading a file. This thing could be Viper with a different GUI and a .bin file requirement to allow a instance to be loaded by the DAW and people would try claiming the same things.

Put a dither plugin on a 24bit audio file in your daw and lower the bit depth. Does it still sound identical, because I did not alter the file loaded onto the channel?

 

It does not mean "identical" code is going to result in identical sound, and that the user saying he noticed a difference is an impossibility. Even a supposedly perfect dsp56300 rom emulator is still only as accurate as the emulation, timing, in/out. If the differences or "errors" are not obvious; he could still notice differences using effects like reverb, the various filters, the vocoder, other modulation.

It requires a good amount of assumptions to say his observation is wrong out of hand, without any actual sound files as proof.

 

Mate, are you serious now? If it nulls it is identical. It's perception that can play at fault. You can listen to the same audio file couple of times and it might sound different based on your changed mood or your focus

 

If the emulation is the result of decompilation, the output should be 1:1, there's no room for interpretation in that regard. Look into FPGA game consoles, for example.

I can't really comment on how this piece of software was made because I didn't follow the development nor I know the original, but this is something that can be tested rather than "I think it's different".

 

Emulation even if it runs the same old code can give bit different result
In game console emulation there are games that don't want to work or work with various artifacts there could be numerous updates iterations and tunings to the emulator to achieve the game running right.
As for OStirus or any others there is only one piece of code ideal operation of which that needs to be attained for a result ) and then start doing a long laborious effort achieving the next synth emulation to work.
DSP56300 project emulation is also not perfect and requires manual refinement for each project, thats why we don't see the quick releases of other synthesizers popping up

 

Last edited: Sep 13, 2025 at 3:24 AM

OsTirus can also use sample rates that weren't possible on the hardware (it was in the og code base I believe, just not implemented).

 

If you actually get into the PAS of differences in audio perception; you can't overlook the factor of human auditory memory. Maybe you have done blind tests yourself, or you can watch a really good one of the Venus Theory youtube guy trying to ID his expensive Arturia (I think Polybrute) and various other plugins intended to sound the same. He would have had better luck rolling dice, about a synth he was using that same day; nevermind one he hasn't used in years. It's because we have extremely short auditory memory to retain any real accuracy about sounds. We can only remember broad descriptions and comparisons for more than a few minutes, if even that.

I never saw a single user proven wrong when they were still testing some of the emulators, that mentioned different builds of the same plugins had differences in sound, even while using the identical ROM in them. In fact, I've never seen any null test results provided at all. That's surprising.

 

One time I made two identical loops for a guy who claimed he could always tell the sound of a Virus apart from other virtual synths. He listened and was like, “Yeah, totally different.” Both loops were straight off my Virus TI2.
I love the Virus, still use it as my main subtractive synth to this day. But man, the myths and folklore around this thing is something special.

 

the "mythology" about these particular emulators is because they are some of the few examples where they used real factory roms for them. Everyone already expects usual emulation plugins to be inaccurate to some degree, and developers market them by whose is closest.
People get out their Plugin Doctor and do some comparisons between them.

 

Word.

 

100%

I beta tested Viper before Adam released v1 and I remember being so impressed with how close of an approximation Adam had gotten to the Virus. This was before the option to import presets, but even still, recreating presets on Viper based on Virus presets got me pretty damn close. When the option to import actual virus presets got implemented, it was even better, not every presets was perfect but you could get even closer approximations. Which is perfectly fine for what I do. I still have Viper for older projects that I used it on, but these days I just use Ostirus so I dont have to deal with latency issues with the TI using the VirusControl app.

Personally, from a production standpoint all I care about is how something sounds and couldn't care less where it comes from as long as it works in a mix. These days there are a lot more options available to get that Virus sound without having to own a Virus. I don't think you'll find a better option than Ostirus, it's literally THE Virus sound. I'm also not saying that to convince anyone. If it sounds off, maybe what you're looking for has less to do with what it's emulating and more with what sounds right to you. All we can do is a make a decision that feels right and start making music.

 

It can be tested. Have at it. Also, if you actually read the thread; nowhere in it did I say it is inherently different. I commented that the usual claims of something being "100% accurate" are not a given just because it is loading a factory ROM. Moreso, that the original comment cannot be ruled out as a total impossibility just because a factory ROM is used. The newest TI ROM is directly from the installer: Virus TI Software Suite 5.1.7.00. The only way to end up with something different is to open an older installer to get the .bin, or using the Virus Control to burn new data back to the TI, saving it there, and then re-exporting it. You can create and store copies of as many versions as you want on your computer, but it can only contain 1 version of the firmware at a time.

Lastly, here are differences (minor or not, doesn't matter), in parameters which cannot even be edited using the old Access VCC plugin or the emulator software interfaces, nevermind the Misery Island editor:

Spoiler

Hidden / Hardware-Only Parameters

• Oscillator phase offsets
  
  
• Internal oscillator routing tweaks
  
  
• Random waveform start index (for wavetables)
  
  
• Minor filter internal routing tweaks (resonance/FM behavior)
  
  
• Envelope iteration counters
  
  
• Hidden envelope shapes
  
  
• LFO initial phase/position
  
  
• LFO sync state
  
  
• Hidden LFO destinations
  
  
• Randomized modulation source seeds
  
  
• NRPN-only modulation matrix entries
  
  
• FX internal dry/wet balance tweaks
  
  
• Delay/panning random seeds
  
  
• Voice allocation states (per-voice assignments)
  
  
• Internal counters affecting stereo/voice behavior
  
  
• Random phase seeds for oscillators and noise
  
  
• Firmware-dependent internal tweaks / initialization values