[LAU] Plasma TV as monitor for studio work.

David Olofson david at olofson.net
Sun Jan 25 22:47:54 EST 2009


On Saturday 24 January 2009, alex stone wrote:
[...]
> For those driving larger LCD screens, what's the general consensus
> for graphic cards, and any extra tweaks you might have had to cope
> with?

I've been using on Apple 30" LCD for a little over three years now, 
for pretty much everything; games, racing simulators, programming, 
and the usual web, email etc stuff.

The first issue I ran into was the fact that 2560x1600 requires dual 
link DVI, and that basically wasn't available on anything but high 
end workstation video cards at the time. That caused me some 
ATI/Linux related headaches at first, and forced me to go for a 
top-of-the-line 512 MB gamer card at twice the cost of any normal 
GF6800 card - but that card worked flawlessly with the 30", and (as 
of course) performed very well.

You won't need dual link for any lower resolutions, but if you go for 
a 2560x1600 display, you should be fine with any reasonably serious 
gamer card. I think pretty much any GF 8800 GT or better should have 
dual link DVI, but beware of "cripled" variants like 8400, 8600 
etc... Don't know about the later generation cards.

Anyway, depending on what you want to do, you may have to consider 
that these large, high resolution screens have a lot of pixels to 
move around! Several times as many as any "sensible" size display... 
This severely impacts dragging large windows, playing fullscreen 
video and things like that.

I'm currently using a dual GF 8800 GT SLI setup with 1 GB VRAM per 
card (rougly equivalent to one of the latest generation nVidia cards, 
according to benchmarks), and that's just about sufficient for 
playing Half-life 2, Doom 3, Quake 4 and similar at the native 
display resolution without dropping below 60 fps too frequently. 
(That is with "everything maxed out", except for FSAA, which doesn't 
really add all that much in these resolutions.)

You certainly won't need a multi-GPU setup for normal desktop work, 
but you should probably stay away from the budget models with some 
disabled pipelines, slow VRAM and stuff.

If you're going to use these new OpenGL rendered desktop environments, 
I'm not quite sure what you need, actually... I've not exactly been 
impressed so far, but OTOH, the HL2 GUI (translucent windows and 
stuff over rather complex real time animated 3D scenes) is perfectly 
nice and smooth, so I strongly doubt it's a raw performance issue.

Finally, from a user POV, it might be worth noting that a display this 
large doesn't really work like a bigger single display. At a normal 
working distance (close enough that you don't need 20+ pixels tall 
letters to be able to read), you can't see that MSN client flashing 
in the corner if you're reading something over at the other side of 
the display! It's more like working with a multihead setup, only 
cleaner, more efficient and more flexible. I totally loved it from 
the start, but apparently, some people just never seem to get used to 
it.


//David Olofson - Programmer, Composer, Open Source Advocate

.-------  http://olofson.net - Games, SDL examples  -------.
|        http://zeespace.net - 2.5D rendering engine       |
|       http://audiality.org - Music/audio engine          |
|     http://eel.olofson.net - Real time scripting         |
'--  http://www.reologica.se - Rheology instrumentation  --'



More information about the Linux-audio-user mailing list