Oolite Bulletins

The reason I raise this, is that running Oolite on a Radeon HD 4730 (latest drivers from AMD),
even if I turn VSync on, I still get some tearing. Furthermore, using shift-F the frame-rate is
not stabilized to the monitor's vertical frequency.

On the other hand on an nVidia 8600GT the smooth scrolling is simply... awesome!
Furthermore, shift-F displays a stabilized frame-rate stabilized and equal to the monitors VSync value, as it should.

I wanted to check how many of the Linux Ooliters get advantage of this.

EDIT: I would also like to avoid ending up discussing which videocard is better.
The purpose is to see how many of the Linux Ooliters get advantage of this.

Getafix wrote:

Furthermore, using shift-F the frame-rate is not stabilized to the monitor's vertical frequency.

I'm not a Linux user, but I've had that problem with my Radeon HD 3650 all the time... I think that AMD's v-sync is not so good.

My old 128MB GF6600GT struggles to maintain Frame-rate at the best of times - but I'll let you know next time I remember to check!

El Viejo wrote:

... I think that AMD's v-sync is not so good.

Well, I've read in an Ubuntu Forums old post that AMD Catalyst Control Center does not actually wait for the vertical retrace.
It instead acts like a frame limiter trying. In my radeon configuration I see no difference with or without vsync on.

DaddyHoggy wrote:

...struggles to maintain Frame-rate at the best of times...

Perhaps, when there are not too many objects you could check if the frame-rate tends to be stabilized to the vsync value.

Conclusion:
Nobody cares about screen tearing issues...

I do! But I've not had time to fire up the old Ubuntu machine - I will post here when I do!

See what you've done now Getafix... you've got me tinkering with my Catalyst Control Centre settings... aaarrgh!

But I think Getafix is worried about Vsync under Ubuntu/Linux or have I confused myself?

No, you're right DH... I'm just trespassing in the Linux forum.
However, after some preliminary testing, VSync seems to make little difference on my Windows/ATI machine.

El Viejo wrote:

No, you're right DH... I'm just trespassing in the Linux forum.
However, after some preliminary testing, VSync seems to make little difference on my Windows/ATI machine.

We've had this at work with students when they do the benchmarking practical - make sure you turn Vsync off we say, only to discover that on the few machines we have with ATI cards (we made a conscious decision to move to nvidia with our last rolling upgrade) it doesn't make any difference! Students get marks for noticing, more marks for commenting (correctly), lose marks for not noticing!

ATI Mobility x1600, Ubuntu 10.04

X.Org radeon driver with the following settings:
EXAVSync - off
XV_Vsync - on

therefore not sure which poll option I'll have to choose ...

No tearing, framerate ususally is ~50 but sometimes drops if there are a lot of objects / ships around.

gizmo wrote:

ATI Mobility x1600, Ubuntu 10.04

X.Org radeon driver with the following settings:
EXAVSync - off
XV_Vsync - on

therefore not sure which poll option I'll have to choose ...

No tearing, framerate ususally is ~50 but sometimes drops if there are a lot of objects / ships around.

Dropping is fine due to too much stuff on screen, but the point is that if your monitor refresh is set to 50Hz Oolite with Vsync on shouldn't run above 50Hz either.

To test, check what refresh rate your monitor is running at and then drop game resolution down until, with little on the screen oolite maxes out at refresh of your monitor. Then you'll know if Vsync is working as intended.

That is an improvement! The last time I asked that, I got 0 replies!

Note that oolite is needed to run in Full Screen mode and not in windowed mode
(maximized window will therefore not give any clues about vSync).

gizmo's post for CDT when in Full Screen mode has been split here.