MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Are All Waves Equal..?
MUFF WIGGLER Forum Index -> 5U Format Modules  
Author Are All Waves Equal..?
josaka
Are all different oscillator waves equal between different makes of oscillator..?

a moog sine sounds the same as a .com sine ..?

an Arp saw is the same as a Corsynth Saw ?

do you think they are the same tone/timbre saw for saw/square for square?
or that there is nothing between different raw sine/saw/squares ?
if you dont think they are the same which waves on which oscillators do you drift towards and why..?
Dave Peck
There are certainly differences between the actual wave shapes in different makes of oscillators. Some of these do have a noticeable effect on the sound.

I don't have any particular 'favorites', though.

I remember there was a thread quite a while ago that included screen shots of several different sawtooth waves from different synth models. Quite interesting to see the differences.
m0n0mania
I wonder about this too, if that's the case, what's the magic mojo sauce in the more expensive Moog clone oscillators? Is it the degree of out-of-tuneness/drift?

Without disappearing into the "it's the whole signal chain, distorting CP3's contribute to the Moogyness" etc discussion again, if a waveform is just a waveform, couldn't you use Doepfer VCOs in your IIIc and get the same sound? (signal levels matched, power supply differences etc ignored for simplification).
Xero
moog isn't even moog...lol...which moog....901...921...model d old osc...model d new osc....etc, etc...

arps have this ability to get this certain nasally tone that just screams arp, not sure what it is exactly. just play one in person and you'll hear it eventually. i'm sure it's more than just the oscillators, it's the combination of the oscillators, the various amplifiers, mixers, etc, etc...and it doesn't even matter which filter....there's just something going on there.

its why you've got clones of moog mixers and stuff like that as well as the oscillators...
Dave Peck
m0n0mania wrote:
I wonder about this too, if that's the case, what's the magic mojo sauce in the more expensive Moog clone oscillators? Is it the degree of out-of-tuneness/drift?

Without disappearing into the "it's the whole signal chain, distorting CP3's contribute to the Moogyness" etc discussion again, if a waveform is just a waveform, couldn't you use Doepfer VCOs in your IIIc and get the same sound? (signal levels matched, power supply differences etc ignored for simplification).


Pitch tracking behavior is an important factor in the sound of multiple oscs but that is different from this topic about differences in the sound of one osc versus another osc due to their wave shapes. For example, some sawtooth oscs have a curve near the top of the wave, some are more mathematically correct, some have odd lumps and bumps in the shape, etc. And all of these differences in the shape translate directly to differences in the harmonic content and thus the sound.

If you listen to a Dotcom sine, you can clearly hear a fizzy buzzy timbre added to the basic sine tone, audible in the background, and you can see this on a scope. It's a tiny sharp transient peak in the waveform, which is not present in other more 'accurate' sine oscs.
KSS
The question you didn't ask is the easiest to answer. Yes, identical wavshapes of identical amplitude and DC offset will sound the same. From there it gets muddy fast.
When was the last time you had a perfect waveshape?

Quote:
Are all different oscillator waves equal between different makes of oscillator..?


Definitely not. But the answer to this is bigger than only the waveshape itself. But let's start there. There are essentially three different means to achieve the standard waveshapes used in subtractive synthesis; ramp/saw core, triangle core, and digital.
Each of these is better at some things and worse at others.

The first can immediately be divided by whether its wave is created by rising to a reset or falling to a recharge. The ramp or saw of the description. Does this matter on its own? At audio rates the two should sound identical, as we don't perceive the change in phase the way we do a change in frequency.

The second has the same possible 'inverted' output one triangle from another across different manufacturers, and like the saw ramp will not be noticed by itself.

The digital means of waveshape generation almost always requires some form of filtering on the output to deal with the granularity of the discrete steps used to form the desired waveshape. As technology improves this granularity is getting smaller, and the filtering relationship changes in kind. Most filters are not perfect and therefore the sound of a given waveshape may not sound the same across its own range, let alone how it compares against another makers oscillator.

And here we begin to see why the answer to your question is so much bigger than the question, and how misleading the simple response often seen in this recurring question fall short. Because there are a large number of variations both within the waveshapes themselves, even within the single maker. And even more when one looks past the wave itself to how it interacts with other elements in the typical synth chain.

The easiest to notice is a pulse wave. Because analog generation of pulse waves is derived most often from the core wave, we see an immediate difference between that result when the core is saw, ramp or triangle. When the usual comparator is creating a pulse from a ramp the pulse grows wider from the far end back towards the near end.

For a saw this is reversed. For a triangle it gets wider in equal measure from the middle.

Agan, these phase relationship changes are not necessarily noticed when the waveform is sounding raw and complete in itself. Digital cores also typically follow a similar derivation, though this is not a requirement for them. So they often share their result with the analog type they've chosen to model.

But if we may not hear the difference in width changing natively, we probably do as soon as this lone oscillator is mixed with something else. Or filtered. Try it. Run your different oscillator outputs through a comparator and notice the differences when you mix or filter them. Then notice the differences when these results are used as CV for FM or LFO duties. Though not often seen in typical analog VCO's, run a sine wave though the comparator and notice how its non-linear width change is easily heard against the linear change of the other three methods. And even those are only as linear as the native waveform being compared.

With that simple observation using only one maker's oscillator, it will be hard to say that all makers VCOs sound the same. Because they do choose different means to achieve their results. And that's before considering the effect of DC bias-offset, power supply/distribution modulation effect purposed or not, and even the level of conformance to the ideal exponential V-Oct curve.

While none of these are about the physical waveshape itself, they each and all can have very noticeable effect on the sounds produced natively, and even moreso on the sounds produced once the oscillator is placed in its normal role as only part of the sound generation chain.

Quote:
a moog sine sounds the same as a .com sine ..?


Perfect sines in subtractive analog synths are extremely rare. Although a few different means of analog sine creation have found widespread use, the details of these circuits almost ensure that sine waves will be more different than nearly any other wave across different makers. Once again the first difference is the core. A sine derived from a saw that's already been converted to a triangle, and then had its already less than perfect peaks nipped off electronically is going to sound different than lopping off the generally much cleaner formed peaks of a triangle core.

But it's still an oversimplification to try and summarize this into some sort of saw vs tri core answer. Too many variables come into play to say all saw cores or all tri cores sound the same.

Again, there is the bigger picture of how an oscillator is used in a practical sense, instead of the navel gazing on a seldom heard single osc, single waveform unchanging timbred drone!

Quote:
an Arp saw is the same as a Corsynth Saw ?


Here you are dealing with not only the polarity of the often misnamed 'saw', but also with the reset method, its resulting time impact and whatever means may be used to trim this effect to a reasonable result. It's not uncommon to have vertical spikes at top, bottom or both at or near the vertical parts of this waveshape. And these artifacts, along with any means used to minimise their effect, can often be heard.

Still, the largest difference will be heard if a saw is compared to a ramp, when both are used as part of a typical subtractive sound creation chain.

Adding to this is the fact that at least ARP always had their saws and square-pulses going 0-10V rather than the -5/+5V seen more often later and up to today. When this bias/offset is 'removed' by a DC blocking capacitor is also introduced the reality of filter reactance. Which means the resulting sound does change over changing frequency. One does not need a 'golden' ear to hear this, and when the osc is being used as a modulator instead of direct sound source it can become even more pronounced in the result.

Not to mention that in the typical sum VCOs into VCF as seen in an unmodded 2600, the mix seen at the filter input is not at all the same when pulse and saw are 0-10 and sine and tri are +/-5V.

Quote:
do you think they are the same tone/timbre saw for saw/square for square?
or that there is nothing between different raw sine/saw/squares ?

For all the reasons above, and many more left unsaid, most assuredly no, they are not the same. This is not the same as saying they cannot be made to sound the same, or to give similar results to a degree mostly un-noticed in a mix. But to do this will often require many nuanced choices and settings, and will not likely arise out of random knob twiddling.

Quote:
if you dont think they are the same which waves on which oscillators do you drift towards and why..?


Anyone reading my replies here can see an ARP bias. But the true answer is that the choice for which osc and which wavshape changes depending on what's being attempted as an outcome.

The most important part of the answer is to learn how the different oscillators are different, and for some that will involve getting a better understanding of what's happening under the hood, so to speak. For others, some casual time with an oscilloscope will show at least some of the differences described here in a way that will translate more readily into practical applications. For others, simply playing and listening will be enough to discern the differences, and more importantly to make appropriate use of them.

Nothing described here falls under golden ear territory, and anyone can get better at hearing differences with practice. Still the old tree in the forest koan holds true, if you can't hear it, does it make a difference? But if you're creating for a larger audience than yourself, best not confuse not hearing something with thinking nobody else does.
josaka
Xero wrote:


arps have this ability to get this certain nasally tone that just screams arp, not sure what it is exactly. just play one in person and you'll hear it eventually. i'm sure it's more than just the oscillators, it's the combination of the oscillators, the various amplifiers, mixers, etc, etc...and it doesn't even matter which filter....there's just something going on there.


pretty sure its down to how the filter drives the sound.. on the 2600 at least.. the filter and detuning/modulation sound key on the odyssey..
the Analog craftsman do a nice clone of the 2600 filter.

cretaceousear
Great reply KSS! Very informative. thumbs up
trentpmcd
Different.

When I received my Boog I was able to recreate every classic minimoog sound I could think of, but the Kieth Emerson sounds just weren't there. I had received my bank of Mos-Labs 901s at about the same time and plugged them into the Boog. I could get that Emerson sound, but not the Wakeman or other min sounds when using the 901s. I put in three dotcoms and had a completely different sound and could get neither the classic mini nor the classic Moog Modular sounds, but still got a great sound, just different.

OK, was it the mixer? I used mos-lab C3PO for 901s, STG mixer for dotcoms and just the straight Boog. I tried lowering the signal and pumping the dotcoms through the C3PO for a better comparison, but it was crappy. I tried boosting the 901s and putting them through the STG, but again couldn't get it to sound right. I spent a lot of time with the C3PO and STG tweaking levels to try to get the same amount of distortion, but I can't rule out that the huge difference in sound was caused by the mixers.

I've used the same trio of VCOs using other filters and can hear a difference, but none so large as using the Boog's filter - note - I don't have the Mos-labs 904 yet, so can't compare there.

Anyway, I think there is a difference.
Rex Coil 7
josaka wrote:
Are all different oscillator waves equal between different makes of oscillator..?
Some are just more equal than others ...


hihi
cornutt
Dave Peck wrote:


If you listen to a Dotcom sine, you can clearly hear a fizzy buzzy timbre added to the basic sine tone, audible in the background, and you can see this on a scope. It's a tiny sharp transient peak in the waveform, which is not present in other more 'accurate' sine oscs.


Yeah, getting a VCO to produce a good clean sine wave is a particularly difficult chore in analog. I know there's one trick that several designers use where they over-saturate an OTA... I don't understand how that works. Anyway, it's a hard problem. I have a Q123 standards module with the built-in A440 reference oscillator. If I tune a Q106 to 440 Hz and then A/B it against the A440 oscillator in the Q123, there's a noticeable difference. But there's also a difference if I A/B the Q123 against my H-P function generator.
Sugarfree
KSS wrote:

The digital means of waveshape generation almost always requires some form of filtering on the output to deal with the granularity of the discrete steps used to form the desired waveshape. As technology improves this granularity is getting smaller, and the filtering relationship changes in kind. Most filters are not perfect and therefore the sound of a given waveshape may not sound the same across its own range, let alone how it compares against another makers oscillator.


The above is a misconception. Digital-to-Analog converters don't produce audio with steps. The filters are only used to remove anything above the Nyquist frequency which otherwise would produce unwanted artifacts.
A "stepped waveform" is just a graphical representation of digital data, and doesn't reflect the actual analog output of a DA converter. Even a "stepped waveform" consisting of only 2 steps (samples), comes out as a perfectly continuous sine.
Rex Coil 7
Sugarfree wrote:
KSS wrote:

The digital means of waveshape generation almost always requires some form of filtering on the output to deal with the granularity of the discrete steps used to form the desired waveshape. As technology improves this granularity is getting smaller, and the filtering relationship changes in kind. Most filters are not perfect and therefore the sound of a given waveshape may not sound the same across its own range, let alone how it compares against another makers oscillator.


The above is a misconception. Digital-to-Analog converters don't produce audio with steps. The filters are only used to remove anything above the Nyquist frequency which otherwise would produce unwanted artifacts.
A "stepped waveform" is just a graphical representation of digital data, and doesn't reflect the actual analog output of a DA converter. Even a "stepped waveform" consisting of only 2 steps (samples), comes out as a perfectly continuous sine.
... "but when I turn the Freq adjust knob I can clearly hear stepping" ...

I think sometimes people conflate digital control with digital waveforms.

Now that I think about it, I don't know how that notion is relevant to what you posted here. It was there rattlin' around in my brainskull, but the thought has evaporated!

... ~shit~ ...


hmmm..... d'oh! seriously, i just don't get it
KSS
Sugarfree wrote:
KSS wrote:
...

The above is a misconception. ... Even a "stepped waveform" consisting of only 2 steps (samples), comes out as a perfectly continuous sine.

Please read what was written again. I carefully avoided any claim of disconinuous multi-step output.

Perfectly continuous and perfect sine are not the same thing.

What I wrote is correct.
Sugarfree
KSS wrote:
Sugarfree wrote:
KSS wrote:
...

The above is a misconception. ... Even a "stepped waveform" consisting of only 2 steps (samples), comes out as a perfectly continuous sine.

Please read what was written again. I carefully avoided any claim of disconinuous multi-step output.

Perfectly continuous and perfect sine are not the same thing.

What I wrote is correct.


you're right, I misinterpreted your post. sorry.
MUFF WIGGLER Forum Index -> 5U Format Modules  
Page 1 of 1
Powered by phpBB © phpBB Group