Heres the run down (with only 2 voices):

JunoX2 1.3 = 35%
Dynamo = 28%
Wasp VSTi = 16%
Square 1 = 15%
FreeSonic = 11%
Ninja = 9%
daOrgan = 9%
Pentagon 1.2 = 8%
Vaz 2001 VSTi = 7%
Image-Line DX-10 VSTi = 4%
CM-101 = 4%

It surprises me that Pentagon 1.2 and Vaz 2001 VSTi are at the bottom of the list as these are very powerful VSTi's...

...also I thought the days of CPU hungry VSTi's were over. I was told Dynamo was the hungriest...but it seems that JunoX2 has beaten it to the number one spot

It would be more helpful if you offered up a rundown of your system as well.

For what it's worth, JunoX2 seems fairly efficient on my Mac - MUCH better than any Reaktor/Dynamo ensemble.

PPG

....good point! But this was a like for like comparison

Anyway, just for you:

PC, Pentium III, 733MHz,
512 mb SDRAM,
Windows 98 SE

Hosts tested with Fruityloops and Cubasis VST...both gave similar results.

ppgwave...maybe it would be useful for you to have stated CPU % figures to demonstrate this??? :

[ 16 March 2002, 14:58: Message edited by: The Artist Formally Known as Alex ]

maybe the mac version is more efficient?

Alex, I am a little surprised at your results. I just tested JunoX1.3 in Orion Pro on my games machine, a celeron 1.3 tualatin. With two voices I could not get the cpu past 5% no matter how hard I tried, with the fat slider on mono I maxed out at 3 percent. Dude my system is more powerful than yours granted, but SEVEN times as powerful??? I dont think so. By the way this was using an audigy with directx drivers set to 25.5 ms latency. What sound card are you using alex?? Something doesnt sound quite right

The reason it makes sense to list the configuration is:
Mac and PC have very different utilizations, some plugs are optimized for SSE/G4 (so Celerons or G3s might suffer), Reaktor 3.05 is much different from Reaktor 2.31....

It just never hurts, when presenting "scientific" data to list the environment in which the data was produced.

PPG

G4 Dual800, 1.5gig ram, MOTU 828, Logic 4.81.

[ 16 March 2002, 16:00: Message edited by: ppgwave ]

- the performance/efficiency of junoX2 seems very
system dependant - on my machine [hp pavilion laptop - mobileduron 850 / 256meg ram] i can max out at 13% cpu usage but it usually hovers around 6-7% with 2 voices in orion

- cheers rob
phuzzy_logical

[ 16 March 2002, 16:16: Message edited by: phuzzy ]

Maybe you should try shorter release times and the fat slider to off. Your comparision is NOT fair to junox:
The hoover sound for example uses 6 Oscilators at once for just one voice!
Or do you wish non - fat but faster vsti's for the future???

regards - feilei - reFX

- just for interest i thought id check some others out on my system [hp laptop - mobileduron 850 / 256meg ram / windows me / crappy internal soundcard] -

synthetic - 11%
craxydiamonds [everything on] - 10%
cronoX [bassline1 preset] - 9%
triangle - 8%
tickyclav - 3%
dx10 - 2%
cm101 - 2%

- all seemed reasonable to me and in the end [on my setup at least] the junoX2 came out there or thereabouts [maxing at 13%] - not that hungry

- cheers rob
phuzzy_logical

Michael/Feilei

I re-assessed by findings...and you'll be pleased to know that I found out the reason why ReFX JunoX2 came out at 35%. If you notice from my first post I was comparing the JunoX2 with Dynamo's Junatik. I was originally using them side by side to see which one I prefered for sound....you'll be pleased to know that JunoX2 came out on top. It is bloody excellent and I'm so glad I bought it...its one of my best synths...although it was one of my cheapest buys!

Now, I just bought Dynamo and I new it would be CPU hungry as I'd seen the reviews. I had Junatik playing 2 voices and the CPU was at a steady 28%. Then I closed this and used the JunoX2 which showed 35%. I then switched off my computer and went into town to get some lunch. On return I turned my PC on and continued to assess the CPU load of all my VSTi's.

Now you have pointed this out, I have gone back to have another look. It seems that JunoX2 shows 9% which is very low for the amount of 12 oscillators being used at once (in hoover playing 2 voices). Then I closed the JunoX2 and opened Dynamo Junatic. It gave a steady 28% again. I closed this, re-opened the JunoX2, and what do you know, 35%. It seems that CPU hangs high even after I've closed Dynamo. I tried all the others on my list and they too gave CPU consumption 25-27% higher than my previous readings. They only gave true readings before because I had re-booted between using Dynamo and them....unfortunately for JunoX2 I didn't.

My apology:

So...I apologise for my mistake and advise everyone who hasn't bought this wonderful synth to go and get it coz it sounds bloody wicked and has a respectable CPU consumption. It is very much worth the money and I'm glad I have it. (erm..does that say it all?)

Thats the second time I've had to apologise today (sorry again William...oops, that was the third time!)

Anyway, here is my new rundown:

Dynamo = 28%
Wasp VSTi = 16%
Square 1 = 15%
FreeSonic = 11%
Ninja = 9%
daOrgan = 9%
JunoX2 1.3 = 9%
Pentagon 1.2 = 8%
Vaz 2001 VSTi = 7%
Image-Line DX-10 VSTi = 4%
CM-101 = 4%

[ 16 March 2002, 21:02: Message edited by: The Artist Formally Known as Alex ]

Very interesting, about VAZ2001. I thought it was low and it appears I was right. Great for the sound quality.

I was surprised about Crazy Diamond, didn't think its load would be that high, thought it was less CPU hungry than Ninja!

Michael

Isn't P1 just great. I did a track in Sonar with 5 Pentagons and 1 Triangle and was nowhere near maxing out my CPU. I think it was at 40-50% with a vocal track playing with 1 EQ and Compressor as inserts, and two intermitant audio tracks (short little noisy things). This is probably the first time I've done an entirely synth based track (as I'm pretty much a rocker) so I was just a little surprised at how many instances I could get away with of that bad boy.

- jorgen - i wasnt knocking synthetic [its actually one of my faves at the minute] - 11% was the highest usage i could wring out of it with lfos on everything etc. - it normally hovers around 8-9% which is fine for something that makes such wondeful noise

- cheers rob
phuzzy_logical

I love looking at stats like this. Thanks everyone.

Juno looks great. I may compare the demo to the Juno emulator I already have on the Pulsar platform called U-KNOW 007. It's meant to be pretty good so I may see what they sound like side by side.

Caleb

Finally got around to testing the JunoX2 and I must admit I wasn't that impressed.

I didn't bother comparing it to my Juno emulator on Pulsar because they were obviously two TOTALLY different synths.

Is the sound drenched in reverb or something?

It does produce a certain type of sound quite well and the range of oscillator types is quite novel, but there's just nothing here that grabs me - would probably make a nice freebie.

Arrgghh! If I try to explain it any further it will just sound insulting.

I hope everyone has fun with this one.

That's what I meant.

However, was there really a Juno that had all those oscillator types?

I don't know much about those hardware synths, but some of those oscillators like vocal and comb I don't imagine being put on a Juno.

Oh well - a big "whatever" I guess.

Caleb

I programmed a basic 2 oscillator lead sound on each VSTi. This included using 1 low pass filter (set to about 64 reso and 64 cutoff, and where a synth had effects these were switched off (except for DX-10 and daOrgan with don't have filters of course which might explain why they are low cpu users).

oh, BONES: why is my "array" of synths "interesting"? Just curious!!

[ 20 March 2002, 18:58: Message edited by: The Artist Formally Known as Alex ]

Just that you have two really expensive ones and a bunch of cheapies, with Pentagon1 being in the middle-ground. The impression I get is that some people have everything under the sun whilst the rest seem to favour either the big commercial products OR the shareware-priced ones. You have some of each, which I found interesting.