Support

If you have a problem or need to report a bug please email : support@dsprobotics.com

There are 3 sections to this support area:

DOWNLOADS: access to product manuals, support files and drivers

HELP & INFORMATION: tutorials and example files for learning or finding pre-made modules for your projects

USER FORUMS: meet with other users and exchange ideas, you can also get help and assistance here

Perceptual Processing vs. Digital Signal Processing.

For anything not directly related to FlowStone

Perceptual Processing vs. Digital Signal Processing.

Postby trogluddite » Fri Jun 07, 2019 6:16 pm

A rather fascinating, though somewhat off-topic, subject of discussion came up somewhere around HERE in another thread. Namely to what extent, and for what reasons, might our plugin GUIs affect listeners' perception of their sound qualities. On the one hand, LeAttol has, somewhat controversially, hypothesised that interactions between GUI and DSP code might empirically alter the sound, whereas most others argued that this is, if anything at all, a cognitive effect of the brain. Picking up from where this led us (and with apologies to adamszabo, singled out here only because his post most interested me as a launching off point)...

adamszabo wrote:I thought we are creating plugins/applications using strict logical guidelines and not philosophical stuff. Multiplication mutiplies, Addidion adds, Subtraction subtracts and so on. I dont think we would be able to program at all of people say that sometimes multiplication creates an addition or any other nonsense. The same with colors, its not taught in schools and not a confirmed phenomenon that a red gui always makes the plugin sounds better. Those are just experiments and not facts.

I would not argue with this in a technical sense - data and program instructions are held in the same memory (the von Neumann architecture of most modern computers), and the CPU has no idea what human perceptual or cognitive phenomena data might represent. If visual data could inadvertently "leak" into audio data, there would be nothing to stop similar "leakage" into the CPU's instruction pipeline, the effects of which would be rather noticeable as random crashes!

However, just because something cannot be reduced to mathematics or formal logic does not mean that it isn't an empirical fact. The specifics of "red == better" may not be a "fact"; but that, more generally, our senses do not operate independently of each other is certainly an experimentally confirmed empirical "fact" about human neurology. I'd encourage any sound-engineer, musician, or plugin designer to take a little interest in perceptual processes, as our understanding of maths and logic does not make us immune to perceptual distortions, and it can suggest interesting new ways to manipulate audio. Hence it is advised (or at least should be) that it is wise to turn off visual displays (whatever their colour!) when listening critically; because it is well established that visual stimuli (among other things) can and do affect our judgement of sounds - and that our minds are very good at hiding this from us.

Regarding LeAttol's hypothesis of a link between GUI code and DSP code, it is reasonable enough to criticise this within the context of the theories of computer science. However, if we find no answer there, this in no way proves that he is not genuinely perceiving a correlation between them - something impossible to prove unless Spogg can teach us all how to do a Vulcan mind-meld! If we look to perceptual psychology, though, we can begin to see why this correlation might be apparent to him, and possibly for many other people. Discussing such factors seems a far kinder and more enlightening way to debate such apparently outlandish theories than accusations of delusion or trollishness because we refuse to look beyond the narrow confines of DSP/computer theory for an explanation.

Lastly; the above may seem like I'm proselytising about a pet subject; and that's exactly what it is! As a consequence of being autistic, many of my perceptual processes work somewhat differently than for most people. Many things normally accepted as "self-evident truths" don't seem at all self-evident to me. Conversely, for decades I believed that everyone else in the world felt their head change shape when they walked across a patterned carpet - this synaesthesia seems as "real" to me as any other sensory experience. Perceptual psychology may not be such a "hard" science as computer science, but when you have wonky perceptual processes like mine, it can be a very useful tool indeed!
All schematics/modules I post are free for all to use - but a credit is always polite!
Don't stagnate, mutate to create!
User avatar
trogluddite
 
Posts: 1205
Joined: Fri Oct 22, 2010 12:46 am
Location: Yorkshire, UK

Re: Perceptual Processing vs. Digital Signal Processing.

Postby wlangfor@uoguelph.ca » Fri Jun 07, 2019 9:31 pm

Hmm, well; I guess the realization and impact of graphics as a whole are, when in regards to vst in general not necessarily in the sense of merely a graphics card.

Unlike other types I guess the video end is merely done by the graphics card like in the case of Juce. I remember noting that a great deal of the popular plugins made on the platform were driven by the free graphic driver architecture sometimes included in graphics cards. (starts with g or something).

Anyways, this was the only rationale I could respect in the sense of sounds seeming of a slightly different colour when using certain colours or imagery in various formats.

And this, though whole-heartedly in theory; is why I wondered if the graphics end of a vst is not entirely driven somehow in the vst; To then belay issues with graphics card entirely. I believe that would be an intuitive approach. And that is obviously where c++ and the vst framework might differ? Perhaps I was thinking that flowstone is relying on legacy graphics interpretors, ergo software rather than hardware. And as a product of this "fix" of graphics issues, therein became this precedent seemingly in correlation with the colours and graphics.

What it seems is that black with a bit of white seems a bit off, but the right combination of colours that reach the rgb values of 255 in great quantity seem to have the most affect. So, in thinking this through; The creator wanted the audio firstly processed, and then the graphics. Meaning that probably the only way to do that would be to put that data in queue. And it is because of that queue that sound is affected for the data is emergent while processing both mediums.
Download VST: LA Bands 15 Band EQ
My Website http://www.leattol.com
User avatar
wlangfor@uoguelph.ca
 
Posts: 508
Joined: Tue Apr 03, 2018 5:50 pm
Location: Penetanguishene, Ontario, Canada

Re: Perceptual Processing vs. Digital Signal Processing.

Postby Duckett » Sat Jun 08, 2019 2:07 am

An interesting perspective- as far as the perceptual side of things, Pantone 448 C is used on cigarette packaging in Australia, as marketing researchers found it to be the least attractive color.

As far as the, ahem, "discussion" on this forum being referenced goes, it reminds me of KVR forum threads, in a way- often people who care enough/are invested enough concerning a niche topic (such as VST plugins) to spend a significant amount of time on forums about them tend to be rather, ah, passionate about their views, and just like Bach was trolled by one of his sons (Bach had gone to bed, and the son played on the piano all but the very last note of a piece, and waited; after a time, Bach stormed downstairs, played the last note, and stormed back up again), people can be absolutely unable to let something be, regardless of how little the matter is of concern to others.

There's a reason why this cartoon https://xkcd.com/386/ is so well-known.

Meanwhile, I'll continue to be fantastic at working in FS on synths and effects that are light-years behind in architecture, efficiency and appearance compared to what was made in SM 10 years ago, let alone today! ;)
We have to train ourselves so that we can improvise on anything... a bird, a sock, a fuming beaker! This, too, can be music. Anything can be music. -Biff Debris
User avatar
Duckett
 
Posts: 25
Joined: Mon Dec 14, 2015 12:39 am

Re: Perceptual Processing vs. Digital Signal Processing.

Postby wlangfor@uoguelph.ca » Sat Jun 08, 2019 6:14 am

Yeah, Sm had a way around some of the issues with graphics; and I liked that. An equalizer could be made in a better way. That's too bad; still though; some of the examples were way to complicated. But I mean, it's not like there's ever been constraints. Only, that now there is a greater memory issue imposed in fs.

But, I mean We can still make great music. I use fl 10 because it is faster. But I mean when we're making 64 bit plugs they'll keep up. the fl 12/20 bridge is slow. Jbridge would probably be faster.

Anyways, on the bright side; Here is My new track made with mostly My plugs (containing code made by people here too!) and it sounds so far so good :).

Image
(click to watch Youtube vid)

LOL. Styles of My smooth plugins. And BTW, don't be hating on Me unless You can make a track as well eh? If You can tell Me what sounds better - Then You can also prove that You can mix and make it sound better.

People tell Me theories? represent. If You can drop a track, then good; Otherwise everything negative said - Theory. All theory.
Download VST: LA Bands 15 Band EQ
My Website http://www.leattol.com
User avatar
wlangfor@uoguelph.ca
 
Posts: 508
Joined: Tue Apr 03, 2018 5:50 pm
Location: Penetanguishene, Ontario, Canada


Return to Off Topic

Who is online

Users browsing this forum: No registered users and 1 guest