Larger monitor caused slower graphics

[DavidX]

A post here called "Will new graphics card slow the computer" by "Jill" reminded me of something that I could not understand at the time and still don't.

I noticed that Jill recently purchased a larger monitor. I'd just like to mention that when I changed from a 14" to a 17" some time ago the graphics did slow down by a tiny amount, barely perceptible to the human eye but certainly measureable by diagnostics (using identical res and color settings). I'm sure this has absolutely nothing to do with Jill's slow graphics speed (probably drivers, as MadMax said) but I thought it was an interesting point. I had no idea the actual monitor, as opposed to res and color settings, would have any effect on speed. I was demonstrably wrong.

Any comments? (Remember: resolution, refresh and color settings were identical on both monitors for testing purposes and I used the same graphics card and the same tests. And both monitors had the correct inf files.)

[LJE2]

The size of the monitor does not effect its performance, the quality of the monitor does, if you have a good to high quality 15" monitor and then you buy a low priced 17" or 19" monitor yes you will experience slower graphics.

[a Bill]

Another point to remember is that when you go from a dinky monitor to a larger model, you normally crank up the resolution. This slows down things a bit too. Just changing from 16 bit color to 32 bit color slows mine down a tad too.

[KillerBug]

Yup, 14inch to 17 inch ment 16bit 640x480 to 32bit 1024x768 for me, but no speed decrease, as the first time I pluged in my 17" was the first time I pluged in the first P2 system I built.

[DavidX]

Yup, you must be right LJE2. The 17" is an ASTVision (probably a re-badged Samsung) from 1995 (bought used). But the worrying thing is that I would have thought the 14" it replaced was even worse - a Packard Bell monitor from 1996! I guess it shows just how slow the AST is! Nevertheless, the AST has a much better picture, even allowing for the natural increase in clarity of a 17". It's crystal clear with no moire and no distortion whatsoever, even at the corners. And currently hooked to an nVidia Riva TNT2 and a Voodoo2, it now moves http://www.sysopt.com/forum/wink.gif

Nevertheless, it is interesting that the quality of a monitor demonstrably affects graphics speed. A point to note for those gamers out there who like to squeeze every last FPS out of their systems.

[SteveC]

I have to say something here - the monitor itself CANNOT possibly effect the graphics performance of a system. It makes no different to the fps reading if you have a top of the line 21" or not even one plugged in! The quality of the monitor makes not the blindest amount of difference either - all that does is makes the picture look better or worse.

The only way you'll see a slowdown is if you get a new screen, bump up the res and colour depth, and the display will slow to a degree. Also if you change your refresh rate, and vsync is turned on, then your frame rate will change according to the refresh rate you have set. At 60hz, you will never go over 60fps, but at 85hz, you can theoretically reach 85 fps. But the screen itself, or the quality of the picture makes no difference at all. It's all to do with the settings on your graphics card.

Cheers,

Steve

[DavidX]

Please read my first post more carefully (particularly the last paragraph), SteveC.

[MadMax]

SteveC~In W9X monitors require a driver. Drivers utilize resources. Resource utilization degrades performance......

Poorly written monitor drivers can reduce overall speed just as easily as poor vid card drivers.

The monitor is a link in the graphics chain. The guns still have to paint the image onto the crt and some do it faster than others. This is impossible to benchmark (except with your eyes) but to say the monitor cannot affect graphics performance is insane. Without the monitor you have no graphics. Without the canvas, you have no painting.

[SteveC]

DavidX - I was just talking in general when I was talking about refresh rates and so on.

MadMax - All monitor drivers are, are a ist of settings for refresh rates and so on. And All crts will 'paint' the screen at the same speed if they're set at the same refresh rate. One screen at 75hz is not faster than another at 75hz.

Cheers,

Steve

[DavidX]

MadMax~What you said makes perfect sense: how can the speed at which the guns paint an image onto the CRT affect a benchmark? Surely, as you said, a benchmark can't measure that.

SteveC~What you said also makes perfect sense: 75Hz is 75Hz, no matter what monitor.

That being the case, can someone please explain to me why there definitely was a measureable difference in graphics speed when the two different monitors were hooked up and when all other factors were identical? This one really has me stumped! Is there something in the equation that I am overlooking here?

[This message has been edited by DavidX (edited 06-26-99).]

[a Bill]

The only thing I can see that you have overlooked is the fact that benchmarks aren't the best guide to go by. They rarely state the same specs since the system changes moment by moment. Benchmarks are just for bragging rights, nothing more. Just like BogoMips.

[Nathan]

There is more to put up, meaning "painting", on a 17", 19", or a 21" screen than there is on a 15' or 14" model. Henceforth, it takes longer for the image to appear on the screen. I find the larger monitors require more video ram on them to bring the picture up on them quicker. Let's say you use the same video card, 4 MB, and a 14" monitor. You install a 17" monitor. Now there is more to paint. That video has to fill the whole screen up with some type of color. That takes up resouces. So on the bigger monitors, you need more video ram to help refresh the screen quicker, bring the colors up to the 16 bit or 24 bit range and to keep the HZ up high as to not give you eye strain.

There is more to it than that, but what you are noticing is right. There is just more to it than I can explain. Others here have already done a great job on that. Bottom line,

17" monitors should have at least 4 MB on the video card.

19" monitors should have at least 8 MB on the video card.

So yes, I notice what you do on the images appearing on the screen on the larger monitors. I don't think I explained it properly, but it's not easy either. I gave it my best shot. Hope it helps.

[MadMax]

SteveC~Really? Try loading all of the W9X drivers for your monitor, one at a time. Leave your monitor at 640x480@60Hz. How many drivers put the picture up correctly? If all the monitor drivers do is provide a list of settings, you would think that at this low and common setting all of the displays would be the same.

The OS must interpret the monitor driver just as is does any other driver, and if monitor drivers were all the same there would be no need for a driver for every make and model. Nathan has the right answer. To explain it in detail would take forever.

[DavidX]

Nathan
Many thanks for your reply. At the risk of labouring the point, can I just ask you to clarify one thing? You said:

"Let's say you use the same video card, 4 MB, and a 14" monitor. You install a 17" monitor. Now there is more to paint. That video has to fill the whole screen up with some type of color."
This is assuming that resolution, color settings and refresh (say 640 X 480 X 16 bit @ 75Hz for the sake of argument) is the same on both monitors. Therefore the total amount of pixels is exactly the same on both monitors.

Are you saying then that although the total amount of pixels is identical, it is the fact that each individual pixel is larger on a 17" that causes it to require more power than the same pixel on a 14"?


MadMax and SteveC
I must admit that when I read through a monitor inf file ('driver') it does look like merely a list of refresh setting and resolution ranges for each type of monitor. In fact, when I contacted AST for clarification they replied:
"Please realize that monitors do not require drivers. Support for the monitor is derived from the installed video controller."
Are they talking rubbish? You certainly need the correct inf file to let Windows know what the maximum refresh capability of your monitor is (or risk killing your monitor!).

Also, when you look at the Driver tab for a monitor in System Properties it says:
"No driver files are required or have been loaded for this device". Does this mean it only looks at them if you need to change the refresh rate or the resolution?




[This message has been edited by DavidX (edited 06-26-99).]

[Nathan]

Well... yes and no. Generally speaking when people get a bigger monitor, they chose the smaller settings so they can get more on the screen. And like MadMax said, it would take a lot more explaining than I am capable of writing here.

I try to keep things simple when expressing things (easier said than done) to people. While I can talk with the best of them, I try not to repeat exactly what they say. When trying to explain to a customer what the problem is and the solution, it's best not to talk too much over their head. Henceforth the reasoning for this because it's hard for me to understand it, let alone try to explain it too.

Sorry, but this is the best I can do.

[This message has been edited by Nathan (edited 06-27-99).]

[BBA]

I think I'll have to settle this myself!

When I get around to it, I'm gonna run a timedemo on my 19" Sony then a timedemo with the same pc on a 14"CTX then a 17" Digital, all at the same res and refresh (800x600 and 60Hz)

I'll make a continuation thread to this when Im done!

BBA

[DavidX]

Thanks BBA! I'll look forward to seeing the results. Maybe I should try another test myself too. The old Packard Bell has long gone but I've got a few other 14"ers around here.

[This message has been edited by DavidX (edited 06-27-99).]

[steves]

Bit of a slanging match developing here, but we seem to be getting to the truth - albeit in a round about way.

Just to summarise:

The physical monitor itself cannot affect the benchmarks. And the monitor 'drivers' are only really a .inf file containing limit/capability information.

However the settings in the .inf file will affect the graphics card setup. As a result switching between plug and play monitors (or from non-pnp to pnp), (without changing resolution/colour/refresh) may have some affect on graphics preformance, as different .inf files will be used. The difference will be small, but David's orginial post was only talking about a small change.

If plug and play is not used (i.e generic monitor settings used all the time) and the same resolution/colour/refresh is used there should be no performance difference between monitors.