The Logic Off Topic list

--- In logic-ot@y..., TazmnianDv@a... wrote:
> >But ... guess what graphics hardware Apple decided to build into 
the 
> >G4 range? 
> ><grin>  
> >Yep, it's ATI again!
> 
> Apple uses the ATI card only for the lowest end G4. All the rest 
use NVIDIA GeForce4 MX (I think for about two years now?).  

Well, when I was wandering around London's Tottenham Court Roard in 
early July and scrutinising the point-of-sale stuff, _all_ the G4's I 
saw were supposed to have Ati hardware onboard. I see that the 
website info for the very newest G4 (which wasn't in the shops then) 
says that initially the new 2*1Gigs will only be shipping with 
nVidia. The laptops apparently still have Mobility Radeons. The 
previous G4s apparently shipped with a mix of nVidias and Ati's. The 
wording for the latest G4 suggests that it /might/ ship later with 
other graphics boards, so perhaps the Radeon might show up again on 
the new G4s. 

> ATI is still making boards 
> for Mac (and PC), and recently introduced the 8500 high end 
> graphics board. I 
> hope that doesn't wipe the >grin< off your face.

Oh, I'm not disputing that the Ati Radeons are powerful, I was pretty 
damned impressed by what the cheapy Radeon All-in-One could do on the 
video capture front when I plugged one into my PC last year and 
realised that I was encoding at higher bandwidth than some broadcast 
places do <!> ... I was just saying that if one was (justifiably) 
criticising PC hardware for being built on third-party components 
that couldn't be guaranteed to work nicely with each other, then 
graphics hardware is notorious for this, and the one graphics 
hardware component that I have personal experience of being a bit 
flaky is the Ati Radeon series ... which also turns up on Macs. OK, 
so for all I know the fault might well be more with Ati's drivers or 
installation routines than with the actual hardware, but these are 
all things that count towards the instability or "delicateness" of a 
system.

With graphics hardware there's often a tradeoff, you might put up 
with a bit of instability to get more power.
Then again, if you intend to use the machine primarily for audio, you 
might prefer to forgo having the latest screaming graphics card in 
favour of something quieter and more stable. 
Personally I'd rather have a machine either optimised for video or 
optimised for audio, if you are mainly running logic, then you 
probably don't need MPEG2 video encoding or decoding hardware on your 
system.




> Now, I'll be the first to be suspicious of company created 
propoganda, but 
> Apple claims this on their website :
> 
> "We've taken the fastest PowerPC G4 processor ever created 
> and then doubled it"
... technically, one could argue that taking a single processor 
and "doubling it" is perhaps not quite the same thing as including 
two of them, but we'll pass over that ...

> "it's no wonder that it outperforms even the fastest Pentium-based 
systems."

The benchmarks I've seen have been rather different, there was one 
with the new G4 2*1Gig that had it running jobs at about two thirds 
the speed of a couple of selected pentium machines (one of which was 
a single processor jobbie that cost nearly as much as it did).
 
http://www.digitalvideoediting.com/2002/07_jul/features/cw_macvspc2.ht
m

A few things jump out at me from the Apple "datasheet"
1 - they don't give enough information for us to know what the tests 
actually involved
2 - the Dell PC model that they quoted as a reference seems to come 
in various configurations for memory speed, etc. For all I know it 
might have come in different configurations in the past. Did Apple 
test their box against an older 133 machine?
3 - While Apple say that they were checking against the fastest PC's 
available, they seem to have been using a model that Dell 
currently list as costing US$1069. By contrast, the Dell machine that 
allegedly ran the software about 1.5 times faster than the latest G4 
in the benchmark is supposed to cost about US$2875, which suggests 
that Dell probably recognise that its a better unit.
So at what point did Apple actually do these tests? Were they testing 
the G4 prototype against the fastest mainstream production PC that 
was available a year or two ago? 

PS, elsewhere, Apple's promotional material talks about their 
processor working "faster than light", which if they'd really managed 
it, _would_ be a major technical breakthrough! 
But somehow I'm inclined not to take that one too literally either 
Methnks the Apple site is somewhat prone to exaggeration.


 
> They further claim that Dual G4 is 300% faster than a Sony 
> VAIO 590G running a 2 GIG Pentium 4 (and the 733MHz G4 is 
> 84% faster) at DVD video encoding. 
> However, I think this is partly a gimmick, since Apple has 
> worked hard to make DVD conversion very efficient.

Yep, quite possibly. I think if someone really wanted to do DVD 
(MPEG2) encoding at silly speeds(ie faster than realtime), they'd 
tend to get themselves a graphics board that did it in hardware 
rather than have a top-range computer being tied up to do the same 
job in software. Or spend out on a faster third-party codec.

I expect that Apple had a lot of incentive to optimise their OS's 
MPEG2 handling because they presumably would have needed people to be 
able to watch DVDs on an iMac without any additional hardware. So if 
you are suggesting that perhaps macs are noticeably more efficient at 
MPEG2 handling in software than an equivalently-specified PC, then 
yes, I'd be inclined to believe it. 

Microsoft probably don't have as much incentive to buy in an 
optimised codec 
(in fact, Windows doesn't currently seem to ship with an MPEG2 
decoder at all. It does include a dedicated DVD player shell, but 
that (and Media Player) will only play DVDs after a third-party 
player like PowerDVD has been installed onto the system, and has 
provided the "missing" decoder code. )

> Are they guilty of misleading advertising? There are laws 
> against that. 

Depends on what their definition is of a "spec sheet". They might 
argue that they are using the phrase to signify a technical document 
that might be issued in advance of the actual product becoming 
available, that the details in it are therefore allowed to be out of 
date when the machine finally appears, and that a spec sheet 
isn't "advertising materials" but a technical document aimed at 
professionals who would know this and bear it in mind. 
I think that using the document like that on the website is 
misleading as hell, but there you go.


> Do you PC-lovers think is total nonsense? 
> If so, inform the Justice department of the United States 
> so they can clamp down on these falsehoods.

I'm not a "PC-lover", to me a computer is just an annoying box that 
you have to buy in order to be able to run nice software, like logic.
Until July 1 I was supremely uninterested in "Mac vs PC" issues, a 
Mac was a faster much prettier box that I couldn't afford and didn't 
actually need, and that's where the thought-processes stopped.
July 1st has changed all that for me and a lot of other WinLogic 
people, we now need to think seriously about whether we are likely to 
be thinking of migrating to Mac in a year's time in order to be able 
to use future versions of logic, because if we are, then our software 
and equipment purchases between now and then need to take that into 
account.
If we are still buying soft-instruments, then we'd need to check 
whether cross-platform upgrades are available, for hardware we'd need 
to check out the likelihood of OSX driver availability, and so on.

Hence our newfound interest in /exactly/ how many bangs you get for a 
buck with a Mac, and exactly how a PowerMac G4 compares with what we 
already have. It's not just about sticking two fingers up at the 
neighbours, a lot of us (post July 1) now have a genuine need to know 
exactly how these two platforms compare.

(and its probably driving the other forum users nuts <frown>)

Erk
PS:  someone's logic benchmarks:
  http://www.digitalnaturalsound.com/logic_dsp/perform_lap5.shtml