Performance and Tiger/Leopard Slimmed

[MrPilot]

Hi!

Was doing some interesting tests, (powerbook 1.67 DLSDHR) and the result was quite surprising to say the least.

I used a high bitrate HD video clip and XBMC to test different setups and how many frames it skipped for the entire clip.

1. Tiger using XBMC gave me 1100 total frames dropped.
2. Leopard slimmed down with no gl.blocker, dashboard, spolight, 3Ddock etc etc gave me 740 frames dropped (nice increase vs tiger)
3. Leopard using spotlight, glim,blocker, 3D-dock with all Quartz effects enabled, dashboard enabled and indexing on gave me only 650 dropped frames!!

This is strange, tiger loads so quickly it seems so "lightweight" yet in-game performance and HD video playback was not as good as a more bloated OS like leopard is. I always thought that the more lightweight a system was, the more performance you got. I really think that Leopard didn't get all that much more PPC optimizations in comparison with Tiger. (Due to intel transition)

Anyhow... the weirdest part of the test was that the slimmed down Leopard yielded a slower HD video playback machine! I tried this test 2 times using terminal commands and "secrets" to verify that slimming down the OS HURT performance... WEIRD!! Comments? Experiences with this?

 

[eyoungren]

Hi!

Was doing some interesting tests, (powerbook 1.67 DLSDHR) and the result was quite surprising to say the least.

I used a high bitrate HD video clip and XBMC to test different setups and how many frames it skipped for the entire clip.

1. Tiger using XBMC gave me 1100 total frames dropped.
2. Leopard slimmed down with no gl.blocker, dashboard, spolight, 3Ddock etc etc gave me 740 frames dropped (nice increase vs tiger)
3. Leopard using spotlight, glim,blocker, 3D-dock with all Quartz effects enabled, dashboard enabled and indexing on gave me only 650 dropped frames!!

This is strange, tiger loads so quickly it seems so "lightweight" yet in-game performance and HD video playback was not as good as a more bloated OS like leopard is. I always thought that the more lightweight a system was, the more performance you got. I really think that Leopard didn't get all that much more PPC optimizations in comparison with Tiger. (Due to intel transition)

Anyhow... the weirdest part of the test was that the slimmed down Leopard yielded a slower HD video playback machine! I tried this test 2 times using terminal commands and "secrets" to verify that slimming down the OS HURT performance... WEIRD!! Comments? Experiences with this?

Leopard seems to get an unfair shake a lot. But I've had nothing but good experiences with it. It's been far more stable for me than Tiger.

 

[Andropov]

Hi!

Was doing some interesting tests, (powerbook 1.67 DLSDHR) and the result was quite surprising to say the least.

I used a high bitrate HD video clip and XBMC to test different setups and how many frames it skipped for the entire clip.

1. Tiger using XBMC gave me 1100 total frames dropped.
2. Leopard slimmed down with no gl.blocker, dashboard, spolight, 3Ddock etc etc gave me 740 frames dropped (nice increase vs tiger)
3. Leopard using spotlight, glim,blocker, 3D-dock with all Quartz effects enabled, dashboard enabled and indexing on gave me only 650 dropped frames!!

This is strange, tiger loads so quickly it seems so "lightweight" yet in-game performance and HD video playback was not as good as a more bloated OS like leopard is. I always thought that the more lightweight a system was, the more performance you got. I really think that Leopard didn't get all that much more PPC optimizations in comparison with Tiger. (Due to intel transition)

Anyhow... the weirdest part of the test was that the slimmed down Leopard yielded a slower HD video playback machine! I tried this test 2 times using terminal commands and "secrets" to verify that slimming down the OS HURT performance... WEIRD!! Comments? Experiences with this?

Did you disable QuartzGL when slimming Leopard down? Maybe XBMC makes some kind of GPU acceleration that is QuartzGL dependent. Also, slimming down a OS is great for a lot of tasks, but Dashboard, Spotlight...etc. only use RAM, and video players don't really need a lot of RAM to run.

 

[wobegong]

Leopard seems to get an unfair shake a lot. But I've had nothing but good experiences with it. It's been far more stable for me than Tiger.

Agreed, dare I say it when I had my MacMini G4 1.42 I remember thinking Leopard was faster when I upgraded it - I'm sure Leopard has optimisations which maybe can only be taken advantage of if you have a processor up to a certain spec.

----------

Maybe XBMC makes some kind of GPU acceleration that is QuartzGL dependent.

Yes it does, XBMC performance is very much linked to your GPU.

 

[Member2010]

leopard

Is your Leopard installation from the ultimate 10.5.8 DMG image that has all the Intel bloat stuff removed or did you just install from the stock 10.5 disc?

::

 

[MrPilot]

Is your Leopard installation from the ultimate 10.5.8 DMG image that has all the Intel bloat stuff removed or did you just install from the stock 10.5 disc?

::

Using the 10.5.8 PPC only superduper with all the bloated stuff removed.

However that shouldn't do anything on performance...

----------

Leopard seems to get an unfair shake a lot. But I've had nothing but good experiences with it. It's been far more stable for me than Tiger.

Yeah, seems great! somewhat slower to load but yeah, once it's up and running, it's great!

----------

Did you disable QuartzGL when slimming Leopard down? Maybe XBMC makes some kind of GPU acceleration that is QuartzGL dependent. Also, slimming down a OS is great for a lot of tasks, but Dashboard, Spotlight...etc. only use RAM, and video players don't really need a lot of RAM to run.

QuartzGL was left on as I felt it would hurt general video performance otherwise. I believe that everything that uses threads and ram does to some extent slow the system down... though you are mostly correct, specially on the highest spec powerbook DLSD/HR or MDD G4's/G5's but slower and older PPC macs will hurt by having dashboard and spotlight etc on.
Oh btw, XBMC doesn't use the GPU at all :/

----------

Yes it does, XBMC performance is very much linked to your GPU.

Not at all, I'm afraid

 

[wobegong]

--------

[/COLOR]

Not at all, I'm afraid

— on Mac OS X 10.6 or later, XBMC uses Apple's Video Decode Acceleration Decoder (VDADecoder) hardware acceleration instead, and on recent Windows systems it uses Microsoft's DirectX Video Acceleration (DxVA). Linux boxes can also use OpenMAX for hardware video acceleration, which is most useful for systems built on Nvidia's Tegra2 platform.

The user interface itself uses OpenGL, OpenGL ES, or EGL, so it, too, can be hardware accelerated. Use of GPU acceleration for both video decoding and the GUI reduces XBMC's CPU requirements considerably, and 11.0 officially introduces support for several more low-resource systems. On the Linux side, this includes Texas Instruments OMAP4 processors. On the Apple side, it includes better support for recent iOS 4.x devices (including recent iPads and AppleTVs). But for those who still rely on their CPUs, regardless of the platform, XBMC can now detect CPU features like SSE and MMX at runtime."

You can google Apple VDADecode but its GPU .264 decoding...... Shame this is only for 10.6/Intel!! My bad
....

 

[Swampus]

You can google Apple VDADecode but its GPU .264 decoding...... Shame this is only for 10.6/Intel!! My bad
....

Are you saying now that you were mistaken when you earlier stated that there was GPU acceleration for XBMC on PowerPC Macs? I've not even tried it yet, but I've been curious about it.

Even CorePlayer doesn't do that, does it? It works well just because it makes good use of Altivec? I don't recall seeing any difference in frame rates or CPU usage between low and high end video cards while watching movies.

It would be interesting if XBMC did things differently.

 

[wobegong]

Are you saying now that you were mistaken when you earlier stated that there was GPU acceleration for XBMC on PowerPC Macs? I've not even tried it yet, but I've been curious about it.

Even CorePlayer doesn't do that, does it? It works well just because it makes good use of Altivec? I don't recall seeing any difference in frame rates or CPU usage between low and high end video cards while watching movies.

It would be interesting if XBMC did things differently.

Yes looks like I was wrong it only uses the GPU for menu transitions (OpenGL) for pre 10.5 (10.6 and above for the above GPU hardware assistance).

Still for me on a G5 XBMC plays high def videos smoother than anything else I have tried (I never used CorePlayer so cannot compare to that).