The DDC situation
The DDC situation
- Subject: The DDC situation
- From: email@hidden
- Date: Mon, 22 Dec 2008 15:02:55 +0000
Hi to all,
As someone who as worked with DDC since 2002, I can say that the situation is bad and it is not getting better. In the early days, I saw the potential for this technology and my team at Sequel Imaging designed the "enable cable". This let us get rid of the graphic card and OS issues. We had our own proprietary USB interface that allowed us to talk to any device that used a VGA port. This got us over the first hurdle and then we were faced with the larger implementation problems. For instance, some displays had the controls reversed, if you increased the number, you decreased the effect. The DDC specification requires what is termed a capabliities string. We have found in many displays, especially cheaper ones, that the capabilities string does not indicate the true display capabilities. If your display has a vga connector and a DVI connector, it is quite probable that the DDC capabilities are completely different based upon the connector. For instance, one very well
known
vendor, emulates a CRT data path when on the VGA connector. This means that the brightness control moves the black point, the contrast control moves the max luminance. When attached to the DVI connector, brightness control is inoperative, and the contrast control controls the max luminance. I believe that we get more complaints about DDC in our display applications than any other area. Here is a break down of some of the issues so you understand why there is so much finger pointing.
1. We don't release a version of software without testing it. So we have normally confirmed that in atleast one instance, the communication works and the application runs fine. I am then left with the question: Where is the problem in a particular situation.
2.In my opinion and direct personal experience about 30 - 40 % of the platforms have some form of hardware, firmware, driver, or other problem that results in a poor implementation. The DDC command structure utilizes a physical I2C interface with a logical ACCESS bus protocol. For me to send one byte to a display, I need to open a driver, which communicates with the display card driver, which is supposed to have an API that allows generation of I2C output. That I2C output may be implemented in firmware on the Graphics Chip, or there may be a real serial interface engine on the chip. The Mac provides an application API at the OS level so it is generally easier to communicate. It is up to the display card vendors to provide a working I2C connection. On a windows machine (non vista) communication to the display card driver is generally through technology provided by one fo three vendors. In Vista it is in the OS, but it is up to the Graphic Card vendor to implement the ap
i. On
the Mac, there is a very low level interface to the I2C bus which again the the Graphic Card vendor must implement.
3. My own feeling is that at least 60-75% of displays do not implement DDC correctly, although for the needs of this group, it is probably more than 30%. The specification is quite complex and subject to interpretation. For instance a "maximum" control value is supposed to represent a physical maximum, not a logical maximum. As an example, a contrast control could have a logical range of 0 -255, but the display will go into saturation at level 180, the spec says that the maximum should be 180. Many vendors fail to indicate the physical port (analog or digital) in the EDID. In at least 5 instances of displays from a single manufacturer, the capbilities string indicated capablities that didn't exist.
4. The windows vista certification process dictated a new "ColorTemperature" control. This lead to complete havoc in the display market because they couldn't get the Vista Logo on the box if they didn't have this ddc control. This control impacted RGB gains (because that is what it changed). This lead to many vendors abandoning RGB gains and using only the SetColorTemp command.
5. There is a new control that relates to backlight brightness, that many vendors have not implemented, hence a software programmer has to examine the "meaning" of the brightness, contrast and backlight commands when confronted with a new display.
6. In general, DDC has been more stable on the Mac than on the PC. This is because it is part of the IOkit. If Apple broke IOkit, shame on them.
As a consumer, you should be putting pressure on VESA because they are the standards org. Their testing solutions are very limited and very old. My direct experience with the Enable cable showed that eliminating the graphics card and OS from the chain still left one with a myriad of problems generated by the individual display vendors. I don't see any solution, either near or long term, to this morass. Realistically, once display port is implemented, no display will work properly for years to come and we will still be trying to work out the kinks.
Regards,
Tom Lianza
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden