Digi-Why was tiger better then Leo using all cores?

[Fidelis]

Ahh... and i just did read in this forum that people with newer MAc Pros are having the problem and have to reduce the number of processor as well. So, it's not a problem related only to G5! Something changed in PT8 regarding RTAS power. We just want to know why.

[woodsdenis]

Originally Posted by Fidelis

Ahh... and i just did read in this forum that people with newer MAc Pros are having the problem and have to reduce the number of processor as well. So, it's not a problem related only to G5! Something changed in PT8 regarding RTAS power. We just want to know why.

It is true to say that the problem exists on all systems, however with a G5 with only 2 processors the problem is way more apparent because reducing
the processor count by -1 leaves you with just 1 for all your RTAS stuff.Not good.
On an 8 core Mac, if you do this the difference is minimal if any. Even if you have a Quad G5 it seems, by reading posts that this is an acceptable workaround.

I said previously in this thread that this problem sits with Apple/Leopard, Digi/Rtas and the software developers. No one is totally at fault. I dont have any issues with the AIR/RTAS VI's in PT8/G5 on my system running both processors. However throw in Omnisphere and the stuttering playback etc starts.

I know, for example, that Omnisphere works much better as an AU on Leopard. One of the main reasons for this I can only assume is that it has to use a VST wrapper for PT8, so does Stylus. You cant blame Digi for that. Spectrasonics view is that this is an acceptable way to do this and has worked well in the past.

http://www.northernsounds.com/forum/...hlight=wrapper

To be honest I cant argue with them I know nothing about coding, but is it really that difficult to code/convert for RTAS. If it is, thats what we should complain about. This is not about taking sides. or in this case taking Spectrasonics to task. I use them as an example as they are a major player in the VI market and code for every major format I know of.

I will however take the DigiTechSupt to task on a quote about Omnisphere and working on a G5.He/she used it as an example of how a G5 in the present day wasn't capable or was barely capable of running this synth. My G5 is the min spec for Omnisphere, there is no doubt it is CPU intensive. The only platform it wont run successfully on is PT8. Reducing a processor is not an option.
It will run on PT7.4/Tiger Logic /Tiger/ Leopard etc with both processors. In the interests of fairness that should be pointed out. It is in my experience it is the most CPU intensive plug I have ever seen and to use that as an example wasn't fair. On the other side I totally accept things march on, and these issues will come up again and again as all the new stuff requires more horsepower, thats life.


I think the real question is why is it so difficult for software programmers to
adhere to the RTAS specs in the way that AIR do. Does it take to much time?
Are Air privy to some secret code that others aren't? Is it too complicated ? This surely is what we need to ask

Denis

[Bentley Ferrari]

Originally Posted by woodsdenis

I think the real question is why is it so difficult for software programmers to
adhere to the RTAS specs in the way that AIR do. Does it take to much time?
Are Air privy to some secret code that others aren't? Is it too complicated ? This surely is what we need to ask

Denis

Maybe, but I don't think it's a question of being too difficult. I think it's more one of simple dollars and sense. For many software makers, VST development will get you more customers and more money than RTAS development. The ROI (Return on Investment) for developing RTAS or TDM is not always as high as it is for VST or AU development, depending on your product. This may change, though, if Digidesign's objectives in designing PT8 prove successful. The new GUI and MIDI improvements should appeal to current Cubase/Logic/Sonar users. In other words, when PT is more widely accepted as a VI-friendly DAW software, and it's user base grows in that direction, more developers will port their plug-ins to RTAS. (IMO.)

[woodsdenis]

Originally Posted by Bentley Ferrari

Maybe, but I don't think it's a question of being too difficult. I think it's more one of simple dollars and sense. For many software makers, VST development will get you more customers and more money than RTAS development. The ROI (Return on Investment) for developing RTAS or TDM is not always as high as it is for VST or AU development, depending on your product. This may change, though, if Digidesign's objectives in designing PT8 prove successful. The new GUI and MIDI improvements should appeal to current Cubase/Logic/Sonar users. In other words, when PT is more widely accepted as a VI-friendly DAW software, and it's user base grows in that direction, more developers will port their plug-ins to RTAS. (IMO.)

I suspect this has a lot to do with it. Does anyone know the percentage of LE v Logic systems out there. I think this is were acceptance and adherence to
Rtas standards will happen.

Denis

[GlennO]

The use of wrappers is not the source of the problem and neither is RTAS. The reason so many plugins (not just Omnisphere) require the "reduce your RTAS processors by one" workaround is because of a recently discovered issue with PT8 on OSX 10.5. The problem does not occur in other hosts like Logic because ProTools uses processors in a different way than other hosts.

The reason that an update to either PT8 or the affected plugins which fixes the problem is not yet available is simply because it is a difficult problem to solve.

Glenn Olander
Spectrasonics

Originally Posted by woodsdenis

I know, for example, that Omnisphere works much better as an AU on Leopard. One of the main reasons for this I can only assume is that it has to use a VST wrapper for PT8, so does Stylus. You cant blame Digi for that. Spectrasonics view is that this is an acceptable way to do this and has worked well in the past.

http://www.northernsounds.com/forum/...hlight=wrapper

To be honest I cant argue with them I know nothing about coding, but is it really that difficult to code/convert for RTAS. If it is, thats what we should complain about. This is not about taking sides. or in this case taking Spectrasonics to task. I use them as an example as they are a major player in the VI market and code for every major format I know of.

[woodsdenis]

Originally Posted by GlennO

The use of wrappers is not the source of the problem and neither is RTAS. The reason so many plugins (not just Omnisphere) require the "reduce your RTAS processors by one" workaround is because of a recently discovered issue with PT8 on OSX 10.5. The problem does not occur in other hosts like Logic because ProTools uses processors in a different way than other hosts.

The reason that an update to either PT8 or the affected plugins which fixes the problem is not yet available is simply because it is a difficult problem to solve.

Glenn Olander
Spectrasonics

Glenn

Thanks for your swift reply to this topic. I think it is great to see you guys
respond like this.

I am not suggesting for one second that Omnisphere is the only culprit in this symptom. The only reason it was mentioned was, because it was used as an example earlier on in the thread and I happen to own it, so i have direct experience of how it works.

Can you answer 3 questions.

1.Why would you use a VST wrapper for Omnisphere on both Tiger/Leopard?
2.Why not a direct RTAS version.?
3.What exactly is the problem in your view that affects PT8 on OSX.

If you dont feel comfortable answering these questions, fair enough.

This is all slightly off topic and is not meant in anyway to turn into an anti
Spectrasonics rant. So please respect that Glenn has responded to this thread. I haven't seen any of the other software companies responding like this.

On a personal note I have used all Spectrasonics for years and never had any issue with them. Omnisphere is an awesome VI and I am only annoyed that along with many others there is an issue with PT8 and not with the
other configurations i have mentioned.


Denis

[GlennO]

Like all multiplatform plugin developers, we do not write a new plugin from scratch for each format. All such developers use a design which, of one form or another, puts a veneer over the plugin proper to make it usable in various formats. The wrapping approach we use has proven to be reliable for our customers, and as I mentioned, is unrelated to this problem.

As for the problem, I can only say that it is due to a complex interaction between PT, OSX10.5, and operations which some plugins commonly perform.

Glenn Olander
Spectrasonics

[Fidelis]

Originally Posted by GlennO

As for the problem, I can only say that it is due to a complex interaction between PT, OSX10.5, and operations which some plugins commonly perform.

Glenn Olander
Spectrasonics

That's what Digi should be brave enough to tell us here! There is a problem here! And WE (costumers) are not crazy.

Glen, thank you so much for your posts.

[PhilBuckle]

So at last we know the truth.
I guess it's easier for Digi to tell us to stay current and update our old computers, thereby throwing the problem over to us and taking some heat of themselves.

The truth is that as long as this "problem" with PT8 exists there won't be a mass migration of composers from Logic and other apps to the PT platform.

Unless being on a newer Mac solves this issue 100%

[c-tone]

Originally Posted by PhilBuckle

Unless being on a newer Mac solves this issue 100%

You still need to reduce the number of processors on a new Mac, so no, it doesn't solve the issue. Being on a newer Mac just makes the problem easier to handle, but you still are losing processing power.