Business ethics

[Nokenoku]

Well, there are clear lies, but then there are also misleading statements in marketing. Those are not much better (if at all) than lies.

There are often sentences in marketing, which make people just think clearly wrong things about certain stuff. If you look at it as a lawyer, there might be no real lies in there, and if you actually already know the facts you understand, what such a sentence is really saying (often really nothing).

But as a normal user, who hasn't much technical background with those things, you read it and almost definetly start to think wrong about certain aspects of those plugins and their competitors.

[Warp69]

I think it is inappropriate for a direct competitor to comment in a other business thread on this matter...or try to subjectively 'police' what does and doesnt sound accurate - again - on the competitor thread...

In reality - it is better IMO to let the users come to judgement about whether something is or isnt accurate...

I respectfully disagree.

It happens in every business - why should this business be any different compared to other tech business?

You also assume that only legitimate 'users' participates in threads - what about users with multiple accounts? Users affiliated with the company or competition? Companies with multiple accounts pretending to be users? etc etc etc.

[Gamma-UT]

I think it is inappropriate for a direct competitor to comment in a other business thread on this matter...or try to subjectively 'police' what does and doesnt sound accurate - again - on the competitor thread...

In reality - it is better IMO to let the users come to judgement about whether something is or isnt accurate...

I respectfully disagree.

It happens in every business - why should this business be any different compared to other tech business?

You also assume that only legitimate 'users' participates in threads - what about users with multiple accounts? Users affiliated with the company or competition? Companies with multiple accounts pretending to be users? etc etc etc.

If you forbid competitors posting then you are more likely to get sockpuppetry going on, which is even more inappropriate. However, a lot people who indulge in using fake accounts aren't quite as good at it as they like to think they are: my favourite is when you see two low post-count 'users' vehemently agreeing with each other seconds after the other has posted.

I think you've got to leave some room for discretion, especially as it's normally the intruding vendor that comes off worst if they don't have a legitimate point. People can read the posts and make up their own minds.

[AdmiralQuality]

False claims are a matter of fraud, not ethics. If a company states something that's a lie, then that's "taking money by false pretenses." And is liable to deception offences, and the Theft Act, and other such blah-blah stuff.

Marketeers tend to exaggerate, but do not lie. My favourite, is anything that's "psycho-acoustic," it always makes me smile when I read that one.

Dave

My favorite marketing exaggerations always have to do with the modelling of analog circuit components. The one I remember best is by our old friend here. Step into the Wayback Machine, Sherman...

http://web.archive.org/web/200602062345 ... m/aclm.php

To get the best possible model of the analog equipment, it is not enough to analyse it on a simple output vs. input basis, and we do not have to. Analog equipment is built from many more or less independent circuitries that should be recognized and analysed, and a partial-algorithm made for each circuit.

At Kjaerhus Audio, we use our own technique developed to transform each circuit into a part-algorithm. We call it "Advanced Component Level Modelling" (ACLM). The basic idea is to give each component its own place in the part-algorithm.

When we develop a plug-in, all analog circuits are designed as if we where going to make a piece of hardware. After that, all circuits are transformed into part-algorithms using the ACLM technique and tested to make sure it works like its analog counterpart (which it does if everything was done right!). Lastly, all part-algorithms are "connected" just like the circuits they are modelling. This becomes the final plug-in.

Isn't that the most shameless marketing exaggeration you've ever read? Yes, we precisely model analog components like transistors (with a single multiply operation), resistors (with a single subtraction operation), and capacitors (no, we don't actually, but we're hoping you don't know what they are).

[Urs]

I think you've got to leave some room for discretion, especially as it's normally the intruding vendor that comes off worst if they don't have a legitimate point.

They even get bashed if they do have a point.

Something's wrong in that. Unless I use a competitor's thread for marketing my own stuff it must be possible to raise points, especially when said competitor uses borderline practices. If the two companies in question do not have any overlap in product range then there's not even any conflict of interest.

For instance, if a synth developer takes side in a shootout between an UAD EQ and a competing Sonnox EQ then that should be perfectly fine. The same even accounts for a thread that is about Minimoog emulations (classics like Gmedia vs. Arturia vs. Creamware) if the "intruding synth developer" does not have a stance in that field.

Urs

[quikquak]

I think it is inappropriate for a direct competitor to comment in a other business thread on this matter...or try to subjectively 'police' what does and doesnt sound accurate - again - on the competitor thread...

In reality - it is better IMO to let the users come to judgement about whether something is or isnt accurate...

I respectfully disagree.

It happens in every business - why should this business be any different compared to other tech business?

You also assume that only legitimate 'users' participates in threads - what about users with multiple accounts? Users affiliated with the company or competition? Companies with multiple accounts pretending to be users? etc etc etc.

If you forbid competitors posting then you are more likely to get sockpuppetry going on, which is even more inappropriate. However, a lot people who indulge in using fake accounts aren't quite as good at it as they like to think they are: my favourite is when you see two low post-count 'users' vehemently agreeing with each other seconds after the other has posted.

I think you've got to leave some room for discretion, especially as it's normally the intruding vendor that comes off worst if they don't have a legitimate point. People can read the posts and make up their own minds.

Lol, yes.
imdb.com is full of 'shills,' and you can always tell if it's somebody that worked on a film posting messages.

Shill:-
http://en.wikipedia.org/wiki/Shill


Dave

[AdmiralQuality]

I think you've got to leave some room for discretion, especially as it's normally the intruding vendor that comes off worst if they don't have a legitimate point.

They even get bashed if they do have a point.

Something's wrong in that. Unless I use a competitor's thread for marketing my own stuff it must be possible to raise points, especially when said competitor uses borderline practices. If the two companies in question do not have any overlap in product range then there's not even any conflict of interest.

For instance, if a synth developer takes side in a shootout between an UAD EQ and a competing Sonnox EQ then that should be perfectly fine. The same even accounts for a thread that is about Minimoog emulations (classics like Gmedia vs. Arturia vs. Creamware) if the "intruding synth developer" does not have a stance in that field.

Urs

I sold a Poly-Ana last night, probably because I posted in the Tone2 ElectraX thread that I liked their GUI with 3D perspective, and compared the bashing they were receiving to my own experiences. Turns out, there was somebody here who didn't know about Poly yet.

I'd apologize, but it means I get to eat tomorrow. (Seriously.)

[Urs]

Isn't that the most shameless marketing exaggeration you've ever read? Yes, we precisely model analog components like transistors (with a single multiply operation), resistors (with a single subtraction operation), and capacitors (no, we don't actually, but we're hoping you don't know what they are).

Hehe... actually... analogue component modeling can indeed be done. You commonly end up with huge matrice operations though that can't be calculated in realtime. The algorithm then looks nothing like the circuit schematic. In essence, you build a spice model or something and then use an iterative problem solver.

You can also do a very accurate model of a Moog filter just with lookup tables. For each input sample, former output sample (as unique state variable), frequency parameter, resonance parameter exists exactly 1 output sample. So you could store that in a 4-dimesnional array with, say, 65536 indices each and have an analogue model that is 16 bit accurate.

[AdmiralQuality]

Isn't that the most shameless marketing exaggeration you've ever read? Yes, we precisely model analog components like transistors (with a single multiply operation), resistors (with a single subtraction operation), and capacitors (no, we don't actually, but we're hoping you don't know what they are).

Hehe... actually... analogue component modeling can indeed be done. You commonly end up with huge matrice operations though that can't be calculated in realtime. The algorithm then looks nothing like the circuit schematic. In essence, you build a spice model or something and then use an iterative problem solver.

You can also do a very accurate model of a Moog filter just with lookup tables. For each input sample, former output sample (as unique state variable), frequency parameter, resonance parameter exists exactly 1 output sample. So you could store that in a 4-dimesnional array with, say, 65536 indices each and have an analogue model that is 16 bit accurate.

Yep, you could. But we all know, nobody does.

[Urs]

I'd apologize, but it means I get to eat tomorrow. (Seriously.)

[dhalfen]

I sold a Poly-Ana last night, probably because I posted in the Tone2 ElectraX thread that I liked their GUI with 3D perspective, and compared the bashing they were receiving to my own experiences. Turns out, there was somebody here who didn't know about Poly yet.

Yeah, well, let's be honest -- Rob has his hands on PLENTY o' synths.

But I would certainly have recommended Poly-Ana to him if I had ANY idea that he was unaware of it.

I don't care if you model unicorns for the circuits -- the thing sounds bloody awesome.

[dhalfen]

Marketeers tend to exaggerate, but do not lie.

Well, not intentionally, anyway, depending who's doing any "proofreading" of the ad copy.

Worst offenders I've ever seen: the Crysonic manuals. They're HILARIOUS!!! (And just about the brightest red flags in the business.)

As a "buyer/user," I pay absolutely NO attention to advertising. I'm interested in 4 things:

(1) How does it sound?

(2) How functional/readable is the interface? (And can I do numeric value-entry like Synth Squad or Largo? -- VERY important to me!!!)

(3) How does it perform on my system?

(4) How soon am I going to run into a limitation that really irks me? And is the developer interested in fixing/updating the product?

You pass these four criteria, and you're likely to get my money (sooner or later -- I'm not wealthy!!! ).

(Urs, you have the BEST interface option -- different set sizes -- currently available, and I thank you for this!!! )

[aciddose]

You can also do a very accurate model of a Moog filter just with lookup tables. For each input sample, former output sample (as unique state variable), frequency parameter, resonance parameter exists exactly 1 output sample. So you could store that in a 4-dimesnional array with, say, 65536 indices each and have an analogue model that is 16 bit accurate.

you can't model a iir filter with a single fir filter urs, come on, you know that.

in reality you'd need a lot more than one previous sample. you'd need more like perhaps several times the wavelength for that frequency and the wavelength of course would be different for every frequency. in addition to that - that's assuming the filter is only linear. which again you know it is not.

so it's quite impossible with a reasonable size table (non-infinite?)

[Urs]

You can also do a very accurate model of a Moog filter just with lookup tables. For each input sample, former output sample (as unique state variable), frequency parameter, resonance parameter exists exactly 1 output sample. So you could store that in a 4-dimesnional array with, say, 65536 indices each and have an analogue model that is 16 bit accurate.

you can't model a iir filter with a single fir filter urs, come on, you know that.

in reality you'd need a lot more than one previous sample. you'd need more like perhaps several times the wavelength for that frequency and the wavelength of course would be different for every frequency. in addition to that - that's assuming the filter is only linear. which again you know it is not.

so it's quite impossible with a reasonable size table (non-infinite?)

I'm quite certain that it's doable.

If you had a spice model the input sample, the filter cv, the resonance cv and the state *always* result in the same output sample, no matter what nonlinearities you have. The question is only if a single state variable is enough to represent the state as a whole. Otherwise we need a 7-dimensional lookup table.

We however do not need any previous input or output samples if the model is based on a delayless loop.

(Note though that the memory consumption for this goes into giga/terabyte territory - I never said it was feasable, did I?)

[AdmiralQuality]

But I would certainly have recommended Poly-Ana to him if I had ANY idea that he was unaware of it.

And here I was thinking I had the place saturated.

Thanks!