LIVE 4 PERFORMANCE TEST

[ultrasource]

Sucks for apple that long time fans like myself will be building amd towers so they can run their favorite apps because macs just don't cut it.
they (apple) gotta get their heads out of their asses and forget that not all mac owners are swigging down cups of latte as they drive to their 6 figure job in their bmw.
my powerbook, honestly, does everything i need it to for now, but there really is no headroom to experiment. as soon as i need that headroom, i'm taking my business elsewhere.
*i'm not trying to fuel the pc vs. mac debate. that's not productive.

[D K]

1)50-54%(*36-40%)
2)12" powerbook
3)osx.3.5
4)g4 1.33
5)768 mb
6)internal
7)5400

bet all the live staff use pc's!
i wonder what the results would be like if apple had 2 or 3 gig processors for lappys (i can probably contemplate that one for a long,long time)
*i noticed a whopping 14% difference in cpu when i switched off the reverb
dk

[cope]

1) Ableton Live CPU Meter = 28-31%
2) Laptop/Desktop? = Laptop, HP nx7010
3) Operating System = xp home sp2
4) CPU Make Model and Speed = Pentium-m BANIAS @ 1.6ghz
5) Amount of Ram / Speed of Ram = 512 MB
6) Soundcard = SoundMAX internal
7) Hard Disk Drive Speed = 4200RPM

[peeddrroo]

1) 44%
2)custom desktop (ASUS P4B533 motherboard)
3)WinXP Pro
4)P4 @ 2Ghz
5)1GB DDR133
6)M-Audio Delta audiophile 24/96 PCI
7)7200rpm

thought my desktop was performing very decently, but it's good to know there's room for improvement! (specially as i'm hitting max cpu quite often).

[Spacerboy]

peeddrroo

try this: http://www.musicxp.net/tuning_tips.php

carefully go through every single step!
its really worth it.

we are looking forward which change in cpu usage
it made to your pc.

lg spacy

[Amberience]

AMD fuckin' kicks ass

[unearthed123]

- 1 -

1 - 47% - 59%
2 - SONY VAIO PCG-Z1V/P
3- Windows XP Pro SP1
4- Pentium M 1.6Ghz(Banias) Intel 855PM
5- 512MB x 2 PC2100 DDR-SDRAM
6- Echo Indigo 2
7- 4200rpm 60GB(TOSHIBA MK6021GAS)

- 2 -

1 - 56% - 58%
2 - Apple Powerbook G4 12inch 1.33Ghz
3- MAC OSX 10.3.5
4- PowerPC G4 1.33Ghz
5- 512MB+256MB PC2100 DDR-SDRAM
6- MOTU 828MK2
7- 5400rpm 80GB

- 3 -

1 - 22% - 25%
2 - Custom Build
3- Windows XP Pro SP1
4- ATHLON XP3000+ nVIDIA nForce2(EPOX 8RDA+)
5- 512MB x 2 PC2700 DDR-SDRAM
6- RME Hammerfall Multiface PCI
7- 7200rpm 80GB 8MB CACHE MAXTOR 6Y080P0(System)/6Y160P0(Data)

Looking this thread,
I think
New Pentium M(Dothan 2MB L2 Cache) is best laptop for Live?
Very good results!

Does anyone test the DUAL OPTERON Machine performance?
(Operating CUBASE SX2,DUAL OPTERON(248x2) is 200% faster than P4 3.2Ghz)

I planned to switch ATHLON XP3000+ to ATHLON64 3500+(socket939).
How about this change?
Can I feel the performance up on LIVE4 & SX2?

[rand]

Well, to run the test I didn't want to close my current applications bacause they are HUGE and they would require too long right now to reload... (I'm running LogicPro with Space Designers, softsynths & BIG EXS streaming samples and such... I'm ALSO running Live4 & it's doc is 19.3 megs... crazy. )
So I duped my LIVE ap and ran it too. ;=)
voila 'for now', I'll update it latter with only Live4 running...
1=peak 20-22
2=PowerMac G5 2x2.5,
3=OSX 3.5
4=PowerPC G5 970fx, 2x2.4
5=4gigs PC3200 DDR ram
6=motherboad stereo out (waiting for RME to fix my hammerfall 9652)
7=HDs are 512mb Sata (with the Live files now), 150 scsi raid, 100ata...firewire here n there.
(Remember I'm running other aps simultaneously. )
rrRand

[peeddrroo]

peeddrroo

try this: http://www.musicxp.net/tuning_tips.php

carefully go through every single step!
its really worth it.

we are looking forward which change in cpu usage
it made to your pc.

lg spacy

thanx spacy. my pc was already set up as explained on musicxp... the swap memory wasn't correct, i've disabled it, but the performance stays the same...
anyway, from what i've seen from other test results, 2GHz processors seem to be around 40%. but these tests are reeeaaally interesting, especially as i'm thinking to buy a laptop.

[Spacerboy]

___________________________________________________________
thanx spacy. my pc was already set up as explained on musicxp... the swap memory wasn't correct, i've disabled it, but the performance stays the same...
anyway, from what i've seen from other test results, 2GHz processors seem to be around 40%. but these tests are reeeaaally interesting, especially as i'm thinking to buy a laptop.
___________________________________________________________

peeddrroo,

i would go for a athlon laptop then. cheaper and faster...
but wait for more results on these cpu´s in laptops...
centrino seems to be terrible good too....

spacy

[peeddrroo]

peeddrroo,

i would go for a athlon laptop then. cheaper and faster...
but wait for more results on these cpu´s in laptops...
centrino seems to be terrible good too....

spacy

agreed, the results for athlons look really amazing.
though i had a really bad experience with a Duron and its VIA chipset: i had horrible clicks on any audio that was recorded (that was with the same soundcard). so i was like 'never again' with AMD.
but these test results might make me change my mind...

[Fernando Carvalho]

1) 25%-28%
2) Toshiba M30
3) Windows XP Professional
4) Pentium M 1.5ghz, 1MB L2
5) 1 GB Ram
6) M-Audio Audiophile Fireware ext-card
7) 7200rpm ext Lacie usb2 HD

[raapie]

peeddrroo

try this: http://www.musicxp.net/tuning_tips.php

carefully go through every single step!
its really worth it.

we are looking forward which change in cpu usage
it made to your pc.

lg spacy

are you sure? I tried all those tweaks and it doesn't made any difference on my machines.

[pentajigga]

1) 36%
2) dell desktop
3) XP
4) P4 2.4ghz
5) 512MB 233 (i think)
6) RME HDSP 9652
7) 7200RPM

couldnt set no input. because RME ASIO does not have seperate section in live for input/output, only has RME ASIO or NO AUDIO. so latency was 11+11=22ms.

[AdamJay]

Okay Guys - thanks to everyone who contributed to this public benchmark

Feel free to keep testing machines and configurations, the more the better.
But i did want to make a couple comments based on the results so far.

I think by looking at how productive and effecient a large variety of machines are with this Application, we can gauge a few things....

1) If you are buying a computer for the sole purpose of using Ableton Live 4, your money would be best spent buying either an AMD Athlon 64 Processor, or an Intel Penium-M Dothan Processor. (all assuming you have no major qualms using the windows operating system) If Ableton Live 4 is not the sole purpose for your purchasing/trading/upgrading a computer, then you may want to weigh in the other uses for the machine and how they perform as well.

2) Pure speculation, but... Perhaps Ableton could focus their attention more to optimizing for the G4 and G5 processors. While its good that they cleaned up their code for Live 4 without using the Altivec engine (this gave benefits to G3 owners as well), it simply isn't fair for the Mac folks. Multi-Threading support on the Audio Engine for Dual G5 owners is a must. I still think altivec would help, it did so much for Reason.

3) If you are planning on buying a G5 iMac for Live, now you have a better idea how well it would perform. If a dual 2ghz G5 is doing 24-26%, with its 1ghz FSB's, i can't imagine the single 1.8ghz G5 iMac with 600mhz FSB doing much better than 30%. Also, if and when the Powerbook goes G5, you could expect it to be equipped similarly to the current G5 iMac - something you should take into consideration if you've been "holding out" for a G5 Powerbook which has still yet to be announced and is still only a speculation.

4) Windows XP SP2 may[\i] give better performance. i was amazed to see spacerboy's and amberience's results with Athlon XP 2400 and 2600 processors under Windows XP SP2. I myself am still weary of updating to SP2 - but a minor performance boost should give us all more incentive. Thanks to Machinate and Rikhyray for testing SP1 vs SP2 on the same machines!

5) Mac users - do not be discouraged. This test simply showed how the new features in Live 4 effected performance on various machines. One of the Main culprits to poor performance on Macintosh computers in Ableton Live is Ableton's Reverb. The Test had the reverb on "Comfort" Setting. When i disengaged the reverb on a G4 800mhz iBook, the cpu meter dropped from 89% to 65%. The beauty of DAW's is the effects are totally interchangable, especially on the Mac side with Audio Units. Apple's own AU FX are quite customizable, AU Reverb has multiple transparency settings that you can tune to your own machine based on how many resources it is taking. As many have said - Logic is a different story, and if Apple owned Ableton like they do Emagic - i'm sure the results would be much different. Don't let this test discourage you, but please let it inform you.

and finally..
6) Its obvious that alot of people are interested in performance. As ultrasource mentioned, Headroom to experiment is something to take into consideration. I myself, like to do my final arrangements in realtime - recording direct to .wav with the Tape IT VST recorder, so i need the headroom to be able to do things in realtime, without glitches, at a reasonable latency. If i used Arranger for my final arrangment, yes i could render that to disk with a machine half as powerful as my current machine. For the nay-sayers that say benchmarking is useless and to just be creative and not worry about it - you simply cannot have that attitude using a software that evidently[\i] performs vastly different on a variety of computers. We all want to get a good value for our money, and we all want to know our options.

okay enough of that....

Thanks again to everyone who participated.