List of weird things i've noticed on my unit

[Duchemole]

Hi
first of all, thanks for building such a nice instrument, that matrix feature is a sort of "dream comes true" idea, but sadly, in my opinion, the synth fails (at the moment) its aim becoming the ultimate monosynth, i feel most of these (if not all) could be improved, it really deserves it
i just want to say that i have an important experience about synthesizers, i'm a professionnal musician, and i use them daily for decades
since the first day, i've noticed some funny things using the Matrix brute (got it on 11/01/17), i've reported some of them to mr Usson, but i wanted to have more experience using the instrument to make a complete report
also, i have recordings of most things i mention there, and can provide them to you whenever you want

- Tuning, i have to say that i'm pretty surprised that the synth does not have background tuning, as it reacts exactly the same as the Andromeda, that becomes TOO in tune, then, vco don't drift that much, otherwise, i think it would have been great to have the option, on the software center to choose how accurate it would have to keep in tune vcos (it contributes to give a slightly stiff sound to the instrument)
also, vco 1&2 are not properly tuned on my instrument, compared to the vco3, when they're on 0, they're too low, i need to put both knobs on 2 to have everything in tune, is it due to bad calibration or something ? (then, most presets sound a bit out of tune, but i don't mind as i never use them, it's just an obervation)

- i have to say that, i never have to use the auto tune feature (i've tried it sometimes), as even if the synth takes a long time to be in tune, it's very stable and in tune once that point is reached, once, i've tried to auto tune the synth after few minutes switching it on (to let a friend hear it), well, it was terrible and was awfully out of tune (about a 5th on a vco), even after 20mn, it was not better, i have to do a power cycle to have it properly in tune

- VCF even being wide open don't let high frequencies to shine as expected, i've made SEVERAL comparisons with many instruments, including minibrute, difference is huge (i've mesured that it drasticaly drops from 7khz), to me, both that lack of high frequencies and drift contribute to hide the real potential of the instrument

- pitch bend can only modulate vco pitch, and as it's not part of the sources in the matrix, it's not possible to change anything about it, it's not a major issue to me, BUT, on ALL synths i'm thinking of, you modulate the cutoff with the pitch bend, as exemple, if you try to play a sine wave using a filter as generator, it's not possible to bend it (but it is possible on the minibrute)
also, if you start to self oscillate the filter, while still hearing vco behind, it makes a really strange effect, as you bend only the vco (maybe it would be nice to have the choice to choose the option on the software center ?)

- While talking about self oscillation, on my unit, i have to use 96 value to have about equal temperament and play sine wave, but i have to tweak a bit to have it properly in tune, maybe it would be nice to have an easier way to get it ?

- arpeggio + sequencer / split points/octave, none of these combinations work, i've seen a post about the arp + seq one, just reporting ...

- about the arpeggio, it would be nice to have the ability to arpeggiate respecting notes order, and dynamic, many synths have this great feature

- also, i don't know how it could be implemented, but it would have been nice to route the sequencer to different vco, as example, 16 steps to each one, that way, using different filters and the 3rd enveloppe to modulate the other vcf2, it could quickly sound as a huge modular system, keeping only the "small" 3vco + 2vcf structure

- i think it could be nice to have the choice to get the external signal non filtered (as example, when no leds are on, if user does not want the signal, he just has to turn the knob off)

- i have noticed a strange problem on VCO 1 & 2 tunings, when making a patch with very small drift, when playing, one of the vco will stop drifting randomly to get the tuning of the other on (no, sync button is off), when i cancel the vco (by pushing filter switch on the mixer), when i switch it back, it drifts a bit, then stops (and sometimes starts again to drift, i have recorded this failure, noticed that the problem could be even worse when the synth was on for several hours)

- about 1 second to change a patch, considering how powerful cpu are nowadays, i think it would be possible to speed up that delay ? (honestly, such a thing would prevent me using the matrix brute on stage, ever ... but maybe it's not a bad thing, considering its weight and size ... maybe it's more a studio instrument ...)

- 2 small POK POK noises when changing presets

thanks for reading all these points, i hope they will help to perfect that great synth (feel free to ask recordings of any of these failures if needed)

[BobTheDog]

The tuning problem sounds weird have you tries the autotune (PANEL+KbdTrack) after it has warmed up (10 minutes for so)?


I just tested the filter cutoff points with the blue noise and on mine the drop-off seems ok, I have included two images. One with res up full and one with res down.

[BobTheDog]

- i have noticed a strange problem on VCO 1 & 2 tunings, when making a patch with very small drift, when playing, one of the vco will stop drifting randomly to get the tuning of the other on (no, sync button is off), when i cancel the vco (by pushing filter switch on the mixer), when i switch it back, it drifts a bit, then stops (and sometimes starts again to drift, i have recorded this failure, noticed that the problem could be even worse when the synth was on for several hours)

This seems to be a problem with the fine tune knobs, there is a dead-zone in the middle and the knob can jump back into the dead-zone. The fine tuning knobs are not good for getting the oscillators slightly out of tune.

If you take the mod wheel in the matrix and map it to osc 1 pitch with a mod strength of 1 and put the mod wheel up you can see that there is much finer control of the pitch available, and it doesn't jump back.

Maybe arturia can look at changing the scaling of the fine tune knob so we get finer control near the centre.

[Duchemole]

The tuning problem sounds weird have you tries the autotune (PANEL+KbdTrack) after it has warmed up (10 minutes for so)?


I just tested the filter cutoff points with the blue noise and on mine the drop-off seems ok, I have included two images. One with res up full and one with res down.

ok, maybe i exagerated a bit in my statement, but there is a serious loss in high frequencies
here are 2 low C saw wave, one from the minibrute, the other one from the matrix
it's very easy to hear the difference between (since i have enough space, i also send you an umcompressed file, i had to produce the square with the sub osc on the minibrute, as it seems to be badly calibrated, and starts with a slightly rec signal)

about the tuning failure, as i said, i tried to run the auto tune probably to early, as the instrument was not warm enough, but still, it went completely mad, i could not nothing except power it off/on

thanks for your interest

(for some reasons i don't understand (3 files, under 2mb total size), the browser refuses to add attachments, when i push "post", it crashes, so here is a wetransfer link)
https://we.tl/Lb0H7sBjcX

[BobTheDog]

My MB sounds exactly the same as yours in your audio example, as you say there is definitely more high frequency content from the mini brute.

I wonder if it is the filter cutoff or slightly different oscillator shapes, maybe someone from Arturia can help us out?

[BobTheDog]

I guess you could try the output of the mini brute into the audio in on the matrix brute and see if it is the filter.

[F5D]

Have to mention that I also noticed that the oscillator waveforms are not as "sharp" as some other synths that I have. I can get nice sounds out of the synth, but when turning the Low-pass cutoff frequency up, I often check whether I still have something limiting the cutoff frequency in the matrix or so. Maybe it is just the character of the synth, slightly damped at high frequencies. Nothing that would prevent programming nice sounds with it, but being such a radical synth I was expecting the oscillator waveforms to sound really bright too, if needed.

One day, somebody who opens the synth and inspects the circuit will find the simple RC-filter that does this, and by changing the values will get full spectrum out. Could be in any part of the synth actually. I am also wondering what kind of filtering did they use with the BBD fx. It sounds a little bit cleaner compared to some other BBD fx that I have. One possibility is that there is some leakage from the BBD fx that they had to filter out, resulting in the slightly attenuated high end. Some BBD fx leak the clock and sound really weird, in a nice way. However, in the MatrixBrute, the BBD fx is parallel to the dry signal and is filtered independently, so this is possibly not the reason here.

[Duchemole]

I guess you could try the output of the mini brute into the audio in on the matrix brute and see if it is the filter.

and that's a pretty good idea (except that i don't have the minibrute to make the test, as my friend has brought it back home)
so, i made that test:

i've recorded 4 tracks, with the same VCO (dotcom)
- track 1: raw vco output (sawtooth)
- track 2: that vco through moslab LPF wide opened though VCA (dotcom)
- track 3: raw vco through matrixbrute ladder filter 12db
- track 4: raw vco through matrixbrute ladder filter 24db

i have been surprised by the results, as i did not have more high frequencies loss when comparing track 2 to track 3/4 (there is a minor difference when switching filter 12/24 db)
does it mean that the problem is about the vco ? not the filter or the mixer ?

[Duchemole]

we are currently sharing some recordings of our units with a friend of mine
his vco sound even darker than on my unit, but we've noticed a very strange thing
on his synth, VCO 2 makes a ramp, on mine it makes sawtooth

[BobTheDog]

Interesting, I guess mines what I would call a positive ramp.

Here are Saw, Square and Triangle:

[F5D]

Just captured the waveforms of the VCO 1 & 2 of my MatrixBrute. In my synth, both VCOs produce the same looking waveforms, so I post only the waveforms of VCO1 here. It seems that the waveforms look the same when using the Ladder filter in either 12 or 24 dB/oct modes. When using the Steiner filter in 12 or 24 dB/oct modes, there is a slight difference in shape, the 24 dB being slightly closer to the Ladder. I post here screen captures of Bitwig oscilloscope device with VCO1 waveform full level, VCO1 mixer full level, Drive, Brute factor and Resonance of the filters at zero, filter output at full level, also the synth output at full level. These two filter modes represent the brightest (Steiner 12 dB) and most damped (Ladder 24 dB) low-pass filters available in the synth.

Edit. Added also VCO1 Saw with 100 % drive with both filters.

VCO1 Saw through Ladder LP 24 dB/oct fully open, Drive 0 %:


VCO1 Saw through Steiner 12 dB/oct fully open, Drive 0 %:



VCO1 Saw through Ladder LP 24 dB/oct fully open, Drive 100 %:


VCO1 Saw through Steiner 12 dB/oct fully open, Drive 100 %:



VCO1 Square through Ladder LP 24 dB/oct fully open, Drive 0 %:


VCO1 Square through Steiner 12 dB/oct fully open, Drive 0 %:



VCO1 Triangle through Ladder LP 24 dB/oct fully open, Drive 0 %:


VCO1 Triangle through Steiner 12 dB/oct fully open, Drive 0 %:

[artElect]

Yeah its interesting that they changed the Oscillators that much from the Minibrute. I did a side by side comparison and the most obvious changes were:

- the Sub-square wave is duller on the Matrixbrute. For me the difference to the Sub-Sine is way to subtle...
- the Triangle Metallizer sounds much thinner compared to the Minibrute. I don't understand why they changed that so drastically, because this was a monstrous sound on the mini and one of the best sounding triangle mods I know
- the ultrasaw also sounds a little bit more lifeless and flatter on the Matrixbrute
- overall less high frequency content on the Matrixbrute. This makes it less aggressive and less "brutal", the Mini screams more. Maybe Arturia wanted to get a "warmer - softer" sound in general on the Matrixbrute, because there were many people complaining about the harshness of the mini (which I never understood - I liked that industrial sound).

Don't get me wrong, this is a really fantastic instrument. But it does not cover all the sounds from a minibrute, it has a different character.

[radray]

As far as high frequency content is concerned have you switched the filters to high pass to see if it changes?  A good test would be using a harmonically rich wave like a saw or square and look at it on a spectrum analyzer display.

Set it up with the low pass engaged fully open take a screen shot then switch to high pass fully open and compare screen shots. This will let you know if the high frequency filtering is caused by the low pass filter cutoff calibration.  If there is no change in upper frequency content then it could be fixed filtering somewhere else in the circuit path. 

If it ends up being the low pass filter cutoff calibration it could possibly be changed with firmware or a trimmer on the circuit board or it might only be able to be changed by changing some resistor values. It all depends on how the system is designed so it's hard to say without the schematic.

[SoundRider]

Didn't Arturia tell us at the early presentations at NAMM 2016 that the MatrixBrute would not be a copy of MiniBrute circuits with add-ons? I think they did.

It's also my impression that the MatrixBrute has been "tamed", but not to the worse. Some of the raw aggressiveness is damped. The sound is less raw or brutal - compared to the MiniBrute - but it became more flexible and refined. I would guess that Arturia wanted to address people, who are in classical bread-and-butter sounds too. That was not the main focus of the MiniBrute, rather the opposite.

I compared the MatrixBrute to the Moog Voyager ... not scientifically ... but to see (hear) the differences. The MatrixBrute is even with the ladder-filter very different ... and that was one selling argument more to me

I see the strength of the MatrixBrute sound-wise, where it's really different compared to many others and this would be:
 
- having two "Brute VCOs" with true exponential slopes (see the oscilloscope images in the previous posts)!
- having three VCOs
- having FM for different VCOs and filters
- the Steiner-Parker filter,
- having two "full independent" filter paths,
- modulation capabilities and
- the capability to modulate modulations.

This opens quite some interesting territory of sound design, which other - sometimes expensive (non-modular) - synths can't follow.

I still love the untamed rawness of the MiniBrute and its - sometimes - chaotic nature of the resonance and the BruteFactor. Now I definitely know that I will keep my MiniBrute.