Geekbech completely unreliable tr4sh?

[bemoll4-]

CL vs Mhz
QPI vs CoreSpeed
Turbo vs No Turbo

The Logic 9 benchmark... well its not that good the question is when the barrier gets reached... when you have playback issues once or regular only the first time but then indefinetly its good etc.

apologies but I'm not sure I understand.. why the evans logic test is not a good evaluation test for comparing your performance?

In my experience it has been very consistent across the board... a better specced (or optimized as you say) osx machine does more tracks

 

[damndoe]

1. The problem is "when" to measure. Even for yourself: when it stops once? is it the max? when you replay it multiple times and it does stop once but then you play it again and it plays indefinitely - does it count?
2. The problem is people use different settings - machine vs machine comparison hard

sure the test is ok. but its not a good benchmark. its just good for testing how well logic runs ;-)

 

[bemoll4-]

1. The problem is "when" to measure. Even for yourself: when it stops once? is it the max? when you replay it multiple times and it does stop once but then you play it again and it plays indefinitely - does it count?
2. The problem is people use different settings - machine vs machine comparison hard

sure the test is ok. but its not a good benchmark. its just good for testing how well logic runs ;-)

i see but the instructions on even's test are quite clear about the points you raised.

i agree it doesn't give you a super accurate reading, but if you do intensive audio work and want to know how well your machine is performing i still found it a valuable and consistent tool. it gives me a better idea than geekbench or the likes.. at least works for me.

 

[damndoe]

Well I did not find things to be precise enough.
But furthermore Logic is not the only DAW and it only measures your system running Logic not your system running lets say heavy recording or for another case heavy sample use etc. Its just some application specific benchmark, like running WinRAR (which actually shows huge benefits for highly clocked RAM even if the CAS latency is skyrocketed).

So I am still looking for something like Cinebench but more broad. And yeah, even Cinebench seems to be a better measurement for me.

 

[dejongj]

I just see these 'tools' as relative comparison tools, not as exact science universal tools. So in the OP's case, if he always tests when Ableton Live is running then the condition stays exactly the same.

Key to any test, ensure you environment and test conditions are the same without it, it is worth jack.

 

[justr]

I just see these 'tools' as relative comparison tools, not as exact science universal tools. So in the OP's case, if he always tests when Ableton Live is running then the condition stays exactly the same.

Key to any test, ensure you environment and test conditions are the same without it, it is worth jack.

I agree.

They are ballpark figures, mainly, and best for relative testing.

I'll use Geekbench to compare some of my systems (along with running other benching apps to look for consistency), but only in a very blunt way-- If the score is less than 20% different, I don't put much faith in it. And that said, I put a lot more faith in comparing two systems when the scores differ by something like 300% (for instance, comparing an old core duo macbook to a monster hackintosh).

The most important benchmarks, though, are daily use scenarios.

Do you edit video and lose time to rendering? Well... how much faster/slower are your systems at doing comparable tasks. Compile code... same.