ARP 2600 Clone?

[ghettosynth]

While I believe in bias, I'm in agreement with IncarnateX.

Based on what I'm reading, I'm not sure that you do (agree with IncarnateX). Which is really of no matter, because he IS overstating the position, whether people agree or not. I could explain this in excruciating detail, but it would be a waste of time and way off topic on KVR. As I've said, this isn't a (science) paper, it's a conversation.

Essentially, when we consider two digital filters (to keep this simple), and ask if one is better than the other with respect to how close we are to emulating an analog filter. Then we have a truth state and a perception of the truth state.

The truth state is how accurately these digital filters actually model the analog counterpart. The perception of the truth state is what any individual hears. What IncarnateX is saying is that any one person's perception cannot be asserted "scientifically" to represent the truth state. This is independent of the bias arguments.

This is too strong. It neglects the actual truth state and the common understanding of sound. Additionally, not all perceptions are equal, and bias that might influence perception does not always go in the desired direction, as you point out below, and as I mentioned above. We absolutely can use perception to answer these kinds of questions and the more experienced a listener is, the more value we should place on their perception. Yes, there will always be variance about their response, and bias is a powerful influence, but, it is far too strong of a statement to throw perception out as being "un-scientific" and not representative of the truth state. This is the kind of pedantic argument that pseudo-science often makes.

At the end of the day, perception corrected for strong visual and suggestion bias, is ALL that matters. It is sound, after all, and it's only function in this context is to be perceived.

He is NOT saying that there is or isn't a difference, and neither am I.

While there are some sounds that can be nailed beyond anyone's ability to spot the difference, there are enough that can't.

Yes, you're preaching to the choir. You know that right?

I'm a big fan of digital synths, in fact I made a move to digital synths in the 90s (didn't we all?!

As did I, well, I didn't completely move to digital synths, but the K2K, Nord G1, and JD-800 became centerpieces of my studio and live rig.

When I was convinced that there was some "analog magic" I might be missing, I went to see what people were talking about. I was very skeptical. I was biased against old analogs. I hated all that crappy synth pop from the 70s and 80s. But I heard the difference. It wasn't totally obvious at first but once I knew what to look for, I could not "unhear" it.

Right, this is what I mean by bias working against you, or going in the wrong direction. You were convinced that analog wasn't better, but you could hear the difference.

Smash cut, we now have some amazing software VA that really does get a lot closer. I'd love to ditch my analogs, I really would. I can't though. They provide character I can't get as easily in software. So, what do you have to say to someone who was biased against analog?

I say thank you for supporting my point of view on this. Perception can be thought of as a mixture of influences. You hear something and some part of what you hear is the truth state of the sound. Some part of what you hear is bias and that will color your perception in some direction. That bias may be due to visual stimulus, or it may be due to suggestion, both have been shown to have a powerful effect on our perception. The bias may not add to the perception in the desired direction, but it does color perception nonetheless. As long as we take steps to control the bias in what we're hearing, we can make qualitative statements of substance about our perceptions. This is especially true when this is done in aggregate, i.e., we yield the power of many perceptions instead of just our own.

What I am saying is that as long as one knows what to listen for, that the truth state of the sound dominates in such a model in cases where there are gross differences. Hence, perception is a reasonable estimator of the truth state.

Or, in more layspeak, yes bias matters, but the differences are so dramatic in some cases, this case in particular, that as long as you know what to listen for, you will be able to hear the difference.

Further, I'm saying that we have other data that informs us that the 2600v cannot be accurate. It is an old design that we know does not take advantage of newer methods that yield greater accuracy. We don't even have to listen to the filter to know that this is true.

[pdxindy]

Or, in more layspeak, yes bias matters, but the differences are so dramatic in some cases, this case in particular, that as long as you know what to listen for, you will be able to hear the difference.

Further, I'm saying that we have other data that informs us that the 2600v cannot be accurate. It is an old design that we know does not take advantage of newer methods that yield greater accuracy. We don't even have to listen to the filter to know that this is true.

It is like 3d rendering... when renderers started adding functionality like global illumination, self shadowing and translucent materials like skin, 3d renders became much more realistic and also took more CPU

[zerocrossing]

Or, in more layspeak, yes bias matters, but the differences are so dramatic in some cases, this case in particular, that as long as you know what to listen for, you will be able to hear the difference.

Further, I'm saying that we have other data that informs us that the 2600v cannot be accurate. It is an old design that we know does not take advantage of newer methods that yield greater accuracy. We don't even have to listen to the filter to know that this is true.

It is like 3d rendering... when renderers started adding functionality like global illumination, self shadowing and translucent materials like skin, 3d renders became much more realistic and also took more CPU

Good point. Look at Toy Story 1. It's not even raytraced. A lot of it still looks fine, as plastic toys work fine for such things. But some of it is horrible by today's standards... Hell, it looked bad by it's day's standards as well, but when you have to render 24 high resolution frames per second of film, there was no way they could get a large enough render farm to render the movie in a timely fashion. I watched some of it with my 2 year old a while ago and my first thought was, "Why don't they render this film?" I bet with a minimal amount of work they could make it look much better... Especially the baby's hair.

Then I realized why. Because, Toy Story is a great film. Like great music played on a pawn shop special, a great story that's well acted transcends it's tech.

[IncarnateX]

it is far too strong of a statement to throw perception out as being "un-scientific" and not representative of the truth state.

Rubbish. I would never say that and never have. A perception is always an individual state of truth, question is how you will make your state more valid than someone who has an opposite state.
The body of evidence I am referring to in my long post about psychology of perception shows that

1. As far as sound concern what we perceive is influenced by other senses as well, especially vision
2. Any perceived difference can be due to training or it can be an illusion made by the brain
3. It can be hard even for the individual himself to know whether he is trained or the brain makes this differences to him
4. If someone has the opposite perception than you and you want to convince him that he is wrong, you cannot do this by just claiming it by reference to your own perception (state of truth) and with appeal to your own authority (I am a trained expert). That is not as complicated a statement as you want to make it, but common sense within science.

Now you haven’t provided any evidence in this respect that you are right that your sound examples cannot be replicated, it is all presented as assertions with appeal to your own authority, e.g. claims about filters, so given that someone would disagree with you, none of this will help you.

Proving that the Earth is round and not flat is one way to show that people’s perceptions are false with respect to external reality. Now you need similar means to prove their perceptions are false and yours are right with repect to correspondance with reality.

Since you and many others do not mind eliminating visual bias in the analog-digital question, you could do this by passing a simple blind tests. That would show that you are not hallucinating or whatever and thus your perceptions are now backed up by actual evidence and can be used to prove that there must be a perceivable difference given that you are trained. But your suggested matching of examples as shown in my previous post will not do.

Most annoying right now is that you argue with me over my head by virtue of others while ignoring my posts and that, Sir, is the most unscientific about your approach right now. That is the act of a coward.

[trimph1]

I keep saying...that. GUI means a lot for a fair number of people. Hence the EYE dominates even sound.

Turning the monitor off to really listen to the sounds might work, although I kind of doubt it would with some. The bias would be that strong.

[ghettosynth]

I keep saying...that. GUI means a lot for a fair number of people. Hence the EYE dominates even sound.

Turning the monitor off to really listen to the sounds might work, although I kind of doubt it would with some. The bias would be that strong.

It does (mean a lot, it doesn't necessarily dominate), but watch the McGurk effect video that I posted earlier. The visual bias goes away immediately when you stop looking at the visual stimulus. If you don't know which source creates the sound you're hearing, and you can't see it, then you have for all practical purposes mitigated the effect of this bias.

I get the feeling that some of you have never tried this for yourself? I do this kind of thing all the time because I'm very aware of how perception bias impacts what we hear.

[wagtunes]

I keep saying...that. GUI means a lot for a fair number of people. Hence the EYE dominates even sound.

Turning the monitor off to really listen to the sounds might work, although I kind of doubt it would with some. The bias would be that strong.

It does, but watch the McGurk effect video that I posted earlier. The visual bias goes away immediately when you stop looking at the visual stimulus. If you don't know which source creates the sound you're hearing, and you can't see it, then you have for all practical purposes mitigated the effect of this bias.

I get the feeling that some of you have never tried this for yourself? I do this kind of thing all the time because I'm very aware of how perception bias impacts what we hear.

This really is true. When I first saw a video of a real orchestra playing to GPO Strings, I swear they sounded real. Then I bought GPO and realized how poor they actually sound next to real strings.

Visual bias, at least for me, is super strong, which is why I absolutely have to hear things without any visuals at all in order to accurately tell the differences between the two and which one I like better.

And even then, if the sounds are close enough, I probably can't tell anyway.

It is rare that I hear something that makes me go "wow, that was amazing." And it is indeed quite possible that my "memories" of the ARP 2600 aren't nearly as accurate as the actual sound of the synth.

That is why I'd love to hear a non visual blind demo between a real ARP 2600 and one of the emulations to see if I can tell the difference or at least pick the one I like better.

Do any such exist? If so, where and I'll give a listen.

[trimph1]

Oh, and don't forget. Because many Arp2600's were hand wired so there could be a wee bit of variance between them. I also redone mine, which was hand wired, a few times since...

[ghettosynth]

Oh, and don't forget. Because many Arp2600's were hand wired so there could be a wee bit of variance between them. I also redone mine, which was hand wired, a few times since...

A wee bit? Absolutely, I'd say that there's quite a bit of variance, I'm not so sure that hand wiring has a huge effect, but certainly just the analog nature will contribute to variance. A larger source of variance will be in the setup of each test. You will not be able to restore the knobs to precisely the same place in many cases and so these small variances may have a cascading effect. But, if we can pick virtually any example from the population, assuming that the device is properly calibrated, and still consistently hear/measure the difference, then for all intents and purposes, that variance isn't really relevant to the conversation.

[EnGee]

I would love a real 2600.

One of m favourite soft synths is u-he ACE. It is loosely based on the architecture of the 2600 and can do a few of it's tricks. Sounds pretty decent to my ears, and at the price, it's a steal

Cheers

Scorb

You made me curious to try ACE...

Yes, you are absolutely right! I had some play with ACE this morning (bought it just this morning from the Market Place) and it sounds fantastic. Not sure how near to the real ARP 2600, but it can be aggressive and it does 'in your face' sounds easily.

It has nice filters and great sound clarity. The cpu is not that bad! I expected it to use more than that. Anyway, one thing it has that is missing in all other emulations (maybe even of course the original 2600 too!!) is the ability to do great pads (sounds strange but it is so similar to the pads in my Blofeld!! Crazy, but I need to compare more!).

I will try to apply the patch book that came with Timewarp 2600 (I think the same as the original one for the real 2600). They are implemented already in both Timewarp 2600 and Arturia 2600V but they sound very different! Sometimes I prefer the 2600V and sometimes the Timewarp one. I'm curious to hear how they sound in ACE

[zerocrossing]

While I believe in bias, I'm in agreement with IncarnateX.

Based on what I'm reading, I'm not sure that you do (agree with IncarnateX). Which is really of no matter, because he IS overstating the position, whether people agree or not. I could explain this in excruciating detail, but it would be a waste of time and way off topic on KVR. As I've said, this isn't a (science) paper, it's a conversation.

Essentially, when we consider two digital filters (to keep this simple), and ask if one is better than the other with respect to how close we are to emulating an analog filter. Then we have a truth state and a perception of the truth state.

The truth state is how accurately these digital filters actually model the analog counterpart. The perception of the truth state is what any individual hears. What IncarnateX is saying is that any one person's perception cannot be asserted "scientifically" to represent the truth state. This is independent of the bias arguments.

This is too strong. It neglects the actual truth state and the common understanding of sound. Additionally, not all perceptions are equal, and bias that might influence perception does not always go in the desired direction, as you point out below, and as I mentioned above. We absolutely can use perception to answer these kinds of questions and the more experienced a listener is, the more value we should place on their perception. Yes, there will always be variance about their response, and bias is a powerful influence, but, it is far too strong of a statement to throw perception out as being "un-scientific" and not representative of the truth state. This is the kind of pedantic argument that pseudo-science often makes.

At the end of the day, perception corrected for strong visual and suggestion bias, is ALL that matters. It is sound, after all, and it's only function in this context is to be perceived.

He is NOT saying that there is or isn't a difference, and neither am I.

While there are some sounds that can be nailed beyond anyone's ability to spot the difference, there are enough that can't.

Yes, you're preaching to the choir. You know that right?

I'm a big fan of digital synths, in fact I made a move to digital synths in the 90s (didn't we all?!

As did I, well, I didn't completely move to digital synths, but the K2K, Nord G1, and JD-800 became centerpieces of my studio and live rig.

When I was convinced that there was some "analog magic" I might be missing, I went to see what people were talking about. I was very skeptical. I was biased against old analogs. I hated all that crappy synth pop from the 70s and 80s. But I heard the difference. It wasn't totally obvious at first but once I knew what to look for, I could not "unhear" it.

Right, this is what I mean by bias working against you, or going in the wrong direction. You were convinced that analog wasn't better, but you could hear the difference.

Smash cut, we now have some amazing software VA that really does get a lot closer. I'd love to ditch my analogs, I really would. I can't though. They provide character I can't get as easily in software. So, what do you have to say to someone who was biased against analog?

I say thank you for supporting my point of view on this. Perception can be thought of as a mixture of influences. You hear something and some part of what you hear is the truth state of the sound. Some part of what you hear is bias and that will color your perception in some direction. That bias may be due to visual stimulus, or it may be due to suggestion, both have been shown to have a powerful effect on our perception. The bias may not add to the perception in the desired direction, but it does color perception nonetheless. As long as we take steps to control the bias in what we're hearing, we can make qualitative statements of substance about our perceptions. This is especially true when this is done in aggregate, i.e., we yield the power of many perceptions instead of just our own.

What I am saying is that as long as one knows what to listen for, that the truth state of the sound dominates in such a model in cases where there are gross differences. Hence, perception is a reasonable estimator of the truth state.

Or, in more layspeak, yes bias matters, but the differences are so dramatic in some cases, this case in particular, that as long as you know what to listen for, you will be able to hear the difference.

Further, I'm saying that we have other data that informs us that the 2600v cannot be accurate. It is an old design that we know does not take advantage of newer methods that yield greater accuracy. We don't even have to listen to the filter to know that this is true.

OK, I got lost in that conversation so perhaps I'm arguing with no one? In that case, I'll agree with myself!

One thing I will say that along with bias we also have varying degrees of hearing loss unless we're children raised in a quite woodland cottage. This alone sets up a totally different issue. I remember when I was working at a hi-fi stereo store and I had one of the local heavy metal guitarists come in. He was enamored with a set of speakers that to me were the shrillness sounding speakers I'd ever heard. So bright. Coincidently a guy who ran a local studio known for heavy metal production (Anthrax recorded there) came in and they both commenced to profess their undying love for these ultra bright speakers (which I thought sounded horrible). Wear ear protection kids!

I'm surprised that no one has done some sort of test that looked at spectral data over time on various synths.

Right now I'm doing a little test that's been pretty illuminating. I wanted to see how close I could recreate a sound coming from a Modulus .002 in software. First off, finding a synth with that architecture was a challenge. On first glance I thought it was a 2 oscillator with a sub oscillator architecture. Not really. It's more like a four oscillator setup. Then the example is using a stack mode so each oscillator, including the sub oscillators, are multiplied by 3 so I'm looking at 12 oscillators per note. So that kind of digital oscillators into an analog filter... I went to ZebraHZ and failed. I even had the parameters of the .002 in front of me and I couldn't quite get there. The .002 sounded rich and clear where Zebra 2 always seemed to sound a bit I distinct and mushy. Maybe it was my lack of sound design prowess, but this simple sound eludes me. Of course, Zebra isn't pretending to emulate anything, but on a deeper level there was a quality in the hardware that I think is what we all talk about when we praise hardware. I super want to be wrong because I don't have the money for a .002, and I want one.

[BlackWinny]

I'm quite happy with the free Arppe2600va. Of course it is a SynthEdit product and quite old, and probably that it won't be able to compete with the majors, but for free it is not bad at all for those who can't afford an expensive one...

[david.beholder]

Yes, you are absolutely right! I had some play with ACE this morning (bought it just this morning from the Market Place) and it sounds fantastic. Not sure how near to the real ARP 2600, but it can be aggressive and it does 'in your face' sounds easily.

Ace is very bread and butter synth from fluffy pads to chainsaw leads but... Filters are different from ARP, too squelchy. Oscillators sound quite close.

Urs mentioned once they might add other filters, closer to 2600 some day.

[david.beholder]

Right now I'm doing a little test that's been pretty illuminating. I wanted to see how close I could recreate a sound coming from a Modulus .002 in software.

Could you please post the sound?

[trimph1]

I'm quite happy with the free Arppe2600va. Of course it is a SynthEdit product and quite old, and probably that it won't be able to compete with the majors, but for free it is not bad at all for those who can't afford an expensive one...

And it can sound rather unpleasant as well. And it does come very close. It is too bad that Conrady passed on.