3GB RAM in a macbook

[poodlebuild]

Does anyone here have experience with 3GB of RAM in a macbook using PT with VI? There seems to be some controversy over whether 3GB is better than 2GB since 3GB means using 2 unequal size sticks and losing the dual channel thing.
There are some bench test results here:
http://eshop.macsales.com/Reviews/Fr...kcore23gb.html

This is OWC, they want to sell more RAM, and yet the numbers don't seem that impressive.
Still, I'd like to know how this applies to VI in PT.

[spkguitar]

I just got new RAM from OWC last week, and now I have 2.5GB installed. I haven't had a chance to do any testing or benchmarking, though.

Anything specific you would like to know? Any specific tests you want to see the results of? I'll crank up Ivory on it this evening and let you know how it goes.

[poodlebuild]

Thanks spkguitar,

So I guess now you must have 1 x 2GB and 1 x 512.
I'm interested in the difference between two same-sized RAM modules and two different sizes as it applies to PT, esp how many VIs you can run at once. I guess before you had two 512 sticks. I don't know if you did any actual tests before to compare with now, but even so if that's what you had, what you have now is so much more RAM that it's going to be a huge difference.
What would be great is to hear from someone who's tested 2 x 1GB vs. 1 x 2GB + 1 x 1GB, since that's the choice I now have to make. Is it worth the extra money or could it actually be worse with more RAM that's uneven? I've seen the OWC bench tests but I can't tell how it would apply to Pro Tools.
Theoretical replies are certainly welcome as well...

[spkguitar]

So I guess now you must have 1 x 2GB and 1 x 512.

Yes.

I guess before you had two 512 sticks.

Yes.

What would be great is to hear from someone who's tested 2 x 1GB vs. 1 x 2GB + 1 x 1GB, since that's the choice I now have to make.

I would imagine that would be quite rare, as it would be an expensive test to run.

So far everything is running very smoothly for me and I have no complaints at all. Since there really isn't any benchmarking test for using VI (similar to the DaveC and DVerb tests... or is there?), it's hard to judge.

Maybe we should come up with a benchmarking test? Using XPand? (most everyone has it now)

[cheekypaul]

I have 3gb. Can I test something for you?

[Stiff]

Problem with using Xpand! as the benchmark test would be that it's so easy we wouldn't be able to run enough instances with track limitations and all!

[sunburst79]

Maybe we should come up with a benchmarking test? Using XPand? (most everyone has it now)

I was on a road trip pondering Protools/RAM and VI's and I was thinking exactly the same thing.

Some type of VI benchmarking IS needed and I was also thinking Xpand and would be the Vi to use.

It's Free and everybody should have it by now

Some thoughts.

The beauty of a Dverb test is that it can be duplicated on any current PT rig and the results are consistent and duplicatable.

A VI or Xpand benchmark test could be equally useful.

We would need a downloadable MIDI sequence or session file to play in Protools.

This would be to keep things consistent

Multiple instances of Xpand running different voices as to stress the CPU, I/O system and RAM.

Ideally we might want several versions of the same test in order to stress the three subsystems separately.

What I don't know is does Xpand stream from disc or just load into RAM?

At any rate a duplicatable VI test is a excellent idea and I would hope that between all of us DUC'ies we could come up with something.

My knowledge of MIDI is not strong enough to do this by myself but I think its a great idea!

Anyone else think that this is worth starting its own thread in the Virtual Instrument category and seeing if we can come up with something?

[cheekypaul]

I was on a road trip pondering Protools/RAM and VI's and I was thinking exactly the same thing.

Some type of VI benchmarking IS needed and I was also thinking Xpand and would be the Vi to use.

It's Free and everybody should have it by now

Some thoughts.

The beauty of a Dverb test is that it can be duplicated on any current PT rig and the results are consistent and duplicatable.

A VI or Xpand benchmark test could be equally useful.

We would need a downloadable MIDI sequence or session file to play in Protools.

This would be to keep things consistent

Multiple instances of Xpand running different voices as to stress the CPU, I/O system and RAM.

Ideally we might want several versions of the same test in order to stress the three subsystems separately.

What I don't know is does Xpand stream from disc or just load into RAM?

At any rate a duplicatable VI test is a excellent idea and I would hope that between all of us DUC'ies we could come up with something.

My knowledge of MIDI is not strong enough to do this by myself but I think its a great idea!

Anyone else think that this is worth starting its own thread in the Virtual Instrument category and seeing if we can come up with something?

with this post we are getting somewhere. excellent.

[poodlebuild]

I have 3gb. Can I test something for you?

Well, I should say you've done enough. Your Pianoteq tip changed my life.
But, maybe you could. Since it would take someone who maybe worked where they had access to ram(or just lot's of money)to test 2 x 1GB vs. 1 x 2GB + 1 X 1GB, maybe this standard testing idea could be done by one person who has one of those configurations and someone else who has the other.
Xpand would be good. Could you really get so many instances that you'd run out of tracks to put it on?
To test RAM you would need something that didn't eat up your CPU first...

[Stiff]

Could you really get so many instances that you'd run out of tracks to put it on?

I wasn't THAT serious, but when I write songs I can have around 7-8 instances of it without noticing any problems at all.

As for the test:

* We should use a preset. Possibly a "heavy" one?

* A midi sequence was a good idea. How many simultaneous notes would be best?

* Mono or stereo?