CurtPalme.com Home Theater Forum

winny · Posted: Sun Apr 29, 2018 3:46 pm Post subject: Poor mans CMS on BG808s?

Thought I would come around to posting a full theater build, but that have to come later.

Long story short, new theater and took parts from two broken 808s and made a working one. Black is freakishly awesome and the white is more than enough, but I’m struggling with that dark gray shades disappear into black so for example Stranger Things or Dark City becomes a bit hard to watch.

Using Linux PC with kodi -> DVI -> DVI to port 3 adapter. 16-235 color space.

The s-models seem to have far more settings than my normal 1209/2, so I’m a bit lost here. Corse convergence will be a future endeavor too, but is there a standard procedure with a test image, gain and offset for each color in the projector to “lift” the dark gray from black?

Thanks in advance!

EDIT: Terrible cell phone photo added.

**kal** · Posted: Sun Apr 29, 2018 4:44 pm Post subject: Re: Poor mans CMS on BG808s?

winny · Posted: Sun Apr 29, 2018 4:47 pm Post subject:

Long term plan in calibration and 3D LUT in kodi, but is there a “cheat” in the meantime to get started?

**kal** · Posted: Sun Apr 29, 2018 5:00 pm Post subject:

I don't really understand what you mean by "cheat" or by just "getting started". You either have gamma correction or you don't. There are many different ways to do this. Depends on your needs/budget/etc. I'm not familiar with doing 3D LUT in Kodi boxes.

Kal

winny · Posted: Sun Apr 29, 2018 5:12 pm Post subject:

Cheat in this regard is tweeting the brightness, contrast, gain and offset (and possibly G2) to get the dark gray visible. Think someone who wants cruise control for his car but only has a brick on the gas pedal. It’s not what you want, but it gets you more cruise control than with no brick.

**kal** · Posted: Sun Apr 29, 2018 5:20 pm Post subject:

I don't know of any cheats.

Kal

winny · Posted: Sun Apr 29, 2018 5:22 pm Post subject:

Let me flip the question around. Suppose you do have perfect gamma, what test image do you display and calibrate to for setting brightness, contrast, gain and offset to their correct values?

**kal** · Posted: Sun Apr 29, 2018 7:16 pm Post subject:

There's no one test pattern.

See: https://www.curtpalme.com/forum_archived/viewtopic.php@t=10457.html

Kal

winny · Posted: Sun Apr 29, 2018 7:27 pm Post subject:

Of course not, but being the authority on calibration, is there any one you would recommend for calibration by eye? Gradient steps?

**kal** · Posted: Sun Apr 29, 2018 7:43 pm Post subject:

No. You can't calibrate by eye. The eye is notoriously bad at this.

A professional who's done this for many years can sometimes get to the point where they've trained their eyes such that they can get in the ballpark, but a meter's always required.

Kal

winny · Posted: Mon Apr 30, 2018 10:48 pm Post subject:

Five hours spent on the theater today. Swapped out the neck boards for Decibels 120 MHz ones and reset everything. Paid more attention to the analog centering and V size (is there an Hsize for R&B somewhere?) and got much better picture.

Spent some time with the gain and offset for each channel and made a compromise to lower color temperature, followed by overall brightness and contrast. Dark gray got better, but still far from perfect. I need to get a sensor and read up on 3D LUTs.

**kal** · Posted: Tue May 01, 2018 12:04 am Post subject:

Before you jump 100% into 3D LUTs, I'd recommend just doing a 2-point calibration with a meter. See how close you get. You may not need a 3D LUT.

Kal

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

Old video standards were made to suit to the characteristics to the CRT displays, so they need very little tayloring actually. This is not like with the modern digital displays, which have to meet the CRT-like characteristics which they don't have.
I'd say the less digital manipulations to the signal the merrier.
As for gamma, there are several specially-breed test pattern to imulate the gamma behavior. As they are, all depending to some extent on display focus and/or bandwidth, the two weak points of CRT, so the results can be very misleading...
Nontheless, here is one such pattern: