KVR Audio

fabi · Post by **fabi** » Sun Nov 20, 2005 10:37 am

bmanic wrote:Some paypal warnings, take them serously:

http://www.paypalwarning.com/

http://www.paypalsucks.com/

- bManic

a lot the information on those sites and the sites to which they link seem outdated, though. not all of it, mind you. so read up, guys, and be careful.

dbmcclain · Post by **dbmcclain** » Sun Nov 20, 2005 11:08 am

WOW! The PDBHS (= Put Dave Back on the HotSeat!) Campaign now has $340 donations + $340 matching funds! More than 25% of the way there! But I won't start sweating just yet...

C'mon AMD'ers this is your chance to show that AMD kicks Intel's butt! As long as we rely on Intel to generate the code for PLParEQ3 you can guess who's going to win that contest!

The compiler folks at GreenHill have been writing compilers longer than I have (which is a considerable time!), and these guys claim to be super AMD-friendly. Check them out for yourself.

We already have the AMD Performance libs. Those were free for the asking... but right now we rely on the Intel Reference Compiler to gen the code...

Cheers!

daverich · Post by **daverich** » Sun Nov 20, 2005 11:21 am

Yeah just looked it up and PayPal in the UK is regulated by the FSA.

So in essence it does have to follow normal banking guidelines.

Kind regards

Dave Rich

dbmcclain · Post by **dbmcclain** » Sun Nov 20, 2005 11:29 am

Here's my pledge to the AMD crowd:

AMD and Intel have been warring over CPU designs for years. AMD has enough friction from this that in all likelihood they haven't made their libraries just knock-offs of the Intel versions.

So that means that if I kept the DSP core as is without changing a line, you might see better peformance... But you will probably see even better performance if I rearrange the DSP core to better suit the AMD library capabilities.

I pledge to you AMD'ers out there, that in standing with my own high standards for excellence in computing, I will make those special alterations for AMD versions. I will provide two AMD versions -- one for widest compatibility with legacy machines, and the other for maximum performance on the latest, hottest, AMD cores available.

So will the contest be fair? In a way, yes, because the DSP core is now specialized to the needs of Intel. So why not give AMD the same kind of head start by rearranging the code to best suit them as well?

Cheers,

- DM

Doug1 · Post by **Doug1** » Sun Nov 20, 2005 11:32 am

Sounds fair to me.

dbmcclain · Post by **dbmcclain** » Sun Nov 20, 2005 1:14 pm

Well, we have a confirmation from Markus, that there is definite hope for the AMD crowd to see improvements, even from the Intel-provided libraries.

While not yet perfect playback for Markus at QL 5, he was able to use one more filter before getting dropouts. That is a sign that we are headed in the right direction. His special compile took out all the stops on P4 instructions, but it still used an old performance library that knows nothing about SSE2 and higher instructions.

So there is definite hope on the horizon! Let's see AMD kick Intel's butt!!! I'll do what I can to make that happen! It will be an interesting contest...

BTW.. the PDBHS Campaign now has $380 contributions + $380 matching funds. Just under 30% of the way to the goal. Let's have some war chants for the AMD crowd!!

Cheers,

- DM

snooky · Post by **snooky** » Sun Nov 20, 2005 1:20 pm

I now mailed for the 3 band version again, I hope that it arraives

Rabban · Post by **Rabban** » Sun Nov 20, 2005 1:41 pm

dbmcclain wrote:So there is definite hope on the horizon! Let's see AMD kick Intel's butt!!! I'll do what I can to make that happen! It will be an interesting contest...

BTW.. the PDBHS Campaign now has $380 contributions + $380 matching funds. Just under 30% of the way to the goal. Let's have some war chants for the AMD crowd!!

Cheers,

- DM

I pledge to you AMD'ers out there, that in standing with my own high standards for excellence in computing, I will make those special alterations for AMD versions. I will provide two AMD versions -- one for widest compatibility with legacy machines, and the other for maximum performance on the latest, hottest, AMD cores available.

Please excuse my ignorant question (and poor wording of the situation) since I know absolutely nothing about compilers and programming (I'm just a musician, dammit):

Since you have Intel-specific DSP in the current version of your EQ, what will happen to the Intel users once you start using the AMD compiler (I'm sure you will get enough donations to get the AMD compiler)? I am an Intel user and I am wondering if the performance will decrease from the current state if you switched to the AMD compiler and started rearranging the DSP code to suit AMD'ers?

I am all for getting the best performance and I know that there are many AMD'ers in this crowd but I am just curious what happens to the Intel'ers when you start using the AMD compilers. Would Intel performance remain the same or start to decrease with the AMD-friendly code?

Or would you have an Intel-specific version of PLParEQ as well as your 2 AMD versions?

Again, sorry for the possibly stupid question since I have no experience in this particular field.

Cheers,

Rabban

dbmcclain · Post by **dbmcclain** » Sun Nov 20, 2005 1:47 pm

Fear Not!! I will not stop providing the Intel-based code. After all, I am an Intel user too!

What will happen is that the Zip file will just grow to include two additional DLL versions. Two for Intel, and two for AMD.

--------------
On another topic... I'm no taxidermist... So if some experts on skinning out there will contact me to discuss what is required to make this skinable, then you can begin to enjoy a new look too.

Otherwise, we will all just have to live with the "Mr Science" GUI...

(no threat intended, but I honestly don't know how to do skinning. I think it takes someone about 20-30 years younger than me to do it...)

------------
Oh! One other thing... Markus informed me that he was able to use 1 fiilter at ASIO 1024 without dropout, but the second one started making dropouts. At ASIO 256, he still has dropouts, but fewer of them. All at QL 5.

We want to be absolutely straight and honest about everything here. But this is still progress!

Cheers!

- DM

Rabban · Post by **Rabban** » Sun Nov 20, 2005 2:10 pm

dbmcclain wrote:Fear Not!! I will not stop providing the Intel-based code. After all, I am an Intel user too!

What will happen is that the Zip file will just grow to include two additional DLL versions. Two for Intel, and two for AMD.

- DM

Wow, thanks for the extremely quick reply. Talk about catering to both camps - 2 Intel versions and 2 AMD versions.

I am glad I donated $20, but I think it is not enough. I might have to donate a little more for the great support for a wonderful product!

Cheers,

Rabban

politcat · Post by **politcat** » Sun Nov 20, 2005 2:44 pm

<bump>

politcat wrote:
dbmcclain wrote:... but it works okay on a Celeron? I always thought those were just sawn-off Pentiums. Which version, plain? or 'NP4? on the Celeron and Pentium?
celeron...yeah it's got me wondering too. but, what do i know.

i always use the plain version on both.

any news on this front?

gd_vm · Post by **gd_vm** » Sun Nov 20, 2005 3:21 pm

Received my invoice....thank you David. C'mon guys!!! DONATE!!!

I am an AMD user and have had no problems in wavelab...all 3 band, QL5.
Havent really tried tracking with it, Ill give that a shot later today.

dbmcclain · Post by **dbmcclain** » Sun Nov 20, 2005 4:21 pm

Wow!! We now have $440 donations + $440 matching funds! That pushes us almost 1/3rd of the way to the goal for the PDBHS Campaign! Terrific!! (.. but still not sweating yet...)

... okay here's a motivator for all you Intel folks who think you have little to gain from this part of the campaign...

You are likely going to buy another computer, and probably sooner than you think...

How'd you like to participate in a runoff between Intel and AMD sponsored by all of us without any corporate ties to either company. Just some real tough, real-life, hard-core audio processing tests, carefully crafted for both processors by an experienced Scientist? Is $20 enough to be in on this wager? Do you think Intel will outperform AMD? Wanna put your money where your mouth is? I am!....

- Cheers,

- DM

dbmcclain · Post by **dbmcclain** » Sun Nov 20, 2005 4:29 pm

... on another front, the Celeron?

I pushed my system hard enough to get the chug-chug of the locomotive on some pink noise here, in real-time streaming, in T2.

When I froze the track, no chug-chug, just pure blissful EQ...

So perhaps it is a host-dependent thing?

I also ran the tests in Adobe Audition. Same result.

- DM

politcat · Post by **politcat** » Sun Nov 20, 2005 5:20 pm

dbmcclain wrote:... on another front, the Celeron?

I pushed my system hard enough to get the chug-chug of the locomotive on some pink noise here, in real-time streaming, in T2.

When I froze the track, no chug-chug, just pure blissful EQ...

So perhaps it is a host-dependent thing?

I also ran the tests in Adobe Audition. Same result.

- DM

i use sonar home studio 4 . anyone have a suggestion?

10-Band PLParEQ is Here!