Max/MSP on Mac-tels.
I have Max/MXP 4.5.4 and I’m looking to buy a spangley new Macbook Pro but before I part with the serious quantity of wedge required I want to know that Max is going to work on it through Rosetta, I’m expecting it to but I want to know if anyone knows for sure.
Also should I be getting Max/MSP sequential updates free, I know wouldn’t get version to version updates but thought I should get sequential onrs. Anyone know?
search the archives of some days/weeks ago, they will probably put out a beta as soon as possible, but for now you shouldn’t count on it.
Cheers for the reply, that’s what I’m getting from cycling 74 too, but it’s good to hear from people who perhaps know more than me. Also I was looking for replies from anyone who perhaps has a Mac-tel or has been running it on a development machine.
On Jan 21, 2006, at 9:39 PM, Joe Bird wrote:
> I have Max/MXP 4.5.4 and I’m looking to buy a spangley new Macbook
> Pro but before I part with the serious quantity of wedge required I
> want to know that Max is going to work on it through Rosetta, I’m
> expecting it to but I want to know if anyone knows for sure.
Max Runtime will work under Rosetta. The PACE protected full app will
not. Even still, for realtime audio work, Rosetta is not a good
option as it will have audio dropouts from time to time. Best bet is
to wait for a Universal Binary version which will be available later
this year (actively working on it, but no date promised, and we won’t
respond to further inquiries).
> Also should I be getting Max/MSP sequential updates free, I know
> wouldn’t get version to version updates but thought I should get
> sequential onrs. Anyone know?
4.5.x are free updates for all registered owners of MaxMSP 4.5.
Just an update to this info: PACE just released a new version of their
InterLok Extensions on paceap.com that fixes issues with Intel Macs. I
have verified that it works on a new Intel iMac (the laptops are not
out yet, of course, but I don’t know why it would be any different).
You wouldn’t know about this update from the PACE site, but it’s
definitely there when you download the Mac OS X Extensions. We haven’t
updated our 4.5.6 installer with these new extensions yet so you’ll
need to get the extensions directly from paceap.com. With this update,
you’ll be able to run 4.5.6 on an Intel Mac.
Our Intel Mac development is moving forward. We showed a preliminary
version this past weekend at NAMM with all but two or three objects
working and I am hoping to have more information about questions people
have raised here very soon. Tomorrow I hope to know the answer to this
question: can traditional CFM Max externs written for the PowerPC run
inside the native Intel version? I think the answer might be yes,
surprisingly, which would render inoperative my previous assertion that
all objects will need to be recompiled. Stay tuned.
On 25 Jan 2006, at 21:05, David Zicarelli wrote:
> With this update, you’ll be able to run 4.5.6 on an Intel Mac.
All totally Rosetta-ated, presumably?
> Tomorrow I hope to know the answer to this question: can
> traditional CFM Max externs written for the PowerPC run inside the
> native Intel version? I think the answer might be yes,
> surprisingly, which would render inoperative my previous assertion
> that all objects will need to be recompiled.
Oh. Apart from anything else, I had assumed that, with MacIntel,
Carbon was deaddeaddead.
nick rothwell — composition, systems, performance — http://
On Jan 25, 2006, at 2:32 PM, Nick Rothwell wrote:
> Oh. Apart from anything else, I had assumed that, with MacIntel,
> Carbon was deaddeaddead.
FWIW, note that Carbon is *not* CFM. It is just one of Apple’s C
APIs, and is independent of runtime architecture. Carbon exists both
as Mach-O and CFM. Also note that CFM programs run without problem
under Rosetta, which is why the CFM MaxMSP 4.5.6 works under Rosetta
(with the latest PACE extensions). Note that for real time audio
processing, Rosetta still presents problems–i.e. despite the fact
that the code executes, there may be audio dropouts resultant from
the code translation at runtime.
Thanx Joshua, that’s great news, I agree, I see no reason why it shouldn’t work on a Macbook since it’s just the same as an iMac in a smaller silver box.
Thanx again for all the patient replies and info.
Just curious, but with this advent, does it mean that externals and pluggo plugins and such will function on both platforms without a recompile, ie win and osx?
In a word – no.
Is it possible to make the upcoming intel MSP object vector-ready with SSE optimization ?
(i.e. fft~, buffir~, index~ etc..)
On around Jan 28, 2006, at 9:26, Salvator said something like:
> Is it possible to make the upcoming intel MSP object vector-ready with
> SSE optimization ?
> (i.e. fft~, buffir~, index~ etc..)
It is surely possible, will presumably need to be done on an
object-by-object basis, and from what I know about AlitVec, it is not
always immediately obvious which processes will benefit by vector
Vector processing is not the answer to every problem.
FFTs can benefit a lot from vector processing (as witnessed by Apple’s
FFT library). I’m not sure that index~ would benefit. There are others
who may have deeper insight into this.
————– http://www.bek.no/~pcastine/Litter/ ————–
Peter Castine | ^
| Litter Power & Litter Bundle for Jitter
firstname.lastname@example.org | iCE: Sequencing, Recording, and Interface Building
email@example.com | for Max/MSP
| Extremely cool
> Is it possible to make the upcoming intel MSP object vector-ready with SSE optimization ?
> (i.e. fft~, buffir~, index~ etc..)
Yes, the newest centrino chips just announced support SSE3
instructions, so at some point your patches may receive a nice
efficiency boost with upgrades of MSP and Jitter.
If possible, Apple advises that you use the Accelerate framework for
your mac optimization needs.
This will ensure a high degree of optimization on any current and
future platforms, and because it’s a dylib your code will benefit
from optimizations without the need for recompiling.
This of course is no substitute for hand-tweaking, but I for one am
too lazy to not use the given libs when it is pragmatic to do so.
Also, gcc-4.0 will automatically optimize some code for the lowest
common denominator if you pass the right flags in Xcode ( fmadd on
altivec for example ).
> If possible, Apple advises that you use the Accelerate framework for
> your mac optimization needs.
Thanks for the tip, but because we produce a cross-platform product,
and both platforms now support the same vector instructions, I think
we’ll probably steer clear of Apple-specific optimization libraries.
Sorry, I didn’t intend to aim that towards Cycling, but rather at any
external-writers who don’t want to learn sse or altivec assembly.
Hey David Z.,
I was just wondering if you now knew the answer to the question about traditional CFM externals for the PowerPC needing to be recompiled.
- David @ VV
CFM binaries cannot be loaded from a native-intel process. No-how, no-
way. You simply cannot run Rosetta-translation from inside of a non-
translated binary. Rosetta only works on a per-process level. This
is according to the people supporting CFM on OS X.
I figured this was the case, David Z’s last post had gotten me hopeful that he had found a work around of some sort.