Thinking ahead to the CPU power 2 years from now

DSP, Plugin and Host development discussion.
RELATED
PRODUCTS

Post

Trojan Badger wrote:Very low latency convolution reverbs will be a hot topic.

Plus there's capability for better physical modelling, especially for, say, wind and string instruments. A truly devastating "virtual guitar" could be on the cards.

And that of course opens the door for a physically modelled human voice - none of the current virtual vocalists based on samples, but a truly tweakable voice.

Truly.
Hrm we already have 0 latency convolution reverbs...

and we have tweakable human-style voices...

where ya been this last year?

Post

Robert Randolph wrote:we have tweakable human-style voices...
I'm pretty blown away by what I've heard of Vocaloid. I wouldn't have guessed we were that close to a totally transparent synth vocalist -- maybe 6 more years and you won't be able to tell it's synth at all. (Vocaloid could fool me in some but not all situations. -- uh let me add that by "fool me" I mean "sounds like a human voice being screwed up by a pitch corrector", not "sounds like a human voice recorded and engineered by people who aren't idiots")
Last edited by Borogove on Sat Aug 07, 2004 1:20 am, edited 1 time in total.
Image
Don't do it my way.

Post

Borogove wrote: I'm pretty blown away by what I've heard of Vocaloid.
what, with how pathetic it is? :hihi:

Post

I don't desire any more plugins or particularly more advanced technology (well, there are a few things I can think of), but it would be nice to have the CPU load always low enough that my host remains ultra-responsive. It'd also be nice for stuff like rendering to be quicker.
Every day takes figuring out all over again how to f#ckin’ live.

Post

Gotta agree there, considering all the hype. If they had advertised it as a shit-hot wordsynth then I might not be so appalled by it, but no - "you'll never need to hire a sessionist again";"let Vocaloid be the voice you've always dreamed of having" - yeah, i've always wanted to sing like a Russian undergoing nasal surgery!!!

Ok..i admit..having a bad day!

Nath

Post

ChimeraMan wrote:Gotta agree there, considering all the hype. If they had advertised it as a shit-hot wordsynth then I might not be so appalled by it, but no - "you'll never need to hire a sessionist again";"let Vocaloid be the voice you've always dreamed of having" - yeah, i've always wanted to sing like a Russian undergoing nasal surgery!!!
Heh, I managed to miss most of the hype, I guess. Mostly I had assumed that we were nowhere near the ability to do even that well yet.

Most of the Miriam demos sucked, also -- I don't know who this Miriam Stockley person is but I pity her now -- but some of the Leon demos really seemed to play to Vocaloid's strengths.
Image
Don't do it my way.

Post

If you compare vocoloid to modern pop vocals, vocoloid does indeed sound very human (if you using modern pop as a reference).

Being that people process vocals so much with pitch modulations, un-natural edits, double takes etc... its no wonders many people who feel vocoloid is very close to a natural singing voice.

Post

If you don`t know it yet Miriam Stockley is the singer from Adiemus, you take the composer Karl Jenkins He then makes a composition of Classical Singing mix it with tribal rythms, and the angels sing. Think Enya Mixed with Ravel and Deep Forest, Without Electronics, and yes you have to feel sorry for her if vocaloid is bad, because she is a great singer.

Post

About Whitout Electronics, that will be Electronic Instruments, I do Think Theres is Used Effects a great deal in Adiemus Albums

Post

I'd just love to be able to have audio at a really high samplerate, and with 64bit precision. Currently this isn't really an option.

Post

I think the power race is gone. With the 486, we had the 10, 20, 33, 40, 66, 80 MHz, and now, we have the 1, 1.1, 1.2, ... The speed isn't increasing like before. Engineers face new problems, and the Moore law isn't valid any more.
I wish we could have a new architecture, more massively parallel... But not dual-core or non-sense like it, not before the bottlenecks are resolved, but that is part of the research field, and no matter what they say, multiple processors stuff is not working as they want : tey can't provide them enough data in an efficient way.

Post

With a bit of luck, maybe someone will come up with a synth that makes sounds in a completely unique and currently impossible way.

I was well excited when the latest additive softsynths came out (still saving for cameleon and dopplemangler), as they offered a way of creating sounds in such a different way with resynthesis. Granted additive technology has been around for yonks, but I've never known it implemented so creatively.

I'm hoping there'll be something else to push the boundary back.

Something I can't yet imagine! That'd be more interesting to me.

:)

Post

Hi

I find it amusing to think of how big the average Hard-drive will be or the Bigger sample libraries/romplers that also seem to get bigger in lieu with CPU speed; so second guessing, in 5 years time the standard H-Drive will be 2 trillabytes and the top notch romplers will have 96,000 presets spread over 20 Dual-Quad layered DVD's giving 1280 gigs of library.

MMMmmmmm

Flipper.

Post

The standard 'hard drive' will be a holographic storage medium, which will offer instant data access to entire pages of non-volatile memory. This will be fu***ng fast! You'll also easily get huge storage capacities.

I've heard that there is a memory device that should be around next year that offers something like 2 terabytes of stoarge, in a package the size of an SD card.
Dunno if it is vapourware though.

Post Reply

Return to “DSP and Plugin Development”