Forum for all other samplers & synths such as Maschine, MVs, Akai S & Z series, Roland, Korg, OP-1, analog synths etc.
User avatar
By richie Wed Mar 21, 2018 11:36 pm
There needs to be a better review made of it. In my opinion, to no fault of his own - didn't really have enough knowledge on the s950 be able to give it a fair review.

It needs to be reviewed by someone that actually has a S950 to do a fair comparison.
By head034 Fri Mar 23, 2018 8:30 pm
I have an s950 and used it for many many years! I bought this plugin and like it alot! For a digital reproduction im very happy with the result.
User avatar
By distortedtekno Sat Mar 24, 2018 12:50 am
I owned an S950 that I sold 10 years ago. This plugin sounds convincing enough for me to get over my seller's remorse. So I'm gonna try out the RE this weekend and share my thoughts when I get caught up. I got a good feeling I'll buy it before my trial period ends.
By Knuckles5000 Mon Apr 09, 2018 1:54 am
I don't know if anyone would be interested in a shootout between this plug and the actual 950 but if so I could probably figure out how to upload one. From just screwing around with it it seemed much closer to anything I've been able to get out of other plugins.
By Knuckles5000 Mon Apr 30, 2018 11:09 pm
Had to work out some issues with my setup and do real life **** but have started working on the comparison. Not done yet but one thing that's surprising me is that the RX950 plug doesn't seem to replicate the frequency response of the 950, at least not entirely. For something that supposedly took a year of study and analyzing circuits I'm surprised that would be overlooked.

Without any documentation of how the plugin works I'm having a hard time figuring out if there is any way to get the plugin settings & daw gain staging anything like how you would actually sample into something, or if that's even the best way to use it. But for all it gets right (I do think it's as good as anyone has gotten thus far) it seems to lack that mid-high bright glassiness at full bandwidth (I could never figure out where in the chain that is happening in the first place).
By Cockdiesel Tue May 01, 2018 12:24 am
Thanks for checking this out for us. Can’t wait to see the results. Most interested on how the filters line up, and seeing how good these emulations are getting or where they could be at in five years.
By Knuckles5000 Sat May 05, 2018 11:04 pm
An extremely unscientific frequency test using a sine sweep and Voxengo SPAN, hopefully not totally **** up:

Original Sine Sweep:
https://imgur.com/gRp3IeL

Run Through S950 at full bandwidth:
https://imgur.com/pCec5bL

Run Through RX950 w/ default gain settings (full bandwidth):
https://imgur.com/p8O9jnQ

Run Through RX950 w/ input and output gain bumped up (full bandwidth):
https://imgur.com/4JSmJbd

Run Through Homemade S950 (at full bandwidth) Impulse Response:
https://imgur.com/lh57FHC

If I'm understanding these correctly it means that RX950 does not exhibit the classic 12-bit Akai frequency characteristic at all, which is kind of odd considering how much effort seems to have gone into modelling it (judging from their instagram with 3D models of the actual volume knobs and such). The plug sounds much more convincing at lower bandwidths, probably because of the steep anti-alias filter. Audio comparisons coming soon.

An interesting thing doing this test is that the 950 input meter spikes way up at those high frequencies as they are being fed into the machine. I don't know if this provides any clue as to when in the signal chain they are being boosted.

By Knuckles5000 Sun May 06, 2018 11:30 pm
Full bandwidth:


Half bandwidth:


Quarter bandwidth:


In all these the first one is the S950 and the second is the RX950 Rack Extension in Reason 10, except the full bandwidth test where it's original sample, then S950, then plugin.

Maybe I'm just a weirdo but even at half bandwidth I miss the brightness of the actual machine. I feel like the plugin does a very good job modelling the things it actually models, but the fact that it doesn't give you at least an option to have that frequency response or w/e it is bums me out. Maybe I am using it wrong?

Saturation and Filter sweep tests next.
By Cockdiesel Mon May 07, 2018 3:15 am
Mad props for this. Excited to see how the filters match up. I’m sure there may be a way to dial them in closer but I see a huge difference like you do with this. At quarter bandwidth the plugin actually sounds like **** to me.

When I get time I’ll try something similar with a dimension d and md500.
By Knuckles5000 Mon May 07, 2018 9:45 pm
Cockdiesel wrote:I’m sure there may be a way to dial them in closer

I imagine there is but I'm having trouble finding it. Part of the problem is there's no documentation or manual whatsoever (supposedly it's coming soon, and I'm hoping it goes into more detail than "turn up the gain to make it loud" or w/e). If I figure out I'm doing it all wrong I'll redo the test. From what I can tell the input gain knob is somewhere between an actual gain knob and a saturator modeled on the 950's preamp behavior.

The unity gain is way over to the left (I am showing my noob ignorance here not knowing what unity is on the 950 rec knob, as I said not a scientific test), and at some point as you turn it to the right it starts saturating, until it's totally unlistenable at max. EDIT: On further testing (comparing peak to RMS) it turns out it does respond to the level of the input signal you feed it. I must have had my gain staging **** up before. Feeding it a drum break peaking at -12dbfs it started to compress a tiny bit right before 12 o'clock, and then more and more the farther to the right you turned it. Feeding it a really quiet signal with input on max still compresses it a fair amount though, so I still don't quite understand the correlation. Feeding it a way too hot signal will compress & distort almost immediately once you start turning the knob, even with the output gain attenuated so it doesn't go into the red.

There is a little tiny Reason-style level meter (you gotta squint) in the top-left corner that goes from green to yellow to red depending on EDIT: the volume of the input signal, but I am having trouble figuring out where the "tapping the red" equivalent sweet spot is, cause like I said it's **** tiny. If I was smarter with this **** I'd be able to do it by ear but they start playing tricks on me after like 15 minutes of obsessive fiddling.

None of this is to say I don't like the plug however. I think it's very good for what it is, and for someone who doesn't care about this level of OCD **** and just wants to mangle some breaks and be in that ballpark of Akai 12-bit sound it would do the job.

I think this is part of why emulation is still where it is tbh. Modelling these old dinosaurs is hard, people who have them don't need plugins, and people who don't have them generally don't need that level of accuracy in a plugin (I use MPC Software MPC 60 & 3000 vintage modes and Maschine's SP-1200 & MPC 60 modes on pads, how would I know whether they're 100% accurate, I don't have the machine to compare).

EDIT: Turns out I said some BS before, corrected where noted
By Knuckles5000 Wed May 09, 2018 9:39 pm
Here's the filter comparison:


I kinda ****ed these up timing wise, as RX950 allows you to do real-time filter sweeps and the actual machine doesn't (unless I need to RTFM again). So on the plugin one I guess I was turning the knob too fast. It should still give an idea of how the curves sound thought. At a casual listen it seems like they did a pretty good job.

EDIT:

Here's a bonus test:


One of these is the S950 sample and one is RX950 and Reason devices. Which do you think is which, and which do you prefer?
By Synthiemant Sun May 02, 2021 10:33 am
Knuckles5000 wrote:An extremely unscientific frequency test using a sine sweep and Voxengo SPAN, hopefully not totally **** up:

Run Through Homemade S950 (at full bandwidth) Impulse Response:
https://imgur.com/lh57FHC



Would you be willing to share your homemade S950 Impulse Response with the community please?