Tal J-8

[zvenx]

......

Well my last audio example posted clearly shows that they are not exactly thoroughly freerunning... That audio showcases exactly what's not possible on an analog synth. J-8 should not be able to do that if it is being called an emulation...

Again I love TAL stuff, but this is beyond weird.

Patrick did some tests again, and that is how his own is behaving.
It is a 14bit Jupiter if that makes a difference.

The instability is in the kHz range, very fast (you can see it with a frequency analyzer). I also think that both have the similar instabilities. Otherwise the VCO's are stable and don't drift. They are free running and they still have almost the same frequency after a few hours. Also didn't notice that the two can drift away from each other.

rsp

[e-crooner]

Never let people know which one is the hardware synth. Let them guess. Otherwise cognitive bias kicks in.

Not in this case, I can clearly hear a difference, and the first one sounds better to me. I am not in favor of either hardware or software, so I don't need to defend the hardware as you imply.

[gentleclockdivider]

System-8 Jupiter 8 PD Jupiter-8 Glide (Factory)

https://soundcloud.com/ggam-c/a?in=ggam ... /jupiter-8


Tal J-8  PD Jupiter-8 Glide (Factory)

https://soundcloud.com/ggam-c/b?in=ggam ... /jupiter-8


 

Two totally different attack stages , second being a lot longer
There is no harm in tweaking one to make it sound closer to the o.g.

[Stefken]

How long has Diva been out? How much new hardware was bought and subsequently died and been binned since then? Just providing some counterpoint.

I knew an u-he synth would pop up in this argument but these are very well supported.

[Teksonik]

Never let people know which one is the hardware synth. Let them guess. Otherwise cognitive bias kicks in.

Not in this case, I can clearly hear a difference, and the first one sounds better to me. I am not in favor of either hardware or software, so I don't need to defend the hardware as you imply.

Just once I'd like someone to run a test like this and lie about which one is the hardware. Say the software is the hardware then watch everyone praise what they think is the hardware.

[zvenx]

......

........ J-8 should not be able to do that if it is being called an emulation...

Again I love TAL stuff, but this is beyond weird.

Patrick did some tests again, and that is how his own is behaving.
It is a 14bit Jupiter if that makes a difference.

,,,,,,,,

rsp

Anyone here who has a Jupiter 8 who can see what it does on their hardware?
rsp

[chk071]

The instability is in the kHz range, very fast (you can see it with a frequency analyzer). I also think that both have the similar instabilities. Otherwise the VCO's are stable and don't drift. They are free running and they still have almost the same frequency after a few hours. Also didn't notice that the two can drift away from each other.

I hear in ED's sound snippet that the phase also seems to be the same with different notes he plays, but, after detuning and tuning the osc's the same again, the phase seems to differ.

I just read here that the synth got some hardware update regarding voice stability: https://megasynth.de/features/roland-jupiter-8/

Anyway, it's strange, but, I can't imagine that it was an oversight, when there was going so much effort in the other sound details. I have no explanation why the sound wouldn't change with different notes though, it does when the phase is free running.

[Sinisterbr]

I sold my SH -101 as it kept going out of tune and I had nobody nearby to repair it. Bought a TAL 101 and pocketed the several hundred bucks difference. Such is life.

The software route is still relatively new but it might turn out the lifecycle of a software synth is maybe like 5 years or so. Without putting an exact number on it: certainly (a lot) shorter than hardware.

You always have the option to "freeze" your computer setup in time.. thus it then behaves just like any other hardware. It lasts as long as it works.

That is not really the argument I want to make.
Rather that software gets absolete/outdated pretty fast ( it is also a fast moving business) and is disposed of quickly.

The same is true for business software. Stuff me make now is often thrown in the bin after some years and completely remade from scratch. Some might survive 10 years if they are lucky and it also depends on your base code. Some have become so big you can’t afford it but these are the exceptions.

But e.g. an sizeable drupal 7 site of 4 years will be completely redone from scratch. Whoop, everything in the bin after a few years.
You don’t do that with a 30 year old Jupiter 8

That's something wildly questionable. If I'd buy a hardware synth I'd have a LOT of problems trying to fix it where I live, when it eventually gets broken. I probably would then decide to sell it.

OTOH, it's far from difficult to make 20-30 years old software run in newer computers, if you know how to do it (granted, not every software, but I bet a modern VST synth, like TAL synths, will).

[kodnin]

In Live 10, Mojave 10.14.6, the back/fwd preset arrowv are erratic and most often jump 2 or 3 presets ahead of the next one. Also, some arps don't synchronise with the DAW tempo.

Yes, can confirm the preset browsing issue. It's odd. If I remember correctly there was also a similar issue on the TAL-Mod back in the day, but I might be wrong.

[EvilDragon]

......

Well my last audio example posted clearly shows that they are not exactly thoroughly freerunning... That audio showcases exactly what's not possible on an analog synth. J-8 should not be able to do that if it is being called an emulation...

Again I love TAL stuff, but this is beyond weird.

Patrick did some tests again, and that is how his own is behaving.
It is a 14bit Jupiter if that makes a difference.

The instability is in the kHz range, very fast (you can see it with a frequency analyzer). I also think that both have the similar instabilities. Otherwise the VCO's are stable and don't drift. They are free running and they still have almost the same frequency after a few hours. Also didn't notice that the two can drift away from each other.

rsp

I would really really REALLY like to hear a counterexample from Patrick's own hardware. Can he really do exactly the same thing as in my audio? Namely:

1. Get osc 2 finetune to completely freeze the phases between the two oscillators
2. After achieving point 1, play multiple notes and have them sound exactly the same with exactly the same starting phases
3. Change finetune to some other value then back to frozen state, which would result in different phase offsets but again constant...


I just cannot believe this is true on the hardware.

[Funkybot's Evil Twin]

How long has Diva been out? How much new hardware was bought and subsequently died and been binned since then? Just providing some counterpoint.

I knew an u-he synth would pop up in this argument but these are very well supported.

They’re like the RME of software. “15 year old product? Here’s an update!”

Maybe it’s a German thing.

[e-crooner]

Just had a look at that patch, it is not really glide, anyway, but an env-to-pitch modulation, which explains why the pitch always goes up.

[zvenx]

I would suggest you email him and ask him.
I only asked him to do the first which was your first example/post on the subject I believe (default, move both to 4', and remove detune).
rsp

[mholloway]

Never let people know which one is the hardware synth. Let them guess. Otherwise cognitive bias kicks in.

Not in this case, I can clearly hear a difference, and the first one sounds better to me. I am not in favor of either hardware or software, so I don't need to defend the hardware as you imply.

eh, somebody is clearly failing to understand how Cognitive Bias works. if the test isn't blind, it means nothing. simple as that. your assurance / opinion / confidence that "you can clearly hear a difference" doesn't mean anything to the rest of us (or to science, heh).

[chk071]

If the software example would have sounded better, I would have posted that the software sounds better.

Don't judge others by your own standards.

Anyway, a single example means absolutely nothing, especially when it's soaked with fx.