Live5 vs Live4 perf test

adhmzaiusz · Post by **adhmzaiusz** » Tue Jul 26, 2005 4:00 pm

stuffe wrote:Would it be worth doing a poll along the lines of

> 20% worse
> 10% worse
> 5% worse
0-5% worse
0-5% better
>5% better
etc etc

Just reading all the posts makes you lose track.

so everyone has to figure the % out on their own compared to how live 4 was for them? nah the point of this is to see the differences between different machines and configurations of each. it makes perfect sense to do it the way it is.

stuffe · Post by **stuffe** » Tue Jul 26, 2005 4:31 pm

ernene wrote:Live 4: 65%-71%
Live 5: 63%-68%

It's not hard to work it out, as per this previous post = 0-5% better on 5.

The way people are doing it now doesn't offer any real performance differential information between different machines and config, as we have no idea what other peoples machines are really like. OK you can compare total RAM, CPU type and speed etc, but you can't say that any 2 laptops running XP with the same total RAM and identical CPU and soundcard will be the same, because you don't know anything else, including the speed of the RAM, XP patch level, Driver or firmware revisions for all hardware, what's running in the tray or background, whether you have email open on a 2nd screen, how defragmented the hard disk is, whether antivirus software is running, laptop or desktop, running on battery or not, connected to a network etc etc ad infinitum....

At least this way you can say that like for like exactly, Live 5 is x% slower/faster than 4. Given enough results in this manner you might be able to gauge whether your machine could handle an upgrade, which surely is pretty much the only reason other than curiosity for doing the test?

Could be wrong, but I know that my mates (identical) laptop is slower than mine, but our descriptions would read the same...

Just a thought anyway.

subterFUSE · Post by **subterFUSE** » Thu Jul 28, 2005 3:48 am

Live 5 Performance:

25% stable, peak 27%
M-Tech D900T laptop
Windows XP Pro SP2
Pentium 4 HT 3.4 GHz (600 series) 2 MB L2 cache
2 x 1024 = 2048 DDR2 PC4200 dual channel 533 MHz RAM
2 x 80 = 160 gig 5,400 rpm SATA hard drives (RAID 0)
M-Audio FW Audiophile

DJSK · Post by **DJSK** » Thu Jul 28, 2005 12:51 pm

I got 20-21% with Live 4.14 and 19-22% with Live 5.01. No real difference here.

treysmith · Post by **treysmith** » Thu Jul 28, 2005 4:42 pm

treysmith wrote:20-21%

This was done at my normatl settings, just a quick test to see where I was at.

Sony Vaio A690
Tascam FW 1884
1.8ghz Centrino
1Gig PC3200
100HDD @ 5400

Actually at 18-20 after deleting config file

ernene · Post by **ernene** » Thu Jul 28, 2005 5:08 pm

adhmzaiusz wrote:
stuffe wrote:Would it be worth doing a poll along the lines of

> 20% worse
> 10% worse
> 5% worse
0-5% worse
0-5% better
>5% better
etc etc

Just reading all the posts makes you lose track.
so everyone has to figure the % out on their own compared to how live 4 was for them? nah the point of this is to see the differences between different machines and configurations of each. it makes perfect sense to do it the way it is.

Well, No. The title says : Live 4 vs Live 5 perf. test. What you are talking about is what Adam Jay wants to do: an only Live 5 performance test, to see how it runs on different machines and configurations, using version 5 devices... This test was made for v4... so it's ok if you want to compare versions, and that's it. Let's wait for AJ to give us a new performance test soon

ztutz · Post by **ztutz** » Sat Jul 30, 2005 6:08 pm

1) Ableton Live CPU Meter %: 48%
2) Laptop/Desktop? Make/Model : HP Pavilion zv6170
3) Operating System: winxp pro sp2
4) CPU Make Model and Speed: AMD Athlon 3800+
5) Amount of Ram / Speed of Ram (if you know): 1G
6) Soundcard (Stock or add on?, usb/firewire/pci ?) Novation X-Station 61 USB
7) Hard Disk Drive Speed (if you know): Seagate 4200/100G

Interestingly, when I reconfigured to use dual monitors, the CPU usage actually dropped a couple of percentage points to 46%. This is a hybrid shared/dedicated graphics card - somehow using two monitors kicked it into a more memory efficient spot.

zekrab · Post by **zekrab** » Sun Jul 31, 2005 8:48 am

ztutz wrote:1) Ableton Live CPU Meter %: 48%
4) CPU Make Model and Speed: AMD Athlon 3800+

ztutz · Post by **ztutz** » Sun Jul 31, 2005 11:06 pm

zekrab wrote:
ztutz wrote:1) Ableton Live CPU Meter %: 48%
4) CPU Make Model and Speed: AMD Athlon 3800+

I know!

The way I look at it, this notebook is no worse than a Mac (ducking and running away!)

Seriously, I've tried tuning XP, and have basically got similar results for my X-Station, for ASIO4ALL, and even for the WDM drivers (although they cannot get low latency - 512 is the best I can do there...) It is just a slow box, because of the graphics architecture, I am guessing. Jacking up the RAM to 2G might help - I'll wait for the 2G sticks to come down in price.

Anybody got suggestions?

zekrab · Post by **zekrab** » Mon Aug 01, 2005 9:46 am

Did you try with delay compensation off in the options?

rasputin · Post by **rasputin** » Mon Aug 01, 2005 6:53 pm

I opened up the Live 4 performance test set in Live 5 on my 2 GHz P4 system. The CPU meter stayed within the 41-43% range. Then I ran it again under Live 4 and was amazed to see the exact same usage! However if you look at Task Manager, the CPU usage it reports for Live 5 is higher, as is the amount of memory used.

This is using an Echo MIA with either the ASIO4ALL 2.6 or the Echo ASIO WDM 6.2 driver, with 512 sample buffers.

So basically I'm thrilled that I get all the new features of 5 witout any particular CPU hit! This is important as my CPU is not the fastest there is out there.

This system has 768 MB of PC133 RAM.

Live5 vs Live4 perf test

Here's some 4 vs 5 info on a 2GHz P4