Feature Request - implement EDID override

[idata]

Hello,

it would be nice if the Intel Graphics driver would support overriding EDID information like described in this document from Microsoft:

http://msdn.microsoft.com/en-us/windows/hardware/gg487330 http://msdn.microsoft.com/en-us/windows/hardware/gg487330

Why is this needed?

In a common HTPC (multimedia PC) setup with a PC, AVR (audio video receiver) and a television, the EDID information stored in the receiver & television that is passed to the PC determines the video and audio capabilities like: supported resolutions, supported audio formats etc ..

The problem is that very often the system only "sees" the EDID information of the television, and not that of the receiver, and so you don't get support for audio formats like DTS, DTS-HD, Dolby TrueHD that are supported by the receiver ...

The latest Sandy bridge CPU and corresponding chipset support all these sound standards, but they often not useable cause of these EDID problems.

Actually it really does not matter what video card / integrated graphics you use, NVidia and AMD graphics have the same problem, but their drivers

allows people to override the EDID information like is described in the above linked document from Microsoft.

Please implement this stuff, shouldn't be that hard to do, for one of you senior driver programmers it's probably just a few hours of work ...

Thank you

Christian

[Allan_J_Intel1]

Hi there!

Your feedback is important to us and I thank you for taking the time to communicate this issue.

Thanks

Allan.

[SHick1]

I continue to experience this issue. When will intel support EDID Override?

[idata]

Currently I am using the very expensive Gefen HDMI Detective Plus to do the job of constantly providing the same EDID info to the HTPC. However, overriding it with an INF file (for example generated by MonInfo) would be a much more elegant solution.

[gbhj]

allan_intel

can information about the following options available in the Intel driver be made available?

1. fakeEDID
2. OEM static mode
3. creating custom DTD

[Allan_J_Intel1]

Thanks for the request.

Allan.

[idata]

So, It's been 5 years - any news?

[TMill18]

Any progress on this? I'm really struggling to get my expansive HTPC to work properly with a connected Marantz AVR. A simple EDID override feature would help me greatly.

[KHave]

Well, this would be an easy thing for Intel fix which would fix a lot of negative experience with their products when used as HTPC. So looking forward to see this one closed soon.

[JTerr5]

Intel: PLEASE STOP IGNORING THIS ISSUE!!! BOTH AMD AND NVIDIA ALLOW THIS. SO PLEASE GET YOUR BIG BOY PANTS ON AND ALLOW THIS FEATURE! HOLY SHIT!

I will NEVER buy another Intel product ever again because of this issues.

[PB7]

Great work ignoring this Intel.

[RSyne]

I also suffer from this issue. I purchased a very expensive Gigabyte Brix I4770R running Intel 5200 Iris Pro. I have this connected to an LG 5.1 amp that should play 5.1 and does via XBOX 360 and Virgin Media Cable Box. The LG amp is connected to an LG 42" TV that only supports 2 channel stereo and it's this restriction that is being picked up by the Intel card and causing windows 8.1 to restrict my sound to 2 channel.

I used to run an Acer Revo (NVIDIA ION) and this worked fine where I could set the sound to 5.1 in Windows.

Reading up on the issue it appears I am limited to using an expensive hardware splitter/switch that spoofs the EDID, running an optical cable to the amp (old school) , trying to hack the registry (really???), or waiting for Intel to implement what appears to be a MUST HAVE feature in their driver set. I can't believe it isn't part of it already.....

The upside is that you can force 5.1 in Netflix and the amp does pick up the audio stream but not all apps offer this functionality.

[RSyne]

O I forgot.... I could buy a new £1000 Sony TV that does support 5.1

*looks at the wife*

[loldf]

Seriously Intel. This is ridiculous. Add support for EDID override!

err... open source?

Does the hardware support the edid being injected via software???

[loldf]

Looks like the clone_update_edid has the capability of reading an EDID from one display and "cloning" it to another within the intel graphics driver...

hmm..z

static void clone_update_edid(struct clone *clone)

{

unsigned long nitems, after;

unsigned char *data;

int format;

Atom type;

if (XRRGetOutputProperty(clone->dst.dpy, clone->dst.rr_output,

XInternAtom(clone->dst.dpy, "EDID", False),

0, 100, False, False, AnyPropertyType,

&type, &format, &nitems, &after, &data) == Success) {

XRRChangeOutputProperty(clone->src.dpy, clone->src.rr_output,

XInternAtom(clone->src.dpy, "EDID", False),

type, format, PropModeReplace, data, nitems);

}

}

from xf86-video-intel-2.99.917/tools/virtual.c

one could say, import a known edid.bin and inject it into a screen? therefor forcing it to a particular mode..

any avid xf86 driver coders available?

[loldf]

1) Add OPTION_EDID_OVERRIDE to intel_options.h

2) Add OPTION_EDID_OVERRIDE to intel_options.c with "OverridEdid", OPTC_STRING, {0}, 0

more later.. come on Intel lets get this edid overriden! it looks like you've already written EDID cloning from one output to another... so it must be supported

[idata]

It seems this is now possible in the linux driver, but still not working in Windows. I've tried two possible ways of doing this and neither one seems to work. One is to specify a custom resolution in the intel graphics control panel, and the other is to create an edid override in the windows registry. Neither one is working for me. The control panel always gives a message about "bandwith exceeded" when this is clearly not the case, and the intel driver seems to ignore the edid_override in the registry. In linux I just specify the modeline I want to xorg.conf or xrandr and this same hardware does exactly what I want. There's no way to do it in windows as far as I can tell.

Thanks

[loldf]

checkout the github.

https://github.com/ArakniD/IntelGraphics https://github.com/ArakniD/IntelGraphics

It now has a option in xorg.conf for;

Section "Device"

Identifier "Device 1"

Driver "intel"

Option "EDIDOverride" "/home/lucas/sony.edid"

EndSection

if the file doesn't exist it'll create it the first time assuming everything is connected OK.

the next time it'll load the file instead.. handy? ;0 I think so!

7776.574] (==) intel(0): Depth 24, (--) framebuffer bpp 32

[ 7776.574] (==) intel(0): RGB weight 888

[ 7776.574] (==) intel(0): Default visual is TrueColor

[ 7776.574] (**) intel(0): Option "EDIDOverride" "/home/lucas/sony.edid"

[ 7776.574] (II) intel(0): Output HDMI1 has no monitor section

[ 7776.574] (II) intel(0): Enabled output HDMI1

[ 7776.574] (II) intel(0): Output DP1 has no monitor section

[ 7776.574] (II) in

Now i'm still having an issue with a blank screen if I turn off the AVR and TV. I hope to resolve this with some further digging.

[loldf]

its 12:31AM and I must sleep... i've reset my AVR with the TV on and the screen came back up ok.. the screen was previously off but the audio came back on ok.

it would appear that its working ok now with an edid override option ;0

w00t

[loldf]

added bug to bug tracker.. perhaps this will get to the correct place now.

https://bugs.freedesktop.org/show_bug.cgi?id=89945 https://bugs.freedesktop.org/show_bug.cgi?id=89945

[ROBERT_U_Intel]

All,

We are currently evaluating the addition of this feature. Please watch this thread for updates.

Thanks

Robert