Emulations - Architecture vs. real emulation

[Meffy]

*sigh* There are no instantaneous signal transitions. They do not occur in the real world. If you believe otherwise, please consult any electronics engineer... besides me, that is. I've wasted enough time on you. Or, if you insist, continue to be wrong; no skin off my back.

[blümchen]

pustekuchen,
I for one am enjoying your commentary
Could I get your comments on a few emulations?

Korg Legacy's MS-20 and Polysix, Arturia's Moog's and CS-80V, and what company that you consider to do the most accurate emulations.

Wonder if Arturia emulated the transistor ladder circuit
in their MiniMoogV.

I tend to believe that most SE creations are not really complete emulations of the hardware but the sound that the hardware makes. Take Superwave-Performer for instance. Some of the presets do a good job of emulating the sound of some classic analog synths but once your start moving filters and envelopes the emulation is gone. So maybe the emulation definition needs to be more specific to include sound, controls and total emulations?

Worse Analog Keyboard?
http://users.informatik.haw-hamburg.de/ ... el_7A.html

Well.
Between the synthesizers you counted above AND THE MOST SE emulations I tested, lie miles of sound quality.

Nobody seriously could ever oversee that. SE emulations sound like cheapest crap compared to those above.

Althought, those above are not 100% authentic BUT earn the label "VA" and "emulation" very justified for several reasons (especially the sound quality they produce in general).

The point is, that there are some hobbyist SE "developers", who seriously believe, that if they use a wavetable or a soundfont and apply some standardized digital filters, that they then have made an "analog emulation". That's ridiculous.

.

[tomg]

*sigh* There are no instantaneous signal transitions. They do not occur in the real world. If you believe otherwise, please consult any electronics engineer... besides me, that is. I've wasted enough time on you. Or, if you insist, continue to be wrong; no skin off my back.

I insist.. Until I see otherwise.. Transistors switch in a fraction of a microsecond "near speed of light" that's absolutly true. Refernace "Art of Electronics Horowitz & Hill Page-64". This switching is what shapes the waveform. A fraction of a microsecond... well that's pretty fast isin't it.

Regards

[tomg]

The point is, that there are some hobbyist SE "developers", who seriously believe, that if they use a wavetable or a soundfont and apply some standardized digital filters, that they then have made an "analog emulation". That's ridiculous.

That's why I always call mine "reasonably faithful". Don't you just love that?

Regards

[shamann]

the Prophecy is nothing else than software running on a dedicated hardware-box...

Aah yes?
Then you own one?

I do, you have a question?

[The Chase]

Does anyone remember this thread here?
http://www.kvraudio.com/forum/viewtopic ... sc&start=0
or here?
http://www.kvraudio.com/forum/viewtopic.php?p=1254354
how about here?

[Meffy]

Transistors switch in a fraction of a microsecond [...] pretty fast

Ah -- at last you've reached reality. :-) The long way 'round, but you got there.

In the future, please remember that "pretty fast" is not the same as "instantaneous." IOW, if you don't mean zero time, do not say zero time. My professors would've given you a big "F" for such sloppiness.

Remember too that in a "fraction of a microsecond," a modern computer can complete rather a lot of operations.

Now I've some rather more pleasant activities to perform. 8-)

[McLilith]

Transistor switches do not have a slew rate. Thats just for opamps.

Umm, actually no. Any sort of electronic amplifier circuit has a slew rate, not just op-amps. The same holds true for digital circuits that only switch between two states. Logic chips have slew rates also.

Meffy: I haven't read this entire thread, but I believe Tomg does know more than you gave him credit for. I think he was just a bit "euphemistic" with his teminology earlier. Lots of people round things off like that. An extremely small quantity becomes zero, even though they really know that it isn't. I personally don't like such a rounding off of the truth, but lots of people do it. Tomg has built lots of analog synth circuitry, and he should know a fair amount about analog synths by this point. I would like to think that he was just "rounding off" things.


take care,
McLilith

[tomg]

Transistors switch in a fraction of a microsecond [...] pretty fast

Ah -- at last you've reached reality. The long way 'round, but you got there.

In the future, please remember that "pretty fast" is not the same as "instantaneous." IOW, if you don't mean zero time, do not say zero time. My professors would've given you a big "F" for such sloppiness.

Now I've some rather more pleasant activities to perform.

I said "near light speed". At 96k/sec a sample comes in every 10.4 microseconds. That's a little slow.

Regards

[tomg]

Transistor switches do not have a slew rate. Thats just for opamps.

Umm, actually no. Any sort of electronic amplifier circuit has a slew rate, not just op-amps. The same holds true for digital circuits that only switch between two states. Logic chips have slew rates also.

Meffy: I haven't read this entire thread, but I believe Tomg does know more than you gave him credit for. I think he was just a bit "euphemistic" with his teminology earlier. Lots of people round things off like that. An extremely small quantity becomes zero, even though they really know that it isn't. I personally don't like such a rounding off of the truth, but lots of people do it. Tomg has built lots of analog synth circuitry, and he should know a fair amount about analog synths by this point. I would like to think that he was just "rounding off" things.


take care,
McLilith

Hi there I have never seen a referance to a transistor slew rate. I even googled "transistor slew rate" and came up empty.

I suppose if you want it to mean "any change over time of anything" but as far as I know the electronic meaning has alwasy been "the maximum rate which an amplifier's output can change, expressed in V/µs and caused by compensation caps."

Regards

[McLilith]

Your f**king smiley bit me.

That's what happens when my smiley detects a moron in denial. Try not to be so stupid next time.

Regards

Well, you simply cannot deny the logic of what I'm saying, and you are not so cleverly pretending that you can. Please - explain how any possible differences in sound quality due to factors that go beyond the human being's power to differentiate can make a difference. Please. And leave out the infinity crap.

Yes I can. The waveforms are not the same = the sound is not the same. It's not that hard to grasp.

Regards

LOL! I can see that logical reasoning isn't one of your strong points. Please point out where I said that digital emulations sound exactly the same as analog. If you read back, I was simply trying to show (which was easy) how silly your actual arguments are. Whether you are a turd or not (not an emulation) is a subjective assessment.

My argument is that software is too slow to emulate a hardware VCO without compromising the sound. That is absolutely true.

If you could have refrained from calling me names like a 5 year old, I might have actually got what you are trying to say but at this point I don't care.

Regards

Hi tomg,

As a fellow electronics person, I "feel your pain" right about now.

I do feel you have a valid point, but you are perhaps not expressing it in a way that the rest of the group understands. You and certain other members of the forum seem to be having orthogonal arguments. In other words, you are actually arguing over different issues which have no bearing on each other, apparently without even realizing it. Each side seemingly misunderstands the statements of the other, and argues against their misconstrued perception of the other side, instead of directly confronting what the other side has said.

I'm probably not making things extremely clear myself at the moment, but I do have an instinctive feeling that you are right about software needing far more speed, but perhaps not exactly for the literal reasons you gave.


take care,
McLilith

[McLilith]

Hi there I have never seen a referance to a transistor slew rate. I even googled "transistor slew rate" and came up empty.

I suppose if you want it to mean "any change over time of anything" but as far as I know the electronic meaning has alwasy been "the maximum rate which an amplifier's output can change, expressed in V/µs and caused by compensation caps."

Regards

I don't care if no one ever mentions "transistor slew rate", my point is that slew rate is not a measurement that can only be applied to op-amps. For starters, here is a web page that applies it to digital circuitry:

Digital Slew Rate

You may hear "slew rate" (more accurately, maximum slew rate) mentioned more often with op-amps, but the term can be applied to other types of amplifiers as well. It can apply to analog or digital, tube or solid state, discrete circuitry or integrated circuitry, small-signal circuitry or power circuitry. For example, it's also a term that is commonly applied to audio power amps. Many of these have topologies quite different from an op-amp.

As for the compensation caps, you don't need to have compensation caps to have a circuit with a maximum slew rate. The presence of compensation caps is one of those things that are common to op-amps, but certainly not a requirement for the existence of a maximum slew rate.


take care,
McLilith

[tomg]

I'm probably not making things extremely clear myself at the moment, but I do have an instinctive feeling that you are right about software needing far more speed, but perhaps not exactly for the literal reasons you gave.


take care,
McLilith

It's weird huh? I felt it go that way .. Like I was attacking all softsynths or something... But all I really was saying is use what you have, be happy you have it and don't expect miracles. Sometimes you get very close to one

Kill the torches and put away the pikes, who knows maybe next time. Hey don't we need a new emoticon?

Regards

[crimsontider]

The point is, that there are some hobbyist SE "developers", who seriously believe, that if they use a wavetable or a soundfont and apply some standardized digital filters, that they then have made an "analog emulation". That's ridiculous.

.

That sounds like a farce to me. I am so far removed from the developer side that I didn't realize "synths" were being created and labeled in this manner.......So they actually just sample the sound or table, throw in some cookie cutter filters and envelopes and walla? Check out my new vintage analog emulation, similar to high performance tires with 1PSI of air in them.

[tomg]

I don't care if no one ever mentions "transistor slew rate", my point is that slew rate is not a measurement that can only be applied to op-amps.

I'm just saying it's not in any of my books or prior knowledge. Terms change over time I guess you could say terms have a slew rate too?

Regards