Contribute
Register

Full range of brightness using ACPIBacklight

Status
Not open for further replies.
So I use the Darwin OS patch with both PNLF versions. And both PNLF work very well. I can use 100 levels of brightness in both case. Which PNLF should I prefer? What do you think?

Ok, that was what I was wondering. The best results should be with the longer patch (based on ACPI 100), and the Darwin -> Windows 2012 patch.

The 100 levels should give you the levels "in between" the major levels. Accessed via Option+Brightness Up/Down.
 
Ok, that was what I was wondering. The best results should be with the longer patch (based on ACPI 100), and the Darwin -> Windows 2012 patch.

The 100 levels should give you the levels "in between" the major levels. Accessed via Option+Brightness Up/Down.

It is correct. I use the longer one and it works as you described. Thank you.
 
It is correct. I use the longer one and it works as you described. Thank you.

Super. Thanks for the feedback...
 
Here is the script to auto generate dsdt patch based on RehabMan's dsdt patch and kpkp formula:
http://www.tonymacx86.com/hp-proboo...htness-using-acpibacklight-10.html#post731697

To use it:
Code:
./brightnesspatch low_level high_level

Hi, this script worked perfect for me on Intel HD Graphics 1st Gen with Mavericks using chameleon bootloader.
The brightness levels in decimal (max and min) can be read using Ubuntu (/sys/class/backlight/intel_backlight/).

Thanks for all.
 
Hi, this script worked perfect for me on Intel HD Graphics 1st Gen with Mavericks using chameleon bootloader.
The brightness levels in decimal (max and min) can be read using Ubuntu (/sys/class/backlight/intel_backlight/).

Thanks for all.

OS X does not use the min/max set by bios. It uses its own values. Max 0xad9 for Haswell and 0x710 for prior Intel graphics. It is best just to use the patch at my repo that uses these values and adjusts to them when the system is loaded. Otherwise you have different behavior before vs after sleep. There are ways to patch the frame buffer to match your bios, but it is generally not worth the trouble.
 
If I understand correctly these values in the 12b_4x40s_BCL may differ from display to display.

Code:
Name (_BCL, Package()\n
    {\n
        3487,\n
        930,\n
        0,\n
        183, 188, 193, 198,\n
        203, 208, 213, 218,\n
        223, 228, 238, 256,\n
        274, 292, 310, 328,\n
        356, 384, 412, 440,\n
        468, 503, 542, 580,\n
        618, 657, 698, 744,\n
        791, 837, 883, 930,\n
        980, 1031, 1081, 1132,\n
        1183, 1236, 1290, 1345,\n
        1399, 1453, 1510, 1570,\n
        1630, 1690, 1750, 1810,\n
        1883, 1956, 2029, 2102,\n
        2175, 2265, 2362, 2460,\n
        2557, 2655, 2767, 2911,\n
        3055, 3199, 3343, 3487,\n
    })\n
To be sure they are 100% accurate, is the best way to extract them from R-W Everything?
 
If I understand correctly these values in the 12b_4x40s_BCL may differ from display to display.

Code:
Name (_BCL, Package()\n
    {\n
        3487,\n
        930,\n
        0,\n
        183, 188, 193, 198,\n
        203, 208, 213, 218,\n
        223, 228, 238, 256,\n
        274, 292, 310, 328,\n
        356, 384, 412, 440,\n
        468, 503, 542, 580,\n
        618, 657, 698, 744,\n
        791, 837, 883, 930,\n
        980, 1031, 1081, 1132,\n
        1183, 1236, 1290, 1345,\n
        1399, 1453, 1510, 1570,\n
        1630, 1690, 1750, 1810,\n
        1883, 1956, 2029, 2102,\n
        2175, 2265, 2362, 2460,\n
        2557, 2655, 2767, 2911,\n
        3055, 3199, 3343, 3487,\n
    })\n
To be sure they are 100% accurate, is the best way to extract them from R-W Everything?

There is no issue of "accuracy" here. Specific values could be customized on individual preference.

Also, the range of values is dependent only on the graphics drivers in use. 0-0xad9 (as above) are for Haswell, and 0-0x710 are for Sandy/Ivy.

This solution is already in the ProBook Installer.
 
There is no issue of "accuracy" here. Specific values could be customized on individual preference.

Also, the range of values is dependent only on the graphics drivers in use. 0-0xad9 (as above) are for Haswell, and 0-0x710 are for Sandy/Ivy.

This solution is already in the ProBook Installer.
Thank you. Yes, the Probook CE installer works well for me.
 
Hi Rehabman,

Any idea why I get different brightness ranges based on which ig-platform-id I'm using?
To try to quantify it, let's say when I use 0xa260006, I can go up to 100% brightness. If I switch to 0xa26000a, I can only get up to about 80%, which in turn means every step in between is a little less bright with 000a compared to the same step with 0006.

I'm using the latest pre-built ACPIBacklight (1016), and Laptop-DSDT-Patch / graphics / graphics_PNLF_haswell.txt applied to my graphics SSDT.
 
Hi Rehabman,

Any idea why I get different brightness ranges based on which ig-platform-id I'm using?
To try to quantify it, let's say when I use 0xa260006, I can go up to 100% brightness. If I switch to 0xa26000a, I can only get up to about 80%, which in turn means every step in between is a little less bright with 000a compared to the same step with 0006.

I'm using the latest pre-built ACPIBacklight (1016), and Laptop-DSDT-Patch / graphics / graphics_PNLF_haswell.txt applied to my graphics SSDT.

Check the framebuffer in each. It has the setting for maximum PWM backlight clock.

In the event of something other than 0xad9, you'd have to modify the _BCL table (and other data) to match.

FYI: I'm working on some changes to this patch (and the HD3000/HD4000) where it will instead keep the max PWM clock at what BIOS sets (as an option), scaling the _BCL as required, and always setting the PWM registers (including max) to what the BIOS originally set. I think it is probably better to set it to what BIOS wants instead of what OS X wants. It will also have the side effect of fixing the problem you mention (not really a bug, but...) because the framebuffer's setting will always be short lived.
 
Status
Not open for further replies.
Back
Top