KVR Audio

alienmachine · Post by **alienmachine** » Wed Sep 24, 2014 2:08 pm

Thanks, Hunter!

Z1202 · Post by **Z1202** » Tue Sep 30, 2014 10:32 am

alienmachine wrote:I started sporadically working on my Eminent 310 emulation again. I was never completely satisfied with its sound and several years ago I found why: my oscillator is based on a square wave, which is filtered to become a quasi-sawtooth wave.

Just for the record. In your current design you use a filtered pulse wave, which you clip afterwards. The clipping is the source of some aliasing (although rather subtle). You can achieve a similar sound in a different way. Just take an (antialiased) sawtooth and HPF it. The HPF cutoff should be equal to the oscillator frequency. This way you don't need to clip. The resulting wave will have a somewhat different shape but practically identical amplitude spectrum (the phase spectrum will be different, but I don't think it is audible in this case).

Anyway, waiting for your new oscs

alienmachine · Post by **alienmachine** » Tue Sep 30, 2014 6:02 pm

Hi there,

Actually, I found a big bug in the oscillators core, after all these years...

I forgot to connect ONE cable, what a SHAME! That is why there is so much aliasing at high frequencies.

Anyway, now I think I nailed the true Eminent strings sound. I will post an example soon.

Thanks for the interest

alienmachine · Post by **alienmachine** » Tue Sep 30, 2014 8:28 pm

Here are some rough examples. I will actually really appreciate your feedback. These are raw examples played on my computer keyboard (yeah, I know, but..

)

https://soundcloud.com/alien-machine/osc-1/s-Ksszw
https://soundcloud.com/alien-machine/osc-2/s-xas5T
https://soundcloud.com/alien-machine/osc-3/s-Hl3Zk

Several things wrong on these: 1) The envelopes are not adjusted properly yet 2) You can hear the artifacts when I'm switching octaves (well, it's the PC keyboard anyway) 3) These examples are at the lowest 44100hz resolution.

Thanks!

PD: I added images but they don't correspond to the demo sounds. They are just to show the waveforms. The images show, respectively, 1 - The raw oscillator output, 2 - Oscillator output after being shaped by two 1-pole LP filters, one with keyboard tracking, and another with a fixed cutoff of around 1.9KHz. 3 - The same as 2 - but with the ensemble on.

Z1202 · Post by **Z1202** » Wed Oct 01, 2014 10:18 am

Sounds convincing to me (and I should say ANY other than yours E310 emulation which I tried or listened to the mp3s didn't sound convincing to me). Whether it's "more convincing" than the old one is difficult for me to tell. For one I would need a direct comparison to the old version (and to the original

). Besides, it's also important to try this thing practically in a mix to have a better judgement. One problem I had with the older version is that the sound was somewhat too dense, taking up the "whole space" in the mix, especially with the phaser on (which I thought might have to do with oscilaltor aliasing, but I'm not sure). I noticed that E310 on JMJ's tracks tends to have the same density problem, but to a somewhat smaller extent.

Waiting for the "final" results.

Z1202 · Post by **Z1202** » Wed Oct 01, 2014 10:28 am

Yet another note. IIRC you are using the stereo version of the KleinStein, made by a simple duplication of the mono version, which means that the L and R channels use two different LFOs. At least theoretically these LFOs can get out of sync. It would be better to duplicate the audio processing part of the phaser directly inside the core cell, and make both channels use the same LFO.

Although I believe that in JMJ's setup the phaser is just used in mono, for obvious reasons

alienmachine · Post by **alienmachine** » Wed Oct 01, 2014 3:29 pm

Thanks for your comments.

On the density issue, my theory is that since string machines use pulse waveforms as their basic waveform, most of them anyway, which is then converted to a sawtooth-like by the shapers, process I'm following here, the harmonic spectra is different. In the case of pulse waveforms, the shorter the pulse is, the wider the notches in the harmonic content of the spectrum are, and the more nasal and hollow the unprocessed tone is. Therefore, even though the sound is rich after passing through the ensemble, it does not overpower the mix, because there are more "free space" in the spectrum for the rest of the sounds in the mix. That is why I actually never liked the sound of my emulation.
The samples here helped me a lot to understand this phenomena: http://wutierson.com/en/news/101-string-machine-project
On the phaser, I never though on that, but you are right. The change you are suggesting is easy to implement and it will actually save some CPU cycles

Thanks and I'll keep you all posted.

HunterKiller · Post by **HunterKiller** » Thu Oct 02, 2014 9:54 am

Oh man I've missed all the commotion with the house move.
Hugo these sound pretty darn good so far.

You've even nailed that vibrato sweet spot.
I'd do some testing for you, but my studio is still in boxes

alienmachine · Post by **alienmachine** » Fri Oct 03, 2014 1:57 am

Ouch! House move! I feel your pain. The last time I moved, 6 years ago, I stopped completely doing anything related to music for a while. What's more, since my move I never turned on the computer I used to created the Eminent emulation. I did a backup before and that was it. Now, that PC is laying in the basement somewhere..

Well, finally, I think I finished it. It sounds good enough. Now, I will modify my original ensemble to create the MkII. If anybody wants to test the ensemble as it is, let me know.

HunterKiller · Post by **HunterKiller** » Fri Oct 03, 2014 9:15 am

Yes I tell you if anything is physical it is a house move, esp the darn furniture.

Hey I was thinking this morning, that it would a good idea to include the Eminent plucks functionality from the Multi Ensemble 5500 into MkII. But to extend the range of the pluck length and also give it finer adjustment.

alienmachine · Post by **alienmachine** » Fri Oct 03, 2014 8:57 pm

Hey Hunter,

According to the manual, the percussion (plucks) feature on the Eminent 310 are available for the Solo registration group (Trumpet, French Horn and Oboe) and the Principal group (Diapason, Principal, Salicional, Octave and String). Not for the Strings Ensemble. Of course, I can implement it but I would prefer to leave it for the whole organ emulation, which it is still work in progress

. My first step was to have the oscillators right, which, I think I finally made it

I'll keep you posted, since I would like you to help me test the prototype when it is ready.

HunterKiller · Post by **HunterKiller** » Sat Oct 04, 2014 12:01 am

No problemo maestro

Btw a lot of people would be interested in the full thing, like A LOT. I could find some people on various forums who are quite knowledgeable (many who own it as well) and send them your way.

alienmachine · Post by **alienmachine** » Sun Oct 05, 2014 11:01 pm

If you can contact them to me, that will be great, so I can ask for feedback

While developing the first version I was in touch with Radek Tymecki, from the zoolook.nl groups. I'm really thankful to him, since he sent me samples of his 310 and provided feedback. I haven't been in touch with him for years, actually.

I'm slowly migrating my new core oscs to the old ensemble. I need to change quite a lot actually, and also the graphics, which it's a lot of work.

Cheers!

HunterKiller · Post by **HunterKiller** » Mon Oct 06, 2014 10:36 am

Lol you've got exactly 310 likes on the old ensemble at the NI site.

Hugo, what contact info can I use to send them your way?

brok landers · Post by **brok landers** » Tue Oct 14, 2014 7:31 pm

just found this... a year ago i made this sound... kinda tries to replicate the pad from the pe 2000 through the misstress of the last track of equinoxe... didn´t listen to it while doing it, just out of memory... but the character should be there... and - it´s synth 1.

https://dl.dropbox.com/u/2633218/synth1 ... ttempt.mp3

Jean michel Jarre VST