Running Headless: How to Force Display Rez?

P

(PeteCresswell)

I am running a Windows 7 box headless - using TeamViewer to get
to it as needed.

Problem is that as soon as I disconnect the monitor and reboot,
Windows reverts to VGA mode and I cannot find a means of telling
it to do otherwise.

HP p6-2107c, using "AMD Radeon HD 6530D Graphics"

Maybe replace the Radeon driver with some sort of special driver?
 
P

(PeteCresswell)

Per (PeteCresswell):
HP p6-2107c, using "AMD Radeon HD 6530D Graphics"

Maybe replace the Radeon driver with some sort of special driver?
Well, I got partway there: took a big breath, crossed my
fingers, and disabled the Radeon Graphics driver.

That gets me a max of 1600 x 1200 - which I can live with.

Still, 'twood be nice if I could fake it into doing 1920 x 1200.

So the question stands: is there a way?
 
P

Paul

(PeteCresswell) said:
I am running a Windows 7 box headless - using TeamViewer to get
to it as needed.

Problem is that as soon as I disconnect the monitor and reboot,
Windows reverts to VGA mode and I cannot find a means of telling
it to do otherwise.

HP p6-2107c, using "AMD Radeon HD 6530D Graphics"

Maybe replace the Radeon driver with some sort of special driver?
There are a couple possibilities.

There could be a utility out there, to force it. There is
PowerStrip, which is used for custom resolutions, but I don't
know if that includes a forcing function (to override EDID).
You should be able to evaluate that for 30 days, to see
if it can do anything for you. For more than 30 days usage,
you pay for it.

http://www.entechtaiwan.com/util/ps.shtm

In terms of hardware, there are two aspects to monitor
operation. The video card has "impedance sensing", and it
can tell if you've unplugged the connector. I keep a set of
"fake" connectors here, like a VGA connector with 75 ohm
resistors on R to ground, G to ground, B to ground. That
creates a fake electrical load. It allows me to fool
the OS into thinking a monitor is present. When I need to test
"dual monitor configurations", when I only owned the one LCD
monitor, I used my "fake VGA" connector to trick the
video card into running dual monitor mode. So that
basically gets the video card to enable the outputs.
On DVI, faking would need something like 100 ohm differential
termination on the high speed signals. I don't have any solder
tail DVI connectors, to build one like that, but I assume it
would work as well. You can also do fake composite and S-Video,
with resistors. (So some of my "fake monitor" experiments, use
composite video terminators.)

( Two video cards, one real monitor, three "fake" monitor connections.
Resolution of monitors: 1280 - 800 - 1280 - 640. Screen capture.)

http://img341.imageshack.us/img341/6043/extendednvidia128080012.gif

The problem with that, is the video driver will limit resolution choices
to "safe" values. Which means I probably can't get 1600x1200 that way.
In the era before multisync monitors became available, there
were resolutions which worked "most of the time" and were considered
"safe" by video card driver writers. And without information coming
from the monitor, the driver may decide to limit the resolution
and refresh to safe values.

There are hardware boxes, which will copy the EDID from a monitor,
and then present that information (read-back mode) when the monitor
is not present. So it's possible to fool the computer, entirely
in hardware, that a monitor is present. This Gefen box is an
example of the technology (a 2Kbit EEPROM in a fancy metal box).
Off brand copy cats of this, may be available for less money.
I don't consider this approach to be cost effective for
your application - this is for people with projection TV sets,
where the set lacks EDID, and the projector costs many times
what the little EDID box costs.

http://www.gefen.com/kvm/dproduct.jsp?prod_id=4714

The Macintosh had a simple scheme for forcing resolution. The
pins on the connector, included an encoding for resolution. I used
to use a "dip switch box", to force resolution to a value my CRT
monitor could use. The PC also had such a scheme, before the EDID
serial clock and data interface came along, but I don't think it
could represent quite as many choices. Resolution forcing by
sense pins, that died out a long time ago. I don't know if
video cards still listen to that or not. In any case, I doubt
the resolution values would be useful ones.

So, test PowerStrip, and see if it does anything useful for you.
I "burned up" my evaluation years ago, and I'd probably have to
do a clean install to be able to test it again.

I've tried to find resolution settings in the Registry, but
it's a maze in there. The programmer at Entechtaiwan probably
knows what to do, but I don't.

Paul
 
P

Paul

(PeteCresswell) said:
Per (PeteCresswell):

Well, I got partway there: took a big breath, crossed my
fingers, and disabled the Radeon Graphics driver.

That gets me a max of 1600 x 1200 - which I can live with.

Still, 'twood be nice if I could fake it into doing 1920 x 1200.

So the question stands: is there a way?
An example of a utility here.

There is no reason to believe such a utility, can exceed whatever
"max" the current driver is implementing. Maybe you'd need to
go back to the Radeon driver, before experimenting with this.
You'd toss a command like this in some Startup thing, then see
if it takes.

http://www.nirsoft.net/utils/nircmd.html

nircmd.exe setdisplay 800 600 24

HTH,
Paul
 
W

Wolf K

(PeteCresswell) said:
Per (PeteCresswell):

Well, I got partway there: took a big breath, crossed my
fingers, and disabled the Radeon Graphics driver.

That gets me a max of 1600 x 1200 - which I can live with.

Still, 'twood be nice if I could fake it into doing 1920 x 1200.

So the question stands: is there a way?
An example of a utility here.

There is no reason to believe such a utility, can exceed whatever
"max" the current driver is implementing. [...]
You can't actually exceed the physical resolution of the screen. Think
about it. On LCD/LED screens, the physical resolution is determined by
the size of the LCD/LED cells. On a CRT, it's determined by the size of
the raster between the gun and the phosphor screen.

You can, if you wish, set a lower resolution, which is accomplished by
dithering. This will produce and/or exaggerate artefacts, which may be
bad enough to bother you. Depends on how far off the physical res you
go, and on what you can see and tolerate.
 
P

Paul

Wolf said:
(PeteCresswell) said:
Per (PeteCresswell):
HP p6-2107c, using "AMD Radeon HD 6530D Graphics"

Maybe replace the Radeon driver with some sort of special driver?

Well, I got partway there: took a big breath, crossed my
fingers, and disabled the Radeon Graphics driver.

That gets me a max of 1600 x 1200 - which I can live with.

Still, 'twood be nice if I could fake it into doing 1920 x 1200.

So the question stands: is there a way?
An example of a utility here.

There is no reason to believe such a utility, can exceed whatever
"max" the current driver is implementing. [...]
You can't actually exceed the physical resolution of the screen. Think
about it. On LCD/LED screens, the physical resolution is determined by
the size of the LCD/LED cells. On a CRT, it's determined by the size of
the raster between the gun and the phosphor screen.

You can, if you wish, set a lower resolution, which is accomplished by
dithering. This will produce and/or exaggerate artefacts, which may be
bad enough to bother you. Depends on how far off the physical res you
go, and on what you can see and tolerate.
But Peter is trying to run headless (i.e. monitor cannot state a
resolution limit, as there is no monitor connected), so there's
nothing to damage.

Headless means, you set up the system, then later, disconnect the
monitor and access the computer through TeamViewer. At that point,
it doesn't matter how the output is set. The monitor cannot be
damaged by the resolution setting, because there is no monitor.

Paul
 
P

(PeteCresswell)

Per Paul:
resolutions, but I don't
know if that includes a forcing function (to override EDID).
You should be able to evaluate that for 30 days, to see
if it can do anything for you. For more than 30 days usage,
you pay for it.

http://www.entechtaiwan.com/util/ps.shtm
Gave it a try, but PowerStrip didn't seem tb having any of it.

Could have been my own ignorance... but it seemed to me like it
could not go beyond the built-in limitations of the built-in VGA
driver.
 
C

Char Jackson

Per Paul:

I started, then chickened out when WebOfTrust gave the site it's
worst rating in all categories.
I don't know what the basis of that might be, but Nirsoft has been
around for ages and have proven themselves to be completely
trustworthy.
 
P

Paul

(PeteCresswell) said:
Per Paul:

Gave it a try, but PowerStrip didn't seem tb having any of it.

Could have been my own ignorance... but it seemed to me like it
could not go beyond the built-in limitations of the built-in VGA
driver.
Try putting back the Radeon driver before you give up on it.

Paul
 
G

Gene E. Bloch

Per Paul:

I started, then chickened out when WebOfTrust gave the site it's
worst rating in all categories.
Two things.

1. On my browser, WOT seems to like the above URL (four green circles).

2. I have downloaded and used a few NirSoft apps over time, no problems,
no viruses...I respect that site & their stuff.

Did you directly click, or Ctrl-C copy, the URL? Did you perhaps type it
by hand and make a typo?
 
G

Gene E. Bloch

Two things.

1. On my browser, WOT seems to like the above URL (four green circles).

2. I have downloaded and used a few NirSoft apps over time, no problems,
no viruses...I respect that site & their stuff.

Did you directly click, or Ctrl-C copy, the URL? Did you perhaps type it
by hand and make a typo?
In fact, I have used that very app above, nircmd, in a script to turn
off the monitor during an overnight backup operation, because the
brightly lit screen interfered with my sleep :)

My real favorite NirSoft utility is the one that translates minidumps
into something almost understandable. The fact that it isn't completely
understandable is *not* the fault of Nir Sofer :)
 
R

Robin Bignall

In fact, I have used that very app above, nircmd, in a script to turn
off the monitor during an overnight backup operation, because the
brightly lit screen interfered with my sleep :)
Can you imagine normal people wanting to sit in their bedrooms for hours
on end corresponding with other people, few, if any, they know or will
ever meet? Me neither.
 
C

Char Jackson

Can you imagine normal people wanting to sit in their bedrooms for hours
on end corresponding with other people, few, if any, they know or will
ever meet? Me neither.
When IRC popularity peaked among my friends, back in the mid 1990s, it
wasn't unusual for some of them to be online and chatting for 15-18
hours at a time, and sometimes longer. I know of more than a handful
of marriages that broke up as a result.
 
P

(PeteCresswell)

Per Char Jackson:
I don't know what the basis of that might be, but Nirsoft has been
around for ages and have proven themselves to be completely
trustworthy.
Based on that, I went back and clicked the "Download 64-bit
version" button, but Chrome popped "nircmd-x64.zip appears
malicious".

I don't know who that is coming from. Avast, maybe?

I think I'll take Paul's suggestion and go back and aim
PowerStrip at an installed Radeon driver.
 
P

(PeteCresswell)

Per Paul:
Try putting back the Radeon driver before you give up on it.
Now it seems to have stopped offering up any choices except for
opening the Screen Rez control panel.
 
J

Joe Morris

(PeteCresswell) said:
Per Paul:

I started, then chickened out when WebOfTrust gave the site it's
worst rating in all categories.
The problem is likely that nirsoft offers a number of administrator-type
tools which can be used for both good and evil. One consequence of this is
that the anti-virus program on my systems routinely flags a number of
nirsoft tools as malware until I tell it to ignore the folder where I keep
them; this could lead some reports to consider nirsoft to be a nasty place.

H'mmm...just went to the WOT web site; it classifies nirsoft as "excellent"
in *ALL* categories, although it notes that the site has "appeared on a list
of malware distributors" published by Malware Patrol.

http://www.mywot.com/en/scorecard/nirsoft.net

ObGrammarPolice: the word you wanted is "its", not "it's".

Joe
 
P

Paul

(PeteCresswell) said:
Per Paul:


Now it seems to have stopped offering up any choices except for
opening the Screen Rez control panel.
Well, it was just a thought. I can't play with it any more here,
as my "eval" for Powerstrip, ran out years ago.

Maybe you'll just have to live with 1600x1200. That should be enough
to administer the remote box.

It's one thing, to perhaps jam something into the Registry to
set the resolution. It's the behavior of the driver that we
can't predict (what it does when the monitor is missing).
In theory, the video card (via the driver) is supposed to turn off
outputs which have no monitor connected. But the Microsoft VESA driver
(whatever runs when the Radeon is missing), might not be as well equipped
to police those rules.

Before impedance sensing became available, video cards could just
keep the outputs enabled. So whether this is an "improvement"
is debatable. It could be, that someone thought it would reduce
emissions from the computer or something. I've never read a
justification for the usage of impedance sensing.

Paul
 
B

BillW50

Before impedance sensing became available, video cards could just
keep the outputs enabled. So whether this is an "improvement"
is debatable. It could be, that someone thought it would reduce
emissions from the computer or something. I've never read a
justification for the usage of impedance sensing.
Having a massive mismatch of impedance with an amp causes all power
going out to be reflected right back into the amp. And if the design
can't handle twice the power of normal, the final will blow like a fuse.

In this case, we are talking about very small power output. So the final
handling twice the power of normal most likely won't hurt the final. But
in any case, it is still good practice to just switch it off if there is
no load to deliver to.

And true emissions do increase when there is a mismatch of impedance. So
that would be another reason to do so.
 
C

Char Jackson

Per Char Jackson:

Based on that, I went back and clicked the "Download 64-bit
version" button, but Chrome popped "nircmd-x64.zip appears
malicious".

I don't know who that is coming from. Avast, maybe?

I think I'll take Paul's suggestion and go back and aim
PowerStrip at an installed Radeon driver.
*shrug* Suit yourself. I don't get the whole "scared to download"
thing. You can download malicious files all day long and not have a
problem until you execute one or more of them. In this case, you could
simply download, scan with your favorite scanner(s), upload to
virustotal, whatever. I see absolutely NO reason to be scared to
download something, especially something from a site like Nirsoft.

Sorry for the rant, but I just went through a similar scenario with
one of my clients. You give them the tools, you give them the
knowledge, and still they freeze.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top