Qu-Bit Nebulae V2

I’m not sure I get the “disruption” here, nor do I believe the intent of Qu-Bit was ever for the Nebulae to have a “specialized single function.” After all, it does come with multiple firmwares to boot, and users are encouraged to play around with the PD/Csound files. That’s why they upload alt firmwares on their website. Nebulae, both v1 and v2, was created for this purpose and is much much more than just a granular sampler. At its core, the Nebulae is a Raspberry Pi with a panel, knobs, and inputs/outputs that can be turned into whatever the user wants. It seems that Orac is a logical outgrowth of this ethos.

That said, I think that Orac looks cool as an alt firmware, and I’m hoping more info/demos comes out about it soon.


That’s a fair point about the open architecture of the Nebulae though the panel layout and interaction model is clearly optimized for its primary functionality. Orac 2.0 provides a higher level construct for managing interaction between multiple functions on the same module. I do think that is disruptive in the modular context. Unless I missed something Norns lacks that with its single function/script at a time execution model. How interesting would it be to layer Orac on top of Norns to provide modular multi-script patch building and management. Irrespective of disruptive or not it is cool to see this type of innovation in the modular world.

1 Like

Disclaimer: I am not officially affiliated with QuBit, and my thoughts/opinions are my own. However, I did help QuBit out a fair bit when the company first began, and can provide some insights about the development of the first Nebulae.

Some background: I introduced Andrew and Jason to the Raspberry Pi back when it was being developed at Berklee, and also gave them advise for how to get started building the firmware for the Nebulae. I also was the one who got the initial PD support working on the device. (Oh! And that’s me playing the bass and ukulele in the QuBit sample pack.)

So, as for the “intent” of the Nebulae: The original idea behind the Nebulae was to use Csound as a synthesis engine to build a eurorack module. Not a reconfigurable module. Not even a Csound module. Just a module that incidentally used Csound for the DSP. In this case, this module was built around the Mincer and Paritekkel opcodes. I think this design choice reflected CsSpectral, an iPad synthesizer being developed by Richard Boulanger (and us), which used Csound as the underlying synthesis engine. We were very used to the concept of using Csound to build synths.

Shortly after building the initial prototype for the Eurorack, Andrew and Jason published a paper in the Csound Journal on building an FM oscillator using Csound, Pi, and Arduino:


Note how this specifically talks about building an FM oscillator, and not using Csound in general. We were testing Csound on the original Raspberry Pi model B, so there were restrictions on which parts of Csound you could use. Things like sampling and FM synthesis with limited voices were all that you could really do. I actually remember being very impressed that they could get something as complex as partikkel running on that thing. AND Mincer!

I think extendability was always something that was going to be there, but as sort of an afterthought. Sort of a “it’s there, so we might as well” kind of thing. It was full-on Csound. It was running Linux. Why not? The original concept there was to allow people to just load their own Csound patches. PD support only came after people were requesting it. It was never part of the original vision. Luckily, it was just Linux, so most of the groundwork was in place.

Linux, by the way, was definitely a means to an end. It was an easy way to get Csound running on an embedded system. If we had figured out how to get Csound running as some kind of realtime OS, we’d probably would have gone with that. The big picture idea at the time was to get Csound into Eurorack, and that was basically it.

The Nebulae was designed as a phase vocoder + granular synthesis module, kind of like how the TB-303/TR-808 was built to be a band-in-a-box for singer/songwriters. QuBit did a very good thing when they decided to let people reprogram the Nebulae. By making the Nebulae hackable, they gave it permission to grow and evolve beyond the intentions of the original creators. I imagine people will find new uses for it for years to come.


Since you seem to have influenced the use of the Pi, do you know why Qu-Bit decided to use the ‘normal’ form factor rather than the compute module? I can’t help thinking it looks like an odd design choice with the Pi slotted onto the back of the module.

1 Like

It’s likely that the original QuBit Nebulae was in development before the Compute model came out and they stuck with it?

The compute didn’t really exist at that point (we were doing stuff around 2012 right when the Pi first came out). Even if it was available, I don’t think we’d know how to use it. We were music students at music school with no engineering backgrounds, so there was a lot figuring stuff out as we went along. While less elegant, using the Pi board gave us a simple solution which yielded more knowns than unknowns. It got the job done quickly.

QuBit has of course evolved quite a bit as a company, and their manufacturing process has matured since the days of hand-soldering Nebs in a tiny Boston apartment. They of course know about the compute now, and they do have the means to use it, but yeah I think it just made more sense logistically to stick with the original design.


Ah okay. This is all very interesting, thanks for taking the time to answer my questions @PaulBatchelor

1 Like

Hello everybody.

I’ve received my Nebulae V2 two weeks ago and I really enjoy it. I understand how the module works and how I can load new samples but I never found how to use “live source” and the “record mode”…

The button “Source” never hold…
Do you undestand what i’m doing wrong??

Thank you so much.

it needs a tap to switch mode.
this is because if you hold it, it provides access to other (secondary) functions.
(i.e. acts like a kind function button)

(its not particularly sensitive to this, once you’ve done it a couple of times its easy)

Yes, I understand that if I hold “source” button I will enter in secondary mode but when I tap one time the source button the “file” button light one time too and the both turn light off and nothing happen… Nebulae still playing samples loaded into the module…

However, when I tap the record button, he’s keeping hold (light on) and I’ve got the two red lights aroundt the speed encoder but I can’t ear if I record something…

Ok… Sorry everybody… I just find the solution for the “source mode” : push hard this fu##ing button!!!

Mine is maybe strange because I really need to push it hard one time to go one the “source mode”… but its ok, it’s working now and it’s open a crazy world for me!!! :wink:

I got a Nebulae V2 a few weeks ago, and in general I’m loving it. One thing I’ve noticed that is quite quirky is the recording overdub mode. I downloaded the new firmware so it will keep recording over a loop until you turn it off, however when I try to do this I’m hearing all of the these pops and clicks and it appears that the loop length gets altered too. I’m used to good old fashioned guitar loop pedals where you can just overdub to your heart’s content without any clicks or pops and this doesn’t seem to be working this way and its quite frustrating. And no, I’m not in the granular mode.

Hey just following up here…
So I’m still really loving the Nebulae, the one main point of contention is this: I want to be able to mute the live audio input that’s going the output of the Nebulae and still be able to add live audio to the buffer without monitoring the live input to the output. The way I’m mostly using my Nebulae is to pull live audio from various channels of my mixer through an aux send, and there’s enough latency of the audio coming the Nebulae that I’m getting phasing and doubling artifacts when I sample live audio.

I know nothing about cSound or how these alt instruments work, but I do a bunch of front end web coding for my job. Can anyone point me in a direction as to where I should look in the code to try make this change for myself?

Shouldn’t it just be as easy as a mute function for the live audio to output?


Looking at the Nebulae Code here:

There’s a variable called “ablenddry” which i’m guessing handles the dry sound and then on line 898 and 899, this bit of code exists:

aoutl = amixl + (ainl * ablenddry)
aoutr = amixr + (ainr * ablenddry)

Guessing if I just remove the “ablenddry” this may work to remove dry sound from output, no?

Is it just as simple as making a copy of this instr file with changes and loading onto USB stick onto the Neb?

Not a Nebulae 2 owner yet (want one eventually though). I must be missing something in what you want to do. Why doesn’t leaving the blend knob fully to the right or left while in live mode work for you?

When you leave it fully to the left or the right it doesn’t add the incoming audio to the loop. It mutes it on the input, not the output. I basically want this to function the way any standard guitar looping pedal would function, where you don’t hear the looped audio until it cycles back around for the first time. Right now its including the ‘dry’ audio into the mix, which really isn’t dry, because its because processed digitally and has a lag. When you add that on top of that audio in my mixer, it creates a weird phased or slapback that doesn’t sound good.

1 Like

Are you 100% sure? That seems like a bizarre design choice that makes me less excited about potentially getting one. I’m just going off the manual which reads:

“When Source is illuminated this controls the blend of the phase vocoder, granular stream, and dry input signal.
When the knob is set to the middle, only the dry signal will output.
When the knob is set fully counter-clockwise, only the phase vocoded loop will output.
When the knob is set fully clockwise, only the granular loop will output.”

I just assumed it would still add stuff to the loop but like I said, I don’t have one to test. Seems like a huge oversight to not have a 100% wet signal achievable with the stock mode.

Unrelated do folks here sometimes have their Nebulae V2 freeze on start up (stuck in the bootloading phase)? I bought mine used and even after updating to the new firmware I’ve been seeing the issue and I’m not sure what to do if I bought mine secondhand :frowning_face:

1 Like

You def get the ‘dry’ signal, but Im putting it in quotes because its not really analog dry, its still being processed through the A/D convertors and theres latency. This is not that big of a deal if you process audio in series, through the Neb. However, if you process it in parallel through an aux send of mixer, theres already a dry signal present on the mixer and if you can’t remove the dry signal from the Nebulae, you get the true dry signal, the delayed dry signal and them the effected signal. It’s annoying.

1 Like

This happened to me too. I eventually had to send mine back and they replaced it. You should reach out to Qu-bit.


Oh thats a defect? Interesting.