Upcoming Synapse OB-Xa: Obsession

[EvilDragon]

Sorry, but yes. That's how biases work.

[Examigan]

There should be a test like Peter did with his synth OP-X Pro II. He randomly changed which one was played first or second, and you had to choose between them without knowing before hand which one was which.

[recursive one]

Based on Synapse’s other work, I’m 99.9% sure my opinion of the OB-Xa will be higher than my opinion of the Roland Clouds Jupiter-8. So once it’s actually out we can get back on topic.

Unless Ingo, ED, or Bones can get permission to post some teaser audio or screen grabs to get this thread back on track.

The guy from Synapse posted these examples on GS a while back from an early build. First half of each clip is hardware, the second half is the software.

https://soundcloud.com/kayveyar/big-pad-2x-1

https://soundcloud.com/kayveyar/resothrust-2x-1

hardware sounds a bit thicker to me and has more "body".

But the difference is very small and almost certainly would completely vanish in a mix.

Also both clips (software) sound to me like they might have been made with Dune 3, especially the second one.

[mholloway]

There should be a test like Peter did with his synth OP-X Pro II. He randomly changed which one was played first or second, and you had to choose between them without knowing before hand which one was which.

The only A/B testing I'd consider of any 'value' would have to be with fairly complex patches with lots of character and movement. These 'let's compare raw oscillator waveforms against each other' videos are as boring as they are useless (nor do they ever succeed in converting the hardware snobs, even when they should). Slightly better but still boring is the "here's the bog standard moog bass sound"....oof.

[e-crooner]

Frankly, I don't hear any difference between on those audio files.

[Cinebient]

There should be a test like Peter did with his synth OP-X Pro II. He randomly changed which one was played first or second, and you had to choose between them without knowing before hand which one was which.

The only A/B testing I'd consider of any 'value' would have to be with fairly complex patches with lots of character and movement. These 'let's compare raw oscillator waveforms against each other' videos are as boring as they are useless (nor do they ever succeed in converting the hardware snobs, even when they should). Slightly better but still boring is the "here's the bog standard moog bass sound"....oof.

That is true but still i see seldom comparisons where they really got into all the weird and nice audio rate stuff, how it reacts if you automate the hell out of it, use midi CC to control some things on the fly and whatever.
Indeed it is the sum of all which is important here and also some unique things no other synth might has to offer. I do not care really how the raw sawtooth OSC sounds or if it can recreate the simplest presets from the past which might 99% of stock synths coming with any DAW can do.
But at the end the market speech and people seems to like and pay for nostalgia.
But nothing wrong to put some modern tech on top of this.

[Examigan]

Here's a comparison between an actual OB-X and OP-X Pro II. These are not complex patches with tons of character, but they're not raw oscillator waveforms either:

viewtopic.php?f=1&t=348343

[Echoes in the Attic]

Here's a comparison between an actual OB-X and OP-X Pro II. These are not complex patches with tons of character, but they're not raw oscillator waveforms either:

viewtopic.php?f=1&t=348343

That was a good comparison and it's a seriously excellent synth. I'll certainly be getting the new Synapse OB, but I won't be getting rid of the OP-X-Pro, still love that one.

[Echoes in the Attic]

accidental quote post...

[v1o]

Looks like we’re going to get 100 pages of waffling before this thing ever gets released.

[Cinebient]

Looks like we’re going to get 100 pages of waffling before this thing ever gets released.

There are threads with 150+ pages before official release.
So we can get better. Lets do 200

[v1o]

An interesting thing in the OB-Xa i that it had 2 separate circuits für 2-pole and 4-pole LPF. The 2-Pole filter keeps most of the low end with high Resonance while the 4-pole filter has a decreasing low end with high resonance.

Details about the 2-pole and 4-pole designs of the OB-Xa could be found here:
https://electricdruid.net/cem3320-filter-designs/

Are Bigtone studios involved in the creation of this? I know they were instrumental in getting the voicing just right on The Legend. Most analogue synth emulations are maybe 90-95% close to the original hardware, but what’s makes a great emulation is the extra 1-2% of accuracy.

[BONES]

No, what makes a great emulation is that it goes way beyond the original without losing the spirit of what it was. That "extra 1-2%" means nothing in a mix.

[v1o]

No, what makes a great emulation is that it goes way beyond the original without losing the spirit of what it was. That "extra 1-2%" means nothing in a mix.

It’s not always about the mix. You want an instrument that can stand out in its own and do convincing solos, which is where that extra 2% comes in. With analogue synths and a decent analogue mixer, you can get a full sounding arrangement with just four or five tracks.

If you’ve been mixing in the box all your life you’re going to be accustomed to layering sounds on top of each other and generally using lots of tracks to create a full sounding arrangement. But that’s a byproduct of two decades of making music on computers.

[PAK]

They only display the text strings 0-255, but the sound is still NOT stepped but processed with higher internal resolution. That's what matters

In terms of avoiding stepping, yes. In terms of extracting maximum detail from any good model? No. You've now limited everything to 256 settings, which is enough for many things, but a filter model with a bunch of resonance detail could easily make use of more than that.

This is especially true when you consider those steps are likely to be distributed evenly around the dial, whereas it's not uncommon for model settings to squeeze a bunch of detail into a smaller range of movement on the dial.

This is made worse, on the Jupiter plugin, because some controls were expanded to include a negative range, unlike the original. Which is cool in terms of features, but then reduces the range of what was already there to 7 bit, because now it has to use the other 128 values for the other direction.. So, in that sense, some controls are indeed 7 bit.. just in each direction for a total of 8 bits

BTW Roland also agreed this wasn't enough when they considered it worthwhile moving their new Zen Core engine to 10 bit. I'll be curious to see if the Jupiter X implements 10 bit controllers like the Fantom. Although I don't like the backwards step to Zen Core, in terms of reduced emulation detail, it's nice to see Roland at the forefront of going beyond 7/8 bit resolution on the hardware side.