virtualbox.org

Hi folks,

I run Virtualbox on a Mac OS 10.7 host with a Debian Wheezy guest, in full-screen mode. Sometimes, it's on a large monitor (2560x1600), and other times it's on a smaller one (1920x1080, I think). When on the smaller monitor, display performance seems snappier (e.g., moving windows is more responsive, and things like that).

The former is connected with a DisplayPort cable, while the latter is connected with a DisplayPort->DVI adapter. I've allocated 128MB of video memory, and 3D acceleration is on. 2D acceleration is off, and VB displays the "suboptimal settings" warning if I turn it on (and then turns it off for me when I start the VM, I think).

My questions:

1. Is there anything I can do to improve performance in the large-monitor case?
2. Should I file a bug?

Thanks!

Reid

reidpr wrote:2. Should I file a bug?

What would the bug be? That it takes longer to do more?

The bug would be that performance degrades when the display size is large.

reidpr wrote:The bug would be that performance degrades when the display size is large.

That's normally called "a fact of life". When there are lots more pixels to push around, performance goes down. If you have some magic recipe for doing more work take less time, please share the details.

Note that you're talking 2 megapixels vs. 4 here.

michaln wrote:That's normally called "a fact of life". When there are lots more pixels to push around, performance goes down.

Wait, what? I have never experienced any performance effect related to screen size alone. Obviously if you're trying to e.g. play a game fullscreen at 2560x1600 vs 1920x1080, it's going to be slower in the former case. But I'm not doing that. I'm moving the same 100x100 pixel window around on two different screens, and it's less responsive if the screen is larger. Same number of pixels to push in both cases.

Certainly, the Mac OS host shows none of these effects.

If I'm missing something, I'd love an explanation. Perhaps you could not assume I'm an idiot before replying.

reidpr wrote:If I'm missing something, I'd love an explanation.

I don't know if you're missing anything. Clearly we are.

If the guest is in both cases running with the exact same resolution, it would have been extremely helpful to have stated that in your first post.

If the guest resolution changes, then it's clearly not an apples to apples comparison. If the color depth changes, ditto. Even if the resolution doesn't change, the scanlines on the host are bigger which is highly unlikely to improve performance.