EDID issues
EDID issues
First off I have no idea where would be the "proper" place to post this so it's here...
MANY MANY people are having issues with their monitors not being properly detected. This is not confined to Ubuntu. It's happening in other distros too. I am writing this as an nvidia user, but this does not appear to be strictly confined to nvidia either.
I have researched this all over the net for several days. In summary, the scenario goes something like this:
"I can't get my monitor above 640x480. The higher resolutions aren't available in screens and resolution."
"Try using read-edid. I bet you have bad EDID data from your monitor"
"Ok I got read-edid installed. I ran get-edid |parse-edid and it tells me the data is corrupt!"
"Yup... your monitor is outputting bad EDID. You need to do *this* to work around it."
OK, stop scenario. *this* usually equates to using a hex editor on your EDID data because it apparently has a bad checksum after somehow extracting it with phoenix edid editor running under wine or.... it goes on and on.
I did all of these things. Using ddcprobe the final line is "edidfail:". I checked into DVI cabling and what pins are carrying the DDC between graphics card and monitor... all kinds of stuff. I had the *gasp!* corrupt EDID on my monitor too!
Ok so this isn't confined to one monitor, one manufacturer, nothing. Does anyone else have a problem with the assumption that ALL OF THESE monitors from a variety of manufacturers have bad EDID data?
Guess what. They DON'T. Last night I started nvidia-settings and used the "Acquire EDID" button it provides. Well... it retrieved EDID from my monitor just fine. Did it really? So I ran parse-edid on that file the nvidia-settings retrieved and created.
parse-edid<edid.bin
Lo and behold it parses absolutely fine! There's nothing wrong with my monitor's EDID data after all. Huzzah!
It seems get-edid doesn't handle the newer EDID version 1.3 data sets correctly. My monitor is using EDID 1.3. It says so in the EDID data.
What I DON'T know is what mechanism is being used to read EDID from the monitor when it's being detected. I am assuming it's not get-edid because that's not included with a default installation.
This issue has also been blamed on the nvidia restricted drivers. I don't know yet if they play a part in this. This weekend I'm going to do another clean install and test if EDID retrieval works on a default install. In other words... can xorg properly detect my monitor in the default install without restricted drivers being there to take the blame?
My guess is that whatever mechanism is reading EDID for monitor detection, has not been updated to handle EDID 1.3 data. this would explain the LARGE number of different monitors having this issue.
The lack of official help on this is astounding. I paged through many bug reports and almost all of them blame the drivers, blame the monitor for having corrupt EDID data responses, suggest that the user has bad cabling... and more.
EDID 1.3 has different header bytes than EDID 1.2 and previous. I am REALLY beginning to suspect the problem lies in the detection tools, and not the data.
Dell Zino HD, AMD Neo X2 e6850, 8GB, 750GB, ATI HD4330 in MXM slot.
Bookmarks