Very Computer

I am looking for a good 19-21" monitor/video card
to run X in 1280x1024 under Linux. I would like a
refresh rate of at least 80Hz.

Of monitors, two that were recommended to me are:
        Nanao T2-20
        Sony 20SE

Of video cards, I was recommended:
        Stealth 64 Video w/4M
        ATI Mach 64 (Ultra pro?)

Thanx a lot!

Alik

Quote:>Of video cards, I was recommended:
>    Stealth 64 Video w/4M
>    ATI Mach 64 (Ultra pro?)

                                -Christian

Quote:>The real question for me, at least, is can you really tell the
>difference between, for example, 75Hz and 80Hz refresh?  Long ago I had
>a course on Human Ergonomics that indicated a 70Hz refresh (assuming
>normal persistence for the phosphors used in commercial monitors) should
>be close to the maximum detectable by the normal human eye.  At 80Hz,
>you should be in the region where changes are indistinguishable to all
>but electronic equipment, unless you have extremely short persistence
>phosphors in the monitor (this is not likely with standard commercial
>monitors).  If there's an ergonomics student out there, or a professor,
>I'd be happy to learn about the current state of knowledge.  Even a URL
>for a current research paper would be interesting!

--
Don Dettke,  Software Engineering Consultant

Quote:>The real question for me, at least, is can you really tell the
>difference between, for example, 75Hz and 80Hz refresh?  Long ago I had
>a course on Human Ergonomics that indicated a 70Hz refresh (assuming
>normal persistence for the phosphors used in commercial monitors) should
>be close to the maximum detectable by the normal human eye.  At 80Hz,

I find that the particular monitor and the persistence of phosphors makes
70, 75, or 85 hz a serious issue. For example, 20"Sony Trinitron monitors show
a substantial improvement going from 70 to 75 hz in terms of human
perceivable flicker. 85hz is even a little better. On the other hand, I have
an old workstation monitor which runs at 35hz interlaced. It was built before
non-interlaced became a buzz-word and employs substantal persistence in the
phosphor to eliminate flicker. You can't tell it's interlaced, ever: the
stability equals that of the Sony at 75 hz.

Robert

The Stealth 64 Video 3000 series exists as two major board types.  The
cheaper Video 3200 (2MB installed, no upgrades) has a 175MHz DAC and the
more expensive 3240 (2MB installed, upgrade to 4MB possible) and 3400
(4MB installed) use the 220MHz DAC.  Obviously, with only 2MB you could
only use 8bpp for 1280x1024 (not enough graphics RAM for more than
1152x900 in 15bpp or 15bpp), and you should be able to get a refresh of
at least 75Hz on the 175MHz DAC, more on the 220MHz DAC part.

The Diamond board is based on the S3 968, and there are several similar
boards:

        Number 9 Motion 771
        STB Velocity 64V
        Elsa Winner 2000 Pro (there's a variant of this that can take
                8MB for 1600x1200x24bpp)

Other boards with high speed DAC's include the Number 9 Imagine 128 and
the Matrox Millenium.  The fastest board that is available, by a rather
long chalk, is the Matrox Millenium, and this too comes with 220MHz and
175MHz DAC's (220 in the end user product, 175 for the OEM board sold to
Gateway and, I think, Compaq, for example).   The second fastest is the
Imagine 128, and the S3 968's take third place, just a shade faster than
the Mach64 - the difference is probably not visible to the human
eye/brain in sequential tests, but will be obvious in a side-by-side
comparison or in benchmarks.  Again, at the risk of provoking some
flaming, I can only speak for our Server benchmarks.  The I-128 under
XFree86 apparently falls somewhere around the performance of a low end
Cirrus logic chipset, or perhaps the Trident 8900.  The Millenium is not
supported by XFree86 as anything other than a slow VGA.  Despite the
mouth frothing elsewhere, one does have to take Server software into
account when selecting boards for optimum performance.

The real question for me, at least, is can you really tell the
difference between, for example, 75Hz and 80Hz refresh?  Long ago I had
a course on Human Ergonomics that indicated a 70Hz refresh (assuming
normal persistence for the phosphors used in commercial monitors) should
be close to the maximum detectable by the normal human eye.  At 80Hz,
you should be in the region where changes are indistinguishable to all
but electronic equipment, unless you have extremely short persistence
phosphors in the monitor (this is not likely with standard commercial
monitors).  If there's an ergonomics student out there, or a professor,
I'd be happy to learn about the current state of knowledge.  Even a URL
for a current research paper would be interesting!

Cheers, JeremyC.
--
Jeremy Chatfield +1 303/298-7478 FAX:+1 303/298-1406

X Inside Inc, 1801 Broadway, 17th Floor, Denver, CO 80202
Commercial X Server - for more information please try these services

 <SNIP> (A portion of Jeremy's comments follows)

|>
|> The Stealth 64 Video 3000 series exists as two major board types.  The
|> cheaper Video 3200 (2MB installed, no upgrades) has a 175MHz DAC and the
|> more expensive 3240 (2MB installed, upgrade to 4MB possible) and 3400
|> (4MB installed) use the 220MHz DAC.  Obviously, with only 2MB you could
|> only use 8bpp for 1280x1024 (not enough graphics RAM for more than
|> 1152x900 in 15bpp or 15bpp), and you should be able to get a refresh of
|> at least 75Hz on the 175MHz DAC, more on the 220MHz DAC part.
|>

 <SNIP>

This is interesting to me.  I recently purchased a machine that
includes an OEM DIamond Stealth 64 Video 3240 with the TI ViewPoint
3026 RAMDAC, which accordingto TI runs in 135MHz, 175MHz, and
220MHz modes; see their info page at:

   http://www.ti.com/sc/docs/msp/showcase/vol11/ramdac.htm    

When I fire up X (XFree86-3.1.2) the server reports that my maximum
clock is 135MHz!! Is this limitation due to the server, or is
the 'OEM' bashing my performance this much?  I guess my question
is, Is this a hardware limitation that was set at the factory, or
do I have unused potential? I haven't attempted to tweak too much,
but I am managing 71 Hz at 1280x1024 (8bpp, naturally, as I have
only 2MB VRAM).  I wouldn't mind a few more Hz of refresh as I catch
a little flicker peripherally on occasion, and on a 17" monitor
greater resolution seems a little pointless.

So now I have two questions:

  (1) Is there any way to test my RAMDAC speed independently of X?
  (2) Would a commercial server be able to use my hardware's full
      potential when XFree is not (assuming that's the case)?

<snip>

Quote:>|> The Stealth 64 Video 3000 series exists as two major board types.  The
>|> cheaper Video 3200 (2MB installed, no upgrades) has a 175MHz DAC and the
>|> more expensive 3240 (2MB installed, upgrade to 4MB possible) and 3400
>|> (4MB installed) use the 220MHz DAC.  Obviously, with only 2MB you could
>|> only use 8bpp for 1280x1024 (not enough graphics RAM for more than
>|> 1152x900 in 15bpp or 15bpp), and you should be able to get a refresh of
>|> at least 75Hz on the 175MHz DAC, more on the 220MHz DAC part.
> <SNIP>
>This is interesting to me.  I recently purchased a machine that
>includes an OEM DIamond Stealth 64 Video 3240 with the TI ViewPoint
>3026 RAMDAC, which accordingto TI runs in 135MHz, 175MHz, and
>220MHz modes; see their info page at:
>   http://www.ti.com/sc/docs/msp/showcase/vol11/ramdac.htm    

Quote:>When I fire up X (XFree86-3.1.2) the server reports that my maximum
>clock is 135MHz!! Is this limitation due to the server, or is
>the 'OEM' bashing my performance this much?  I guess my question
>is, Is this a hardware limitation that was set at the factory, or
>do I have unused potential? I haven't attempted to tweak too much,
>but I am managing 71 Hz at 1280x1024 (8bpp, naturally, as I have
>only 2MB VRAM).  I wouldn't mind a few more Hz of refresh as I catch
>a little flicker peripherally on occasion, and on a 17" monitor
>greater resolution seems a little pointless.
>So now I have two questions:
>  (1) Is there any way to test my RAMDAC speed independently of X?
>  (2) Would a commercial server be able to use my hardware's full
>      potential when XFree is not (assuming that's the case)?

     DACspeed 220

to your "Device" section.  Diamond's web page (back before it was down
all the time) said the 3240 and 3400 DO have the 220MHz RAMDAC, so it
should be safe.

--
Don Dettke,  Software Engineering Consultant

 -     DACspeed 220

 -to your "Device" section.  Diamond's web page (back before it was down
 -all the time) said the 3240 and 3400 DO have the 220MHz RAMDAC, so it
 -should be safe.

Actually, the OEM version of that card only has a 175MHz RAMDAC, not the
220MHz version in the retail card.  In any case, it is faster than
135MHz.
--
Rob Coleman                                   Years go by, will I still be
Master's Program                              waiting for somebody else to
Computer Science / Georgia Tech               understand?

Quote:> >So now I have two questions:
> >  (1) Is there any way to test my RAMDAC speed independently of X?
> >  (2) Would a commercial server be able to use my hardware's full
> >      potential when XFree is not (assuming that's the case)?
> I believe I saw a post a while back indicating that there was a problem
> detecting the max dot clock for this card.  Try adding
>      DACspeed 220

first you should read README.S3 to learn how to get the
clock limit of your RAMDAC.  there is no software way to
probe this; you have to check the chip label on your card,
there is no other way.

and yes, there are Diamond cards with 135 MHz ramdac.
I don't want to know what happens to this cards if you run them
with 220 MHz (but yes, the XFree86 S3 server can do this if you ask it...)

then you should read the server & config file man pages to learn
how to set the DECspeed in your config file...

Harald

--
All SCSI disks will from now on                     ___       _____
be required to send an email notice                0--,|    /OOOOOOO\
24 hours prior to complete hardware failure!      <_/  /  /OOOOOOOOOOO\
                                                    \  \/OOOOOOOOOOOOOOO\
                                                      \ OOOOOOOOOOOOOOOOO|//
Harald Koenig,                                         \/\/\/\/\/\/\/\/\/
Inst.f.Theoret.Astrophysik                              //  /     \\  \

It is possible that flicker in the ambient lighting can cause a perceived
problem in viewing the display, due to modulation of the display contrast
ratio - but this is WAY down the list of worries.

Senior Engineer, Displays & Human Interface | Note: The opinions presented
Workstation Systems Division                | here are not those of my employer
Hewlett-Packard Co., Ft. Collins, CO        | or of any rational person.      

I recently posted about this same subject (damned Diamond cards...grrr..) and
how managed to solve it. I too ended up with an OEM DS Vid 64 card, but used a
wonderful program called xvidtune to pump up all of my specs until they are
exactly as I want them (maxxed out, that is (-; ). I wish that I could be more
help in locating it (I got a ton of mail last time from people trying to find
it) but I must admit that I don't know where it is. I installed with Slack 3.0
though, so try digging through a slackware mirror to find it. It is truly an
amazing program, allowing runtime adjustement of everything but color depth.
Using it, I was able to get 1600x1200 at 70.3 Hz refresh, with only 175MHz dot
clock. I am really happy with that, as I have no way to get a 220MHz card. Any
way, it enabled me, who had never done this sort of thing at all before, to
get everything set up just the way I wanted it, I hope everyone else that got
saddled with the OEM version can use it to their advantage as well. good luck!

--michael
running linux for one whole month now :-)

Quote:

>I read Diamond's web page when I was shopping for a video card a couple of
>months ago and it left no question that the 3200 had a 175MHz RAMDAC and
>the 3240 and 3400 had a 220MHz RAMDAC.  It didn't even mention a 135MHz

    If you visually inspect to RAMDAC, look for 'RGB526' and CF22
    on it.  This is the 220 Mhz version (there is also a  175 Mhz
    variant with CF17 printed on it).  You can find this information
    at http://fnctsrv0.chips.ibm.com/products/display/rgb526/rgb526.html.
    Info for 220Mhz RAMDAC can be found at www.ti.com.  Maybe the 3200
    uses to IBM RAMDAC, and the 3240 and 3400 have the TI???  I certainly
    have read in linux NG's that there does exist 175Mhz versions.  It's
    all rather confusing...

--
Paul Colucci
Computational Fluid Dynamics Laboratory
Dept. Mechanical and Aerospace Engineering
State University of New York at Buffalo