Hi-

I have some very exciting news from BionicFX. The beta is still being constructed. However, I would like to finally answer your questions about latency.

These are REAL test results:

CPU: AMD Athlon 3200+
GPU: NVIDIA 6800

BionicFX AVEX Engine
Processing: 1 tracks, 1 channels, 1 streams

Host clock: 44100
Host buffer: 512
Host latency: 11ms

AVEX latency: 2.08ms
Computing performance: ~0.0002 GFlops

---

What do these numbers mean?

The latency of the BionicFX engine is 2.08ms. (Not much more than native.)

If you have your host drivers set to a low latency, then the total audio latency will be host + 2ms (in this case 13.08ms total). On Cubase you should experience NO latency, and when we release plug-ins for Logic the latency will be as good or better than other plug-ins and far better than any DSP solution.

Disclaimer 1:
The complexity of certain plug-ins is going to increase the latency. But, this is where another number above is important. The BionicFX engine requires 0.00026 GFlops. This means that far less than 1% of your GPU is being used by the engine and 99.999% can be used for audio processing. Tests running algorithms which performed 5000 floating point operations (FLOPS) added virtually no additional time.

Disclaimer 2:
Increased plug-in counts will cause the latency to increase. How much? I can't say yet, but will share results here in the next few days.

So, thanks for your continued support and interest.

The digital audio revolution is coming soon!

j1

Edit: time 1s removed

you never replied to my beta request

and I don't make them often

Why it took 0.192s to render 1s audio? Does it mean this GPU-powered plug-in still consumes 19% of CPU?

DSP cards take 100-500ms?

Wow, when i used uad-1's they had nearly identical latency to what you describe.

What a crock of shit

Aleksey Vaneev wrote:

Why it took 0.192s to render 1s audio? Does it mean this GPU-powered plug-in still consumes 19% of CPU?

Ignore that number. It's meaningless outside of the engine and has no relevance to performance. (I should have left it out of my original post)

Edit: Error correction.

You must be kidding.. UAD-1 can get down to 64 samples. It uses either 1x or 2x multiple of your soundcards buffer size.

jamesOne wrote:

Robert Randolph wrote: DSP cards take 100-500ms? Wow, when i used uad-1's they had nearly identical latency to what you describe. What a crock of shit Not true at all. The UAD requires a 4096 sample fixed buffer last time I checked. That's a lot more than 2ms.

You should check again. Refer to the post above mine. It goes based upon the host buffer.

jamesOne wrote:

Aleksey Vaneev wrote: Why it took 0.192s to render 1s audio? Does it mean this GPU-powered plug-in still consumes 19% of CPU? Ignore that number. It's meaningless outside of the engine and has no relevance to performance. (I should have left it out of my original post)

So you mean he picked up on something that shows that this "technology" has a flaw?

Im in a nasty mood tonight.

mike.0 wrote:

You must be kidding.. UAD-1 can get down to 64 samples. It uses either 1x or 2x multiple of your soundcards buffer size.

Great.

Then you can compare the performance of the UAD to your NVIDIA card at 64 samples at some point.

Good point.. but dont forget that its the character/quality of the plugins that really matters.

BTW, just to be a complete bastard.. I must tell you that my firewire powercore runs at 64 samples absolutely stable.

mike.0 wrote:

Good point.. but dont forget that its the character/quality of the plugins that really matters. BTW, just to be a complete bastard.. I must tell you that my firewire powercore runs at 64 samples absolutely stable.

This is good to know and I'm glad you are having fun with the PowerCore. It's a great product with killer plug-ins.

Original post edited.

I'm very excited about these plugins. I hope they use less latency than my UAD-1 and Poco and sound better. Why? Because then many of us will be able to use our gpu's instead of buying expensive DSP's. I just hope the plugins aren't priced too high so that all the benefits go out the window. Which is a tall order nowdays what with poco elements selling at $299 (with rebate in US) plus a suite of plugins.

Hum. yet another convolution reverb and Im wondering if it will work with ATI cards... will it make use of ogl rendering cards? Only the higher-end nvidia?

Will there be a DX version?

jamesOne wrote:

GPU: NVIDIA 6800 Time to render 1s: 192ms

Does this mean you could run five 1s convolutions on an NVidia 6800 with negligible CPU hit?

Hmm... five free convolutions...

Forever,




Kim.

Jeez wrote:

jamesOne wrote: GPU: NVIDIA 6800 Time to render 1s: 192ms Does this mean you could run five 1s convolutions on an NVidia 6800 with negligible CPU hit? Hmm... five free convolutions... Forever, Kim.

The time to render 1s is meaningless and should have been left out. It doesn't have anything to do with the performance.

The scalability won't be announced until the plug-ins are optimized, but here's a teaser: you can run more than 5.

(Thanks for the enthusiasm and the thoughts on pricing Munchkin. Your position on pricing is shared by a lot of people and duly noted.)

Randolph - ATI cards are 24-bit in the pixel shader, which is not good enough. They basically cut off the numerical precision. NVIDIA cards are 32-bit through the entire pipeline, and therefore sound better. DX and/or OpenGL specs will be announced later.

Last I looked the x800's are capable of 32-bit...

Then again I dont really care THAT much, cant stand ATI after that infamous SE incident.

Robert Randolph wrote:

Last I looked the x800's are capable of 32-bit... Then again I dont really care THAT much, cant stand ATI after that infamous SE incident.

The pixel shader is still 24-bit:

http://www.firingsquad.com/hardware/ati_radeon_x800/page3.asp

http://endian.net/details.asp?ItemNo=3906

http://www.hardwareanalysis.com/content/article/1710.2/

While "capable", the ATI cards still output 24-bits in the pixel shader hardware, then upsample to 32-bits.

This adds new meaning to the phrase, "it colors the sound," but it's exactly what happens to the audio.

I see.

jamesOne wrote:

This adds new meaning to the phrase, "it colors the sound," but it's exactly what happens to the audio.

Forever,




Kim.

So is a GF4 TI 4200 enough to run this plugin...? And will there be a way to run it in Sonar?

elwood wrote:

So is a GF4 TI 4200 enough to run this plugin...? And will there be a way to run it in Sonar?

No and surely Yes...

The only supported video cards seem to be the last generation of nVidia (6200/6600/6800)....
They are the only one which compute pixel shaders with a 32bits precision..

But i think that the motorala DSP of the powercore are 24 bits precision. So the 24 bits of the last ATI card should be enough to get good results....

Aleksey Vaneev wrote:

Why it took 0.192s to render 1s audio? Does it mean this GPU-powered plug-in still consumes 19% of CPU?

No, it means the plug used for testing uses 19% of the GPU.

In any case, I'm not shelling out on some megabucks high end graphics card for now... I don't play computer games, and my Athlon 64 will do plenty of SIR instances anyway

elwood wrote:

So is a GF4 TI 4200 enough to run this plugin...? And will there be a way to run it in Sonar?

TI4200:
With the TI 4200 it is not a matter of having "enough" processing power. The early NVIDIA cards simply do not have the hardware to support the processing.

Sonar:
Yes, Cakewalk was nice enough to send us a copy of Sonar to test the plug-ins. So, we will make sure it runs well in Sonar.

griels wrote:

Aleksey Vaneev wrote: Why it took 0.192s to render 1s audio? Does it mean this GPU-powered plug-in still consumes 19% of CPU? No, it means the plug used for testing uses 19% of the GPU. In any case, I'm not shelling out on some megabucks high end graphics card for now... I don't play computer games, and my Athlon 64 will do plenty of SIR instances anyway

Griels-

Hi. Good luck with the new website.

It has not been tested yet, but we are hoping the plug-ins will run well on the low-end 6200.

The PCI Express version of the mid-range 6600 are a lot less than $200. The 6200 should cost even less.

I have had the same good experience with the Athlon running plug-ins in ProTools.

jamesOne,

I know the benefits of true convolution vs fft based solutions.

Still it would be nice to hear short demo examples on the same IR using bionic reverb vs. maybe SIR or Voxengo Pristine Space.

Could you maybe arrange something like this? I'm sure most of us would be very interested to hear them.

jamesOne wrote:

[Sonar: Yes, Cakewalk was nice enough to send us a copy of Sonar to test the plug-ins. So, we will make sure it runs well in Sonar.

Why could you not try with the demo version of Sonar ?

Ixox wrote:

jamesOne wrote: [Sonar: Yes, Cakewalk was nice enough to send us a copy of Sonar to test the plug-ins. So, we will make sure it runs well in Sonar. Why could you not try with the demo version of Sonar ?

No VST<->DX adapter, one would assume.

JamesOne - $200 doesn't sound so bad... Sadly my mobo has no PCI Express slots (was too impatient to wait for PCI-X Socket 939 boards after my old machine died). Incidentally, are there any laptop graphics chips that support your plugins?

Kingston wrote:

jamesOne, I know the benefits of true convolution vs fft based solutions. Still it would be nice to hear short demo examples on the same IR using bionic reverb vs. maybe SIR or Voxengo Pristine Space. Could you maybe arrange something like this? I'm sure most of us would be very interested to hear them.

Not going to happen.

Our goal is to put out a great reverb as a demonstration of the AVEX technology and a glimpse of the future. There are a lot of plug-ins planned.

The guys that made those products (Christian and Aleksey) have made important contributions to the small studio market. I would rather see AVEX help their products at some point.

griels wrote:

Ixox wrote: jamesOne wrote: [Sonar: Yes, Cakewalk was nice enough to send us a copy of Sonar to test the plug-ins. So, we will make sure it runs well in Sonar. Why could you not try with the demo version of Sonar ? No VST<->DX adapter, one would assume. JamesOne - $200 doesn't sound so bad... Sadly my mobo has no PCI Express slots (was too impatient to wait for PCI-X Socket 939 boards after my old machine died). Incidentally, are there any laptop graphics chips that support your plugins?

Last time I spoke with their product manager we were tossing around the merits of the DX native vs. just running through the adapter. No decision was made, but they are convinced that it will run flawlessly through the built-in adapter.

They happen to be in the same city I am, so I will be spending some time in their office to make sure it runs well.

I just checked and the 6600 can be purchased through mail-order for less than $140+shipping. These will be even cheaper in the future.

I have not started testing the engine on laptops yet. But, I will find out from NVIDIA which chips have the hardware support and post it on our website at some point.

jamesOne wrote:

griels wrote: Ixox wrote: jamesOne wrote: [Sonar: Yes, Cakewalk was nice enough to send us a copy of Sonar to test the plug-ins. So, we will make sure it runs well in Sonar. Why could you not try with the demo version of Sonar ? No VST<->DX adapter, one would assume. JamesOne - $200 doesn't sound so bad... Sadly my mobo has no PCI Express slots (was too impatient to wait for PCI-X Socket 939 boards after my old machine died). Incidentally, are there any laptop graphics chips that support your plugins? Last time I spoke with their product manager we were tossing around the merits of the DX native vs. just running through the adapter. No decision was made, but they are convinced that it will run flawlessly through the built-in adapter. They happen to be in the same city I am, so I will be spending some time in their office to make sure it runs well. I just checked and the 6600 can be purchased through mail-order for less than $140+shipping. These will be even cheaper in the future. I have not started testing the engine on laptops yet. But, I will find out from NVIDIA which chips have the hardware support and post it on our website at some point.

Great, thanks! Wonder how important the AGP bandwidth problems are in practical terms (compared to the symmetric PCI-Xpress)? I mean, the PCI-based UAD/Powercore seem to handle a decent number of tracks.

In any case, I won't be shelling out for a new card in the near future, maybe some time next year.

By AGP bandwidth problems are you talking about the slow readback? We have found ways to resolve that issue and getting very good performance over AGP.

The PCI Express video cards are appealing, because of the bi-directional rates. But, if I am not mistaken they also use faster memory and that is where you might see even higher performance.

jamesOne wrote:

By AGP bandwidth problems are you talking about the slow readback?

Yes.

jamesOne wrote:

We have found ways to resolve that issue and getting very good performance over AGP.

Cool

jamesOne wrote:

The PCI Express video cards are appealing, because of the bi-directional rates. But, if I am not mistaken they also use faster memory and that is where you might see even higher performance.

jamesOne wrote:

Kingston wrote: jamesOne, I know the benefits of true convolution vs fft based solutions. Still it would be nice to hear short demo examples on the same IR using bionic reverb vs. maybe SIR or Voxengo Pristine Space. Could you maybe arrange something like this? I'm sure most of us would be very interested to hear them. Not going to happen. Our goal is to put out a great reverb as a demonstration of the AVEX technology and a glimpse of the future. There are a lot of plug-ins planned. The guys that made those products (Christian and Aleksey) have made important contributions to the small studio market. I would rather see AVEX help their products at some point.

Are you saying you won't do the comparisons because you want to be friendly with other developers and not compete with them?

Funny thing, I would've thought even christian and aleksey would like to hear some comparisons (if they haven't already)?

Hi Kingston-

From the E-mail response, I think a lot of people are going to be interested in how the plug-ins sound. This announcement is about latency.

You have requested that we post audio comparisons between our products and potential competitors or partners. I can't think of a single plug-in maker that offers this: here is how our product sounds and here are some links to files we made from our competitors' products.

You are welcome to make such comparisons yourself in the future. We will also be interested in the feedback from everyone on how to improve our products.

My sentiments are not about friendliness and/or competition. I do have a lot of respect for everyone doing DSP programming --- even making simple plug-ins is not a trivial undertaking.

If you have any additional questions about latency or performance of the BionicFX engine ask away, and thanks for the continued interest.

Kingston wrote:

jamesOne wrote: Kingston wrote: jamesOne, I know the benefits of true convolution vs fft based solutions. Still it would be nice to hear short demo examples on the same IR using bionic reverb vs. maybe SIR or Voxengo Pristine Space. Could you maybe arrange something like this? I'm sure most of us would be very interested to hear them. Not going to happen. Our goal is to put out a great reverb as a demonstration of the AVEX technology and a glimpse of the future. There are a lot of plug-ins planned. The guys that made those products (Christian and Aleksey) have made important contributions to the small studio market. I would rather see AVEX help their products at some point. Are you saying you won't do the comparisons because you want to be friendly with other developers and not compete with them? Funny thing, I would've thought even christian and aleksey would like to hear some comparisons (if they haven't already)?

How will bionicfx work in general? Is it a wrapper or chainer type container that loads normal native VSTi plugins? Or does it require special bionicFX-made effect plugs?

thanks

Markus

xRAVENx wrote:

How will bionicfx work in general? Is it a wrapper or chainer type container that loads normal native VSTi plugins? Or does it require special bionicFX-made effect plugs? thanks Markus

BionicFX runs audio algorithms directly on the GPU of your video card. This is an important distinction from DSP cards that employ the DSP chips as co-processors. Think of it as an external piece of hardware, but only has a 2ms response time and potentially 40+ GFlops of processing capability (although the observed is lower).

Plug-ins are being created by BionicFX. You can not simply port existing code, it must be reworked from the ground up. We will pursue partnerships at some point with existing plug-ins that we like and you recommend.

Have you thought about making an SDK available for plugin developers, so they can offer bionicFX versions directly?

I don't mean to sound arrogant, but I honestly hope the quality of the plugins offered will not make people fall back on using native versions, because these might sound better.
IR1 takes very little CPU on a grown up DAW, also there are very fine compressors and EQs available by elemental audio and sonalksis natively that eat little CPU.

I hope you guys will make some sort of SDK available to already established plugin developers for purchase.
An entirely proprietary approach would be a waste of the technology. I'm not saying you shouldn't get rich with this technology, I'm saying I hope you guys are getting rich in a way that leaves the user with a very wide choice of plugins to run on your technology, even when the plugins might not be bought via bionicFX.


best

Markus

Ixox wrote:

elwood wrote: So is a GF4 TI 4200 enough to run this plugin...? And will there be a way to run it in Sonar? No and surely Yes... The only supported video cards seem to be the last generation of nVidia (6200/6600/6800).... They are the only one which compute pixel shaders with a 32bits precision.. But i think that the motorala DSP of the powercore are 24 bits precision. So the 24 bits of the last ATI card should be enough to get good results....

well there goes the initial announcement of ALL FX series Nvidia cards being supported. screw this. I'm not gonna spend 1000 AU on a 6800,+ whatever each plugin costs, when I can get a powercore for that price, proven reliable technology with a highly respected brand name to back it up. did i mention, the 6000 series of cards sound like a HOOVER!
and what, 5 reverbs on a 6800 ultra? not very exciting. if it ws way more than 5 you would have said more than 10, for example

Quote:

We will pursue partnerships at some point with existing plug-ins that we like and you recommend.

This is just about the best news since the initial announcement. As if it wasn't exciting already. I can hardly wait.



Best of luck getting the good ones onboard.

*cough* Universal Audio *cough*

jamesOne wrote:

xRAVENx wrote: How will bionicfx work in general? Is it a wrapper or chainer type container that loads normal native VSTi plugins? Or does it require special bionicFX-made effect plugs? thanks Markus BionicFX runs audio algorithms directly on the GPU of your video card. This is an important distinction from DSP cards that employ the DSP chips as co-processors. Think of it as an external piece of hardware, but only has a 2ms response time and potentially 40+ GFlops of processing capability (although the observed is lower). Plug-ins are being created by BionicFX. You can not simply port existing code, it must be reworked from the ground up. We will pursue partnerships at some point with existing plug-ins that we like and you recommend.

one word comes to mind: Voxengo ... jeez if UAD doesn't snatch up Alexey real soon they will have a serious problem

ttoz wrote:

well there goes the initial announcement of ALL FX series Nvidia cards being supported. screw this. I'm not gonna spend 1000 AU on a 6800,+ whatever each plugin costs

Yes - I just checked out the Green Guide for prices on 6800 cards. My eyes popped at the 4 figure prices. I just did a search on some Oz sites for the lower models 6200 etc and could find almost nothing.

I guess these procesors will trickle down to lower price points in time. Here was me thinking my GeForce 4200 card was going to be my passport to audio heaven LOL.

The fan noise issue is a big one too.

Eg

JamesOne: When will the first synths be released? Tassman is a special candidate for AVEX ... or the waldorf synths. Do you have asked some of the developers?


Urs plan to release his plugs to the native platform. What's about asking him for porting his plugs also on AVEX platform? www.ursplugins.com

egbert wrote:

ttoz wrote: well there goes the initial announcement of ALL FX series Nvidia cards being supported. screw this. I'm not gonna spend 1000 AU on a 6800,+ whatever each plugin costs Yes - I just checked out the Green Guide for prices on 6800 cards. My eyes popped at the 4 figure prices. I just did a search on some Oz sites for the lower models 6200 etc and could find almost nothing. I guess these procesors will trickle down to lower price points in time. Here was me thinking my GeForce 4200 card was going to be my passport to audio heaven LOL. The fan noise issue is a big one too. Eg

At the end of the day it was supposed to even work on fx5200. any fx series card. so the developer started something and made all these waves and now it was all bullshit. nil, NIL interest from me. the fan noise is a BIG issue, believe me. and you can't get passive only cooling for those puppies.
did you know you can get 2 powercore elements for less than 1 grand aussie! not to mention all the included plugins. throw in a virus synth and i bet it would sitll cost less than a 6800 ultra and one bionic plugin

James,
Please ignore all the naysayers here, and continue on with this fascinating work.

People: these are preliminary results of newly emerging technology. Just to prove that it really can work. Give the guy a break. I wouldn't blame him if he stopped posting here until he has solid final stats, but I hope that's not the case.

What many of you don't seem to realize is that the processing power in todays $500.00 vid card will be available in a year or two for $150.00 or less, and generate much less heat than today's offerings. It happens all the time.

Quote:

hope you guys will make some sort of SDK available to already established plugin developers for purchase.

IMO I think this would be the key to being a HUGE success.

JonnySun 1.0 wrote:

JamesOne: When will the first synths be released? Tassman is a special candidate for AVEX ... or the waldorf synths. Do you have asked some of the developers? Urs plan to release his plugs to the native platform. What's about asking him for porting his plugs also on AVEX platform? www.ursplugins.com

We are investigating synths. Nothing is firm.

I want to do one, simply because the name: Bionic Synthesia is too obvious to pass up. (synethesia is the medical term for hearing colors, perceiving different senses together). It adds another meaning to "music that's out of sight".

ttoz wrote:

At the end of the day it was supposed to even work on fx5200.

The specs have not been published anywhere. Please point me at a specific post or location and I will correct it.

BK wrote:

James, Please ignore all the naysayers here, and continue on with this fascinating work. People: these are preliminary results of newly emerging technology. Just to prove that it really can work. Give the guy a break. I wouldn't blame him if he stopped posting here until he has solid final stats, but I hope that's not the case. What many of you don't seem to realize is that the processing power in todays $500.00 vid card will be available in a year or two for $150.00 or less, and generate much less heat than today's offerings. It happens all the time.

Thanks for the support BK.

Here's a couple of things to note:
1. We are coding for the future, not introducing a technology to band-aid previous generations of hardware. So, even if a 6800 sounds high now, it shouldn't down the road. And, to reiterate, that's the development platform, not the product specs.

2. To all of our Aussie friends. You should see less expensive cards in the future. There's always some distribution lag time. In the US, 6xxx cards range from $130 to $500 and these will become even cheaper in the future.

jamesOne wrote:

2. To all of our Aussie friends. You should see less expensive cards in the future. There's always some distribution lag time. In the US, 6xxx cards range from $130 to $500 and these will become even cheaper in the future.

Thanks for the thought James

God forbid that the streets of Melbourne should be lined with panhandlers holding up signs - "Brother can you spare a 6800."

To JamesOne:

What's about Matrox-Cards? As far as i know they offer triple-head-feature. That would be important to me and others if buying new computer and graphic card. NVIDIA-Cards seem to only offer 2-Monitor out ...

I would be glad if there is planned to support matrox cards. I don't need 40 Flops or more at the moment, but 3 Instances of reverb and maybe some effects should be managed by actual matrox cards, don't they?

JonnySun 1.0 wrote:

NVIDIA-Cards seem to only offer 2-Monitor out ...

The Nforce 4 SLI boards are going to offer two PCIE Video slots so you potentially put two FX6800s in there and, assuming all outputs are still live in SLI configurarion, you could get 4 Digital video outs and twice as much number crunching power.

Hm, what's about PCI Stress? (If a Powercore and UAD, maybe also Creamware Scope 14 DSP-Card is put in the mobo?)

JonnySun 1.0 wrote:

Hm, what's about PCI Stress? (If a Powercore and UAD, maybe also Creamware Scope 14 DSP-Card is put in the mobo?)

Well Sir - I'm glad you asked that!

With new PCI Express (with blue crystals) your PCI Stress days are over!

PCIE is not a shared bus architecture - every slot gets its own bandwidth which is more than the whole PCI bus got under PCI 2.x

In short, its a "whole new ballgame.

jamesOne wrote:

ttoz wrote: At the end of the day it was supposed to even work on fx5200. The specs have not been published anywhere. Please point me at a specific post or location and I will correct it.

when this was firts announced right here in this very forum it was made clear that ATi cards were not supported but all geforec fx cards were. i couldn't be botherd finding the thread myself, if you want to, be my guest, but I do have a very good memory.

regardless, 6800's are hot and noisy, and at the end of the day I could afford one if i really wanted to, but don't.
I will reevaluate this in 2 or 3 years time.
cheers and good luck anyways

Whos up for the bet, that there will be people saying:

"The Nvidia might be more on the power side, but an ATI Card sounds better (possibly because of the material they used to actually build the card) (Possibly because it was the same they use for High Quality Audio Cables)"


Also: Make it possible to have Synthedit exported to a Bionic plug, and you have a lot of friends.

I have really high hopes for this, and assuming the final product will have what it takes, I would definitely be ready to purchase a high-end graphics card to use it.

But I'm just very afraid that this will flop (and i'm not talking about a floating thing) and will not satisfy the expectations I have.
Not saying that I think it won't work out, i'm just afraid that it won't.

I'm hoping for the best!

jamesOne, this all sounds pretty interesting. However, convolution computation performance is the 'show maker'. Its real-time performance should be known before judgements could be made on this technology.

BTW, if GPU's performance is that great, I think it could be possible to 'emulate' 64-bit calculations which are necessary for FFT calculations.

skOre wrote:

Whos up for the bet, that there will be people saying: "The Nvidia might be more on the power side, but an ATI Card sounds better (possibly because of the material they used to actually build the card) (Possibly because it was the same they use for High Quality Audio Cables)"

this is all fascinating, but in the end the quality of an audio plugin always comes down to the DSP algorithms and the user-interface (ie parameters and their range) ...... look f.e. at the new princeton digital reverb, it only takes 5% of the CPU, but seems to sound much better than any other native reverb..... proof of the fact that not raw DSP power but the efficiency and smartness of the algorithms are important....

for the user interface: look at the UAD plugins, f.e. the LA2A and 1176 compressors.....just 2 or 3 parameters, and it doesn't matter how you set them, it will always sound amazing! this is a very big part of the succes of these unit, the parameter range and UI is so highly optimised, you can't do wrong with those....(imho this also is true for synths, f.e. look at the succes of the minimoog or the roland juno series, these synths have super optimised user interfaces)...

Olaf

JonnySun 1.0 wrote:

To JamesOne: What's about Matrox-Cards? As far as i know they offer triple-head-feature.

I doubt the Matrox cards will work. I will dig into the specs at some point, but don't hold your breath.

It must be possible ..

if we were serious about this there are cards out there that have 10 times the gigaflops power even of a 6800 ultra, at similar price points. but they are not gamers cards. but they do 2d just fine.

Such as?

ttoz wrote:

if we were serious about this there are cards out there that have 10 times the gigaflops power even of a 6800 ultra, at similar price points. but they are not gamers cards. but they do 2d just fine.

ttoz, good thoughts.

There are video cards that perform 400 GFlops?

Are you referring to the Quadro or Fire GL cards?

Last I checked, the Quadro line of video cards offered approx. 2x's the performance and cost around $2500. Different ballgame.

I would be interested if you can find some professional class cards that you think might work. A lot of the pro class cards use faster DDR3 memory and we will be able to quantify the difference in speed between the PCI Express cards and the professional cards.

There's a lot more to the performance of AVEX than a faster card, so even if the card is faster for 3D or CAD work it doesn't mean better plug-in performance. This is theoretical though, because AVEX has not been benchmarked on a wide assortment of cards yet.

Anandtech has offered to help with this in the future.

whyterabbyt wrote:

Such as?

i talked about them in the first thread.
oh well, i'll go look it up.

ttoz wrote:

jamesOne wrote: ttoz wrote: At the end of the day it was supposed to even work on fx5200. The specs have not been published anywhere. Please point me at a specific post or location and I will correct it. when this was firts announced right here in this very forum it was made clear that ATi cards were not supported but all geforec fx cards were. i couldn't be botherd finding the thread myself, if you want to, be my guest, but I do have a very good memory. regardless, 6800's are hot and noisy, and at the end of the day I could afford one if i really wanted to, but don't. I will reevaluate this in 2 or 3 years time. cheers and good luck anyways

You haven't seen the passively cooled 6800 Gigabyte card then? The 6600 series will be pretty cheap too, and aren't the 6200 even cheaper?

oh SHIT. here is the (now defunct) link, from the original bionic fx thread.
http://www.e-pc.com.au/product_info.php/cPath/33/products_id/223

it doesn't seem they are even a computer supplier anymore.

the basic card was like 200 gigflops at around 1100 AU.
I can't even remember the name of the actual card/brand.

Chris Ochre wrote:

ttoz wrote: jamesOne wrote: ttoz wrote: At the end of the day it was supposed to even work on fx5200. The specs have not been published anywhere. Please point me at a specific post or location and I will correct it. when this was firts announced right here in this very forum it was made clear that ATi cards were not supported but all geforec fx cards were. i couldn't be botherd finding the thread myself, if you want to, be my guest, but I do have a very good memory. regardless, 6800's are hot and noisy, and at the end of the day I could afford one if i really wanted to, but don't. I will reevaluate this in 2 or 3 years time. cheers and good luck anyways You haven't seen the passively cooled 6800 Gigabyte card then? The 6600 series will be pretty cheap too, and aren't the 6200 even cheaper?

I have now. interesting. just went to giga's site. it's the baby model with only 128 ram though

is that going to matter for audio purposes?

Chris Ochre wrote:

is that going to matter for audio purposes?

i don't know. only James can answer that

ttoz wrote:

Chris Ochre wrote: is that going to matter for audio purposes? i don't know. only James can answer that

What's the question?

Oh, ram? Specs haven't been announced. We need some more time on that one.

JamesOne: Don't forget: Without many and good plugs for a resonable price the new technology will disappear into nothingless i fear.

Best technology is useless if you don't have the tools to use it.

Same story with ATARI Jaguar: (for example) The world's first 64-Bit Video-Game Machine and rebirth of the ATARI Company. A machine with many and big muscles. But: There were no good and not many games for it. They were expensive, too.

Please try to look to this point. Givin' this more attention 'cause it would be to sad if such a good idea - converting audio data into graphical and back - would fail. I fear that.

JonnySun 1.0 wrote:

Givin' this more attention 'cause it would be to sad if such a good idea - converting audio data into graphical and back - would fail. I fear that.

exactly what I was saying.

Hi James,

How do the dual PCIE graphics SLI boards fit with your plans - these boards do are set up to process alternate lines as I understand it.

Eg

What is BionicFX?

OK... http://www.bionicfx.com/

But my English is just average... Will this thing make it possible to encrease the performance of any vst? Or does it require custom plugins?

psynus wrote:

OK... http://www.bionicfx.com/ But my English is just average... Will this thing make it possible to encrease the performance of any vst? Or does it require custom plugins?

Custom plugins.

nuffink wrote:

psynus wrote: OK... http://www.bionicfx.com/ But my English is just average... Will this thing make it possible to encrease the performance of any vst? Or does it require custom plugins? Custom plugins.

K. What a pity. Somehow. It's the same thing that keeps me from purchasing receptor...

November passed... december passed... And still no news about this??? Or did I miss something?

Lets wait for NAMM 2005

jamesOne wrote:

I want to do one, simply because the name: Bionic Synthesia is too obvious to pass up. (synethesia is the medical term for hearing colors, perceiving different senses together). It adds another meaning to "music that's out of sight".

What a pity that this thread is a zombie, I actually thought the news about bionic fx were recent.

Anyway,

(1) It's syne-s-thesia.

(2) Add the pards from 3D-labs (Wildcat Realizm) to the list, they should have the necessary 32 bit shader precision that is required for the DSP computations. Although they're not consumer cards.

Were they at NAMM? Any news?

I looked for them, but couldn't find them. It's possible they were in a suite however.

I'm still waiting patiently - any kind of update welcome...