I Think so. The old version is out of sync because of the bugs but still, the sound is a little bit ''better'' at the same volume. It sounds more ballsy.

Comparison:
October 4, 2021 build sound: ''old''
July, 06, 2024 1.3.15 build sound: ''new''

https://filebin.net/c6dymct0vv9gl3c0

 

This "change" might be a result of settings being different or less "chorusy", so hence being ballsy in the old version ?

 

I dunno man, the default settings are same. In fact I had to boost the output +6db in the new version to even get close to the old output but its more than just default output its that all patches sound a bit duller/muddier in the new version for some reason. Its missing a bit of high end detail or something and lacks the clarity of the old version. Glad I still have that 8mb version lol.

 

It could be, but it is a rough road going against bias and problems with differences in levels and how 0.1db can influence us and so on, i hear it but i havent compared side by side. But if it was like this, i would probably have old buggy version that sounds "better" than super polished that lost it´s charm.

 

Adding that +6dB to the VirusTI2 audio channels is entirely normal to me (in Logic) when using either of digital connections. Usually I will add a Logic stock Gain Utility plugin at around +7dB, actually. With it set at +6dB in the plugin, it's very close in levels.

 

Another crappy example.
https://filebin.net/ru39h4s0wyqj6q77

According to reticular this version sounds the closest to the pre gui version (October 4, 2021) https://futurenoize.com/dsp56300/builds/osirus/archived/1.1.9/

I have not tried it again yet.

 

Placebo can be rough, so i wont claim anything, i also think renders from both versions cannot be nulled since the file is not repeatable based on synth architecture. Waveform wise it looks different, but it is a rabbit hole in my opinion, unless several people can test it and confirm you could claim you like this version better than the newest and in that case i would agree.

I wonder what would devs comment on this topic and if it were true does it mean that new synths they are putting out carry the same thing over..

 

It does. Other synths like Vavra sounded dull to my ears to from the very beginning. I would really like to hear an explanation about this topic. They definitely messed up the sound somehow perhaps unintentionally but then again to reduce cpu usage..who knows... My best guess is those cpu/latency optimizations that were done compromised the sound somehow. Either that or something else entirely I have no idea what I'm talking about.

 

I thought the same when i heard Vavra demos..but i know depending on the time in a day or setting i feel differently.
If there were optimisations that degrade the sound i would rather get a new cpu to run it rather than having it being polished to the brim and slim. I still hope a setting is different or something in that neighbourhood, or maybe the options should be changed i think the new version has some.

I guess uptill now nobody had an idea to compare new vs old versions

 

For some reason the sound of the new versions were bothering me so I reverted back to the old version to try it side by side. Yup, I was right. Its definitely not placebo. Default settings are either totally different or the sound ''engine'' is inferior.

 

to be deleted

 

I noticed that "old" files had circa 0.7 db more volume than "new" files in both comparisons.

In the second example I can easily pick new or old file blindly: the new one is slightly duller.

In the first example the playing timing is different so I can pick them just by hearing the different playing.

Are you sure you're not playing patches with slightly different velocities for some reason?

You should test both versions with the same MIDI file. Are you sure sample rate conversion settings in the synth are both the same on new and old versions? Just to be sure that all settings are equal.

 

The only thing that has changed is the gain and latency/sync. The sound itself has not changed.

Worrying about this is a waste of time. (reminds me of audiophiles who swear they can hear differences between certain equipment when the difference is only a lack of gain matching).

 

I know, but it can be a legitimate difference that occured and nobody had time to do a comparisson thru several releases.

 

Afaik only a change in SRC algorithm could cause a change. Code is the code, but natively they don't run at 44k or 48k

 

Who's to say the code hasn't changed. We need a response from the devs on this topic. Why not try it side by side yourself in your daw. The sound is worse. I can hear it, and its not subtle.

 

Just level match and do a null test to be sure.

 

If anyone wants to use them side by side I can send the October 4, 2021 version (pre gui) to you.

 

You can't alway do null tests, for example when there are free running oscillators or random modulations etc...

 

I compared the pre GUI version 1.1.9 to the current one and they sound the same after level matching.
I recorded a file with the default patch where I switch from one to another and it's impossible to distinguish them.