Repro-1 (out now)

[deastman]

I have a 2 year old Dell Inspiron i7 laptop, and most of the presets are maxing out my CPU.

For real? A single instance?

Yup. The only conditional I'll put on that is that I was using onboard audio with ASIO4ALL and didn't have my audio interface connected. Windows 10, Ableton Live.

I can probably get you better details over the weekend, schedule permitting. Anything you'd like me to test?

[Cinebient]

I have a i7 quadcore 2.0 Ghz notebook and i can run 12 instances without trouble.
Even 6-8 with some FX on top are no problem here.

[Urs]

Anything you'd like me to test?

Nothing specific right now.

But as CPU usage pans out to be the main point of criticism, I guess I'll work double shifts to bring it down.

(I was expecting weird things like people posting audio samples of Repro-1 vs. their Jupiter-8 boutique minisynth or something, so I'm quite glad that the negative aspects are actually measurable on objectively quantifiable so far)

people who post Repro-1 vs. JP-08 threads will blissfully be chuckled at

[Urs]

To put some perspective on it, here's a demo track I did with 12 instances on a ordinary spec'd single processor hexacore dustbin Mac Pro:

https://soundcloud.com/uheplugins/urs-h ... miniscence

Each core was taxed a max of 40%... hence my surprise when a single instance maxes out anything recent.

(Btw. we internally have nanosecond-accrate timing information for each module in the synth, hence we have a very good picture of performance issues, unlike what some DAWs sometimes show)

[Funkybot's Evil Twin]

The only proper CPU usage result you're gonna get from task manager. Open your DAW and run a single instance, there you have it.

Studio One was showing 12-14% per single instance of a simple patch, with 22-25% per for instances that used a lot of effects. However, 8 instances with various patches only shows 33-34% so that single instance number definitely isn't real. Meanwhile, Task Manager only showed between 6% for a single, simple patch, and 8-9% for a more complex one. This is with an i7 4790k @ 4ghz in Win 10.

So definitely not light on the CPU at the moment, but not horrible either and Urs already said it'll get even faster. The S1 CPU meters do show some quick spikes. Not sure if that's real or not. Urs, if that's anything of concern let me know, otherwise, I'm assuming it's just something in the beta code and will report if it's in later releases. The CPU spikes haven't turned into an audible issue yet, so it may not even be real.

[zvenx]

Anything you'd like me to test?

Nothing specific right now.

But as CPU usage pans out to be the main point of criticism, I guess I'll work double shifts to bring it down.

(I was expecting weird things like people posting audio samples of Repro-1 vs. their Jupiter-8 boutique minisynth or something, so I'm quite glad that the negative aspects are actually measurable on objectively quantifiable so far)

people who post Repro-1 vs. JP-08 threads will blissfully be chuckled at

My download won't play chords. Obviously it's broken. (yeah bad joke)

Firstly this sounds fantastic and I am enjoying it a lot and wasn't didn't think I would.

Secondly I especially wish it used less cpu especially when it wasn't triggering any audio (yes ED pointed me to why it doesn't), I would have loved that as an option at least.

Thirdly although by far the minimoog hardware was more preferable to me than the Pro One, that you used velocity, aftertouch in such usable ways it is making me enjoy playing this more than that other vst I recently purchased and love too.

rsp

[Urs]

Secondly I especially wish it used less cpu especially when it wasn't triggering any audio (yes ED pointed me to why it doesn't), I would have loved that as an option at least.

I feel very compelled to lift the idea of a "stand by during silence" option off of the other great emulation that came out recently.

[wagtunes]

^What DAW is that, with no resource usage meter?

Cubase 7. Unless I am blind as a bat and just don't see it.

[Cannabis]

Kinda of sneaky way of getting people to buy the product.
1- Free public beta
2- User goes and try the plug In, makes some presets, starts some projects with it.
3- And before it knows, it is hooked.
4- Then it is given the option to buy the the plug in at a slightly reduced price, or his beta will expire

You get someone from a FREE BETA to BE THE OWNER OF A PAID LICENCE.
MONEY IN THE POCKET.... RIGHT? NOTHING WRONG WITH MAKING MONEY WITH YOUR EFFORTS
BUT
Isn't BETA for testing purposes for a small circle of users, who report back the problems?
PUBLIC BETA???? OR MISLEADING BETA?!! OR MASS SALE ?!!!

[Ingonator]

^What DAW is that, with no resource usage meter?

Cubase 7. Unless I am blind as a bat and just don't see it.

Had used Cubase 7 quite a long time ago but here in Cubase Pro 8.5 it seems to be "VST Performance" in the "Devices" menu (here in the german version it is the "Geräte" menu).
I could also open/close this by pressing F12.

[Urs]

Kinda of sneaky way of getting people to buy the product.
1- Free public beta
2- User goes and try the plug In, makes some presets, starts some projects with it.
3- And before it knows, it is hooked.
4- Then it is given the option to buy the the plug in at a slightly reduced price, or his beta will expire

You get someone from a FREE BETA to BE THE OWNER OF A PAID LICENCE.
MONEY IN THE POCKET.... RIGHT? NOTHING WRONG WITH MAKING MONEY WITH YOUR EFFORTS
BUT
Isn't BETA for testing purposes for a small circle of users, who report back the problems?
PUBLIC BETA???? OR MISLEADING BETA?!! OR MASS SALE ?!!!

Hahaha, what have you been smoking?

[thecontrolcentre]

[wagtunes]

^What DAW is that, with no resource usage meter?

Cubase 7. Unless I am blind as a bat and just don't see it.

Had used Cubase 7 quite a long time ago but here in Cubase Pro 8.5 it seems to be "VST Performance" in the "Devices" menu (here in the german version it is the "Geräte" menu).
I could also open/close this by pressing F12.

Thanks. It's better than nothing I guess but it doesn't show actual percent. Just a couple of lines that go from 0 to 100 and you have to estimate the actual percentage, plus it's for the entire project. To get one instance I'd have to either unload everything else or take a measure before inserting plugin and then a measure after inserting plugin. But like I said, I guess it's better than nothing.

And yes, this thing does look like it sucks up about 20% CPU with one instance. But that's with all 5 FX activated for this particular patch.

[fluffy_little_something]

The variation in CPU load from one computer to another seems extreme. When my old AMD system handles it way better than Deastman's i7 laptop, things are getting really weird Does that plugin prefer certain settings (DAW sample rates etc.)? Or multi-threading?

[Ah_Dziz]

I don't know what the hell is going on here at the moment, but....

Me neither

As for the Minimoog comment - the modulation ranges and capabilities are much larger on a Pro-One. It does audio rate PWM, FM, Sync and FilterFM at the same time, and - as EvilDragon already said - over an extreme 12 octave range. It's nuts. The oscillators need to be oversampled like there is no tomorrow.

Furthermore the filter is a very different architecture. I guess everything has been said about the Moog Ladder filter. But there isn't a single paper about the CEM 3320 out there, just a spec sheet and a patent, maybe a blog entry somewhere. It works completely differently from anything I've seen, and that's what makes that sound. It's also currently the most demanding algorithm in Repro, and I have high hopes that this is gonna change.

Well you did a great job on it. I wanna load up two instances on my laptop grab some stomp boxes for my sm 58 and go start a corny old school industrial band. I'm having lots of fun making bizarre sound fx with the osc self modulation. I forgot how crazy that can get especially in lfo mode.