Plug designers who aren't audio engineers.

[Urs]

No problem. ... Problem?

Why is the second example problematic in any way?

Urs

[Spratman]

I think perhaps what Champ is suggesting is that it's possible to get a real "vintage" processor that is sold as a vintage processor. But the thing is, I've yet to find anyone who knows exactly what "vintage" actually sounds like. People can compare software with whatever they want and if you're canny you can play on that. I think a lot of people do/will. But the real clever developers will only allude to selling "vintage" products and let the audience sell it for them...

Spratman

[Urs]

Ah well... okay... that's me then, I always smirk when I come across features like "accurately modeled" or "ultra warm vintage so and so". This has always to be taken with a grain of salt or two, regardless of the developer's qualification.

But I naturally would disagree if there were claims that one has got to be any sort of audio/electro engineer to create audio software that sounds good.

Cheers,

Urs

[whyterabbyt]

I'd like to know the percentage of designers of amps, compressors, eq's and the like that were actually audio engineers. I mean isnt an assumption that a lot of them weren't just EE's who were working on audio gear to fit a given spec anyways?

[saddlemirestudios]

A plugin can be developed a couple different ways. The first being a programmer expiriements blindly with his computer swapping in different code and numbers until it sounds good. The other being an engineer starts the design from schematics and theory which is then developed into software. I find that both of these approaches can be successful, however engineers really are the ones that come up with the industry changing plugs first. I dont think any computer nerd would have just thought to use the fouier transform of a sin wave and other advanced calculus techyniques to implement this whole convolution/dynamic convolution movement. For an engineer however, from my own expirience and lots of homework, would just see that as an obvious solution. The computer nerds however do have critical part in the industry since more and more of the audio recording process is based around the computer. Getting plugs like URS or Sonalksis down to such low CPU drain and stability is an art form itself.

[championrabbit]

I'd like to know the percentage of designers of amps, compressors, eq's and the like that were actually audio engineers. I mean isnt an assumption that a lot of them weren't just EE's who were working on audio gear to fit a given spec anyways?

That's a truism to an extent; the vast majority of 'standard' studio gear was designed (with the possible exception of mics) at a time when the field was new enough so that there was nothing to be an 'expert' in if you see what I mean? Joe Meek for example was plenty trial and error.

These days however we must accept that there are certain fundamental standards in gear design that most manufacturers attempt to achieve; in essence they are attempting to recreate the sounds of the 'standard' studio with their own little twists or better pricing.

Recording gear can fairly accurately be broken down into two groups:

1. Recreations of studio 'standards' such as G-Series channels, LA2As, Pultecs, Marshall JCMs, Fender Tweeds, Neve strips, Joe Meek comps, DBX comps etc.

2. Vanilla-flavoured stuff.

New gear that is esoteric in nature but has no root in 'standard' seems pretty non-existant in terms of effects. You don't encounter new desks that boast a colourful sound that is other than that of a Neve or EMI or SSL etc; marketing would be impossible.

I would suggest that those involved in the recreation/tweaking of 'standard' gear would by neccesity need a profound understanding of what makes the 'standard' gear subjectively good. In other words they would need an audio background as well as EEing.

Non?

[helium]

You can't directly compare EE with DSP. Things that are easy to do in EE are hard to achive in code and vice versa.

e.g. you can create the most crazy wave-shaper the world has ever seen with a simple b-spline or whatever that can be tweaked in various ways (Smart Electronix's Cyanide uses this). This is almost impossible to recreate in real world but pretty simple in software.
They build complex cirquits (well, not too complex, OK) to achive amplitude-modulation which is a simple multiplication in DSP.

So sometimes you have to abstract some hardware-stuff while it's pretty easy to add some creative new stuff.

[Urs]

The first being a programmer expiriements blindly with his computer swapping in different code and numbers until it sounds good. The other being an engineer starts the design from schematics and theory which is then developed into software.

I'd say most times it's inbetween. No serious software developer works blindly without thinking just like no EE writes a line of code without having learned the programming language.

Most principles in DSP are unbelievably easy if not trivial. Just look at the source code examples on the web. However, this has nothing to do with the musical feel & flavour. And knowing what's going on in an analog circuit does not make the DSP knowledge any better. Transforming a circuit into code is relatively easy to achieve, but transforming the sound of a circuit into code requires knowledge that goes beyond skills in programming or EE.

Building great sounding audio software requires creativity, no matter if your product is meant to mimic something existant or if it's something new.

Skills in EE and CS can be gained easily be reading, experience, exploration etc., you don't need to have a degree in that. You also don't need a degree in AE to create the sound you're after. Qualifications might be helpful, but they are not mandatory for success. They can sometimes even be an obstacle ("It must sound identical, because the maths are perfect").

Now, basically what I want to show is, the skills you need to write good audio software are passion, creativity and the gift of being able to look beyond the border of your knowledge. No one brings all the knowlege needed per se, even the most accurate mathematic model does not sound like the real thing, even the most elegant looking code can be improved, even the most transparent mix can be made more transparent.

Also, there's a fundamental difference between analog gear and software implementation. The software implementation can only sound as good as a digital recording of the original. This might be a lesser problem with effects boxes, but it's definately a big problem where air or microphones are involved, such as speaker cabinets, accoustic instruments etc. You can simply not expect your most expensive Genelec monitors to sound like a Marshall or a Stradivari, no matter how good the plugin is. You can only expect them to sound like a recording. The same accounts for Neves, SSLs whatsoever. It can be disappointing to encounter the loss when you finally record your proper sounding master mix from the console to 44.1kHz CD format. That said, the level of abstraction in any software implementation is pretty high, even on the side of sonic properties. I wonder how this affects the work and skills required for people who try to create software simulations, dunno.

Cheers,

Urs

[Rangtangtang]

What, what, what, whats going on, what, a emulation of a recording, a recording, a recording, a recording, a recording, a recording, a recording, a recording, the sound of a moog through speakers or the sound of a recoded moog...........?????????

[Funkybot]

I'd like to know the percentage of designers of amps, compressors, eq's and the like that were actually audio engineers. I mean isnt an assumption that a lot of them weren't just EE's who were working on audio gear to fit a given spec anyways?

If you're a TapeOp reader and read the articles about the guy's designing new boutique hardware (and occasionally guys from bigger companies like Presonus), you'll realize that most of them actually ARE audio engineers, or at least started out primarily as such. You'll also realize that like plug-in developers they need to test their designs by giving them out to studios who'll use them in real world situations, and oftentimes the engineers will come back with suggestions about everything from sound and functionality (just like software beta testers).

In the early days this may not have been the case but keep in mind that Fairchild, Neve, Trident, etc., are only a few of the companies that must have been creating gear back then. It's the better gear we remember, but there was probably a lot more crap available at the time that probably was just being built to spec. We just don't know about them, cause no one talks about them. It's like listening to a classic rock station and saying "man, music was just better back then." Sure it is! When you're only hearing the hits, from the better bands and not listening to any of the thousands of bands from the 60s and 70s that sucked and went nowhere.

Also, I'm not all about vintage per-se. If I had money for a vintage Neve strip, I'd probably buy a Chandler TG-1 instead (maybe even two) then a few mics as well with the savings, and I'd end up with more and probably better (though this would obviously be objective) sounding gear. So, vintage (as in, it's old so it has to be good) doesn't mean shit to me.

Anyway, there's some fuel to add to the fire, and maybe something to even think about.

[loophead]

Could any of us in a blind test say 'what speakers are in this marshall' ?

Could you, in a translatable way, explain the physical properties of a speaker cabinet in action. To what depth ? In what level of scientific measurement ?

Could you tell me the difference between this or that model Neve, and how it differs from this or that model SSL... in a quantifiable way.

No 'emulation' is anything more than a 'map'. The real world is real. The soft world is NOT.

One brings to bear what skills knowledge, and as URS says sense of adventure to bear on a given issue. Sometimes we make great strides sometimes we dont.

I find it quite apropos that a user is upset with a designer for his 'perceived' lack of knowledge in some area of his creation... which I emphatically disagree with (why it is not your lack of understanding with dsp)... thats ok - we are only talking here.

It is just that I find it amusing that these perceptions act like blinders on both sides but no one is going to see there own set of headgear... thats why we call them blinders... one is also blind to the fact of wearing them.

Who cares who built it. How they got to it. Do you find it useful ?

What gauge strings were used in that piano emulation...? who cares !

[Funkybot]

Loophead you have a point, but so does the original poster. To not know what type of speakers were modelled in a cabinet is a huge issue IMO. There's so many different kinds of Celestions in the Marshall range by itself that make a huge difference on the sound (something anyone would clearly hear), and to not know which were modelled by a certain would be quite disheartening. That'd be like a mastering engineer not being sure if he used M-Audio BX5's or Genelecs. Does this knowledge necessarily degrade or improve the quality of one's work? No, but it's certainly something that would dishearten some potential clients.

[mm_FX]

Just an extra £0.0109734 :

A lot of programmers I know simply don't have the time to make as much music as they would like. It gets hard when, after staring at a computer screen 25 hours a day (usually while stressed out and strung out on caffeine) you need to look at a computer screen again for creative reasons :/

And yes, I definitely think that coder types can fall into the trap of only listening to things in isolation, or paying more attention to analysis tools then their ears (just as a musician can while demoing an instrument/effect for example).

Therefore, most coders are heavily dependent on user feedback. Something I've heard time and time again, both from programmers and from beta testers I know, is that the problem a lot of the time can be a lack of tangible, quality feedback from users and beta-testers.

First of all, good, diligent beta testers are very hard to find these days. Secondly, good, diligent beta testers who have a really good ear, both musically and technically, are even harder to find

The problem is that most people who really know their chops inside out, in engineering terms, often simply don't have enough time away from client-based work to help a programmer out with detailed feedback.

If a software company/bedroom coder really wants to pay attention to this stuff to a very high level, they can't get away from paying for the expertise. This of course is incredibly difficult or impossible for most small coding operations.

[xsin1x]

marketing (or greed)will ruin all good products.
haven't you guys heard of the new, fftbasedspectralultrawavemorphingdistortameterfilterama?
it rocks!

[championrabbit]

What gauge strings were used in that piano emulation...? who cares !

Either it's important or it isn't.

If it isn't important then we can stop worrying about samples and emulations and models and impulses.

Who cares, right?

If it is important then the answer to your 'who cares' is 'most users care'.