How can we simulate an audio output in dRambo?
Basically I’m sending an oscillator to a dRambo oscillator set to PULSE and sending
the exact same oscillator signal out via a usb audio interface into a CEM3340 oscillator.
The thing is the oscillator being used to modulate both the internal and external oscillator
stops being the same once it reaches the real world, other factors arrive such
as the D/A that the usb interface uses, the voltage itself going into the CEM3340 chip and
another thing that comes into play is the CEM3340 chip itself but that’s a whole other thread in itself.
I digress,… I started thinking that not only does the dRambo oscillator modulate in a different way
to the external oscillator, it also occurred to me that even the oscillator shape itself plays a part
in regards to said modulation.
So on that note, is there a way to simulate an audio output in dRambo
to further experimentation in basic analog modelling?
I’ve tried the Shaper module and that’s got me much closer, any suggestions?
Screenshot for the basic usb audio interface simulation.
It takes time for the signal to go out through the D/A’s hence the Delay module
and the Shaper module adds BIAS which makes the PULSE oscillator shape less square
and finally the AMP module to attenuate the osc signal as the signal is hot.
It's worth finding out what really happens in your setup and Drambo's new precise scopes can help a lot here.
No matter if it's analog or virtual analog (i.e. digital), the mechanisms of synthesis and modulation are the same and there's usually not as much influence from real analog components as you might expect - the difference you've noticed is more likely caused by other factors that you might not yet be aware of. DC offsets, noise, different levels, hum from bad shieldings etc can often change the sound more than the ADC/DAC, filters and OpAmps.
Before trying to simulate anything, I would first measure the hell out of this setup to get a clear picture about what I actually need to do.
I’ve been using dRambo’s new oscilloscopes to look at the waveforms of the oscillators.
Sending them through an external compressor changes the shape so much and you can even see the noise introduced.
In regards to my rig, it’s been calibrated to -12db on all inputs and outputs
and when I’m experimenting I will push that to -7db.
All inputs and outputs are balanced when they can be and D.I’d when they can’t.
I’m quite particular about noise so I try to make sure my noise floor is within an acceptable range,
at least twice a month I will go through most of the inputs and outputs to check for distortion using a sine wave.
I don’t have a real oscilloscope yet but that’s on the horizon.
On another note I noticed an offset happened when I was dialling through the PULSE width modulation
on the CEM3340 so I did this thing to achieve something close to what I was seeing visually on the oscilloscope.
At a certain point the volume of the CEM3340 increased and then decreased so I used
a Graphic Shaper module to set the offset whilst using a 2nd Graphic Shaper module to
increase and decrease the amplitude of dRambo’s oscillator.
And the other Graphic Shaper module for the offset placement.
If you get an offset while changing pulse width, that sounds like your CEM3340 is AC coupled with at least a capacitor in series somewhere in the path. Could also just be your audio interface, they're usually AC coupled too.
yes, if u get "really different stuff" it points to da/ad something ?
while playing with this remember that what comes out of your speaker is a different story than what the scope shows you
remember a single sample plop (Dirac impulse) will result in something like a sine wave with envelope out of your speakers because of its physical limitations ...
also remember your ears can't tell if its saw up or down , that make looking at scopes quite weird ...
not sure how PWM is done here
sometimes it doesn't go into "nothing" and back
(i find that more musical)
FWIW WTosc does PW similarly...
two examples here...
1) an ‘imperfect’ Pulse generated by Spline (generate) and than positive and negative PW is achieved via Spline (phase shape)
2) stock Pulse with PW effect
In both case PW will act similarly to what you see from external OSC
If the ‘shape’ of the sqaure/pulse-wave is important pass it thru an all pass filter and adjust the filter frequency and observe the change in shape.
While the ‘sound’ of the square wave doesn’t change the altered shape has quite an impact when used for modulation and lower frequencies.
Just having coffee. :)
Nice one this was going to be my next question in regards to the Wavetable oscillator.
im not sure if I would still call this square
See pulse and saw of same frequency mixed ...
so if you have bleed through on the osc level ... it would look exactly like this too
if we invert the saw it looks closer to @gravitas 1st screenshot
only looking at scope is tricky business ...
@lala the earlier screenshot is a screenshot of the oscillator shape generated in dRambo.
The actual CEM3340 shape with no filter, modulation or effects monitored
via a Komplete Audio mkii audio interface is as follows;
to the left,
and to the right.
@rs2000 I think the KA mkii has DC coupled inputs as well as outputs though
it doesn’t have much of a useable output voltage without an external amplifier/booster.
this is interesting
it looks like saw down and then saw up ?
it won't sound different if its only saw up or saw down ,
but it will modulate differently ...
that’s what I thought also.
Here’s another trio of screenshots again left, center and right
this time however it’s the direct signal from the oscillator itself.
The earlier trio of screenshots showed the signal coming out through the VCA.
Note that in the center position it’s slighter more square shaped.
maybe its this?
if u turn the level of the triangle down its not so extrem ...
@lala the CEM3340 has outputs for square/pulse,
saw and triangle so maybe there is bleed on the chip itself
even though on my version I don’t have a direct out for the triangle output.
it must be triangle bleed trough
I think the same also.
drambo is so cool
I can say ah its this and that - make some wild assumptions
and prove it to myself a minute later :) 😂😎
I’m having fun at the moment playing with oscillator shapes and what audio modulation does to the sound.
I always wanted to know how that "weird analog looking pulse" happens
mystery solved :)
the analog saw is the same triangle bleed through ^^
It’s not bleed through... in most analog synths perfect shapes don’t or rarely exist... skewed triangles, weird looking sawtooth, pulses like above... just look at moog model 15 waveforms... :)
oh wow, wouldn't have thought its that easy 🤭🤫
as far as I understand it a lot of osc have a triangle core and do the other waveforms by wave shaping? ...
so its no wonder to me that there is some bleed through from that
a lot of things that I couldn't explain to myself came from all kinds of real world sluggishness like simple bleed tru and feedback ^^
When I finished my subharmonicon attempt I ended up with the conclusion that D stock OSCs are a bit too sterile for that harmonically rich application and started playing around with WTosc to get something warmer out of it... mostly based on model 15, also trying to get the ‘past noon’ overdrive thing baked in... we’ll see at some point... was busy in a past few weeks with moving back to UK🤞
they could be made of triangle, but it’s intentional... I mean most moog synths share this or very similar shape for their pulse and sawtooth waveforms...
I think they are a bit heavier on bass
hm, looking at the spectrogram
what the bleed through does is mostly changing level of 3rd and 5th harmonic
You’re back in the U.K? Wicked.
You know there’s got to be a dRambo meet up at some point in the real. :)
Found this wiki for the CEM3340
hm, how to tune that thing?
the Curtis chips I know you had to turn on for half an hour, then press some weird key combo to autotune
and it would tune the osc and filter then to be in key
u just wait and tune it? no autotune?