MOTUnation.com

I was just talking to somebody today that knows and the way I understood it is that things like plugs (i.e. eVerb) that Motu includes with DP use Altivec but not MAS itself. I think that makes a great deal of sense because the plugs usually need the complex math. A lot of what MAS does is move data around. Mixing involves some multiplication but the plugs need the real number crunching power. I know that certainly the case with Altiverb and will be curious as to how the Intel version performs.

Aramis wrote:My opinion is that anything Altivec is optimized ....by definition ....

Aramis

Say what ?!

Code optimization is related to programming economies and how
compilers are adjusted by a coder to produce optimum or efficient results.

I doubt MOTU has done *everything* possible in DP to produce
the highest code optimization possible.

Shoot me...

Kris...

bang

I agree with mastermix.

Giving an example for differences between,
"supporting" Altivec and "optimized" for Altivec.

If a person could play baseball no matter how bad or good
but could actually play, this is "support".
On the otherhand, if a person is really good at playing
baseball even in Major League, this is called "optimized".

Sorry if this example is confusing....
I just want to say its a big difference between,
"could do it" and "can do it and being very good at it".
And thats the difference between, "support" and "optimized".

Oh well, this is just my opinion anyways...

It would be great to get an official comment about this from MOTU,
but I think their busy developing the IntelMac version and the updates
for DP5 for PPCversion.

P.S
I really appreciate all your replies. Thanks

MOTU is a bunch of really sharp, competitive guys, and the notion that they would not take full advantage of Altivec, which has been around since before DP v.4.00, I find absurd. Remember, DP is MAC ONLY, so Altivec is central to everything they do, unless you think they're still writing the application for guys with G3's.

tommymandel wrote:MOTU is a bunch of really sharp, competitive guys, and the notion that they would not take full advantage of Altivec, which has been around since before DP v.4.00, I find absurd. Remember, DP is MAC ONLY, so Altivec is central to everything they do, unless you think they're still writing the application for guys with G3's.

I may be naive but I agree with that statement ......
I frankly believe they do use Altivec in the best way they can .....
I use MOTU stuf since long time and they have always been one of the best piece of software on the Mac ....
Aramis

i'm pretty sure i remember someone from motu explaining here on this forum (before the big crash) that dp was not optimized for altivec and that the amount of work it would require would not yeild significant enough gains to warrant the cost of such an endeavor. seems like that was a good move considering the intel switch.

jr213 wrote:the amount of work it would require would not yeild significant enough gains to warrant the cost of such an endeavor.

I`ve heard the same. But not here.

What I`ve heard other then this is,
the programmers for DP changed many times for the past years,
and the programmers who came afterwards just
analyzed a software which had no documentations or data about it,
but just the sorcecode. And they`ve mostly extended it,
and not re-write it. They never made clear specifications for DP
to be passed on, or keep full complete documents of this software for the future back then.
And the codes left are very old, and haven`t been designed with anticipation.
They would need to rewrite everything from scrap.
And its not cut out for multiple specialized development teams
to work on one software, like in these days.
Theres really alot of patch work done every where inside of DP.
DP was never meant to live this long,
and there wasn`t anyother software that lived this long either at that time.
They didn`t look 10 years or 20 years ahead of them back then,
when they started building DP.
So optimizing DP for Altivec would cause them to rewrite this
almost from start. And I also heard that this also maybe causing the
Virtual Instruments non-efficiency, and if you want to make the VI`s
efficient too, they`ll need to reprogram this too.

Again, this is just the things I heard.

You can be sure of this much: when DP goes Intel, it will HAVE to be rewritten from the ground up. The programming languages which were around when DP was written will not directly translate to Apple's Intel environment. If MOTU has been nursing legacy code for 20 years now, it will soon bury it and start anew. That could be very good news for us. In fact, I've been wondering if DP5 is partly the product of this effort. It has bugs that shouldn't be there if they are using the legacy code. Maybe we're currently beta-testing and debugging the new, blindingly-fast DP for Intel Macs.

Shooshie

A rewrite of DP is very unlikely. --

Code improvements is more likely.

If MOTU could, they would have done it by now.

Getting rid of the MAS engine and rewriting a new audio
engine would be a herculean task. The economics of such
an effort is the real show stopper. Many things underneath..
such as Protools and legacy hardware compatibility would
suffer a great deal.

Just observe how slower DP gets with each new version...
factoring in all the periodic Apple hardware improvements.


Kris..

I hope MOTU rewrites it for the Intel version..
But I have the same worries as mastermix.
If they don`t rewrite it, it will get more heavier and heavier.
I heard that its really the core part of DP thats making it heavy.
Like the part where DP tries to sync MIDI and MAS.
So as long as they don`t rewrite MAS or DP, it wouldn`t get
CPU efficient.

Just tested DP 5.01 with a project that had regular cpu spikes, glitches, etc. This project was created with DP 4.61 and completed with DP 5.0. I just opened it to see if things got any better with 5.01.

Nope.

MOTU: Your number one issue is CPU optimization. Most other features work great. However, in the 21st century where we can now use VI's for an entire production, DP needs to evolve. If that requires a total rewrite (which is *highly* likely for Intel), then so be it.

Very funny thread ....There is a lot of , ( I heard ) and suppositions here ....
When I see some VI's being updated to solve problems with Logic ,
I question the facts that some of you are trying to establish here ....
I have seen spike problems in my DP that were occuring only when using some patches in Groove-Agent-II for exemple ....So I think it is not so simple to compare a Daw to another ....

Aramis

Ah! I agree with you Aramis.....