Well, I figured I might as well post this in case others find it interesting and/or amusing. First off, as they say, I do not recommend trying this at home. I only did this because the display is an older CRT, was expendable if I screwed up and I had the service manual in front of me. I can't emphasize enough that by doing something like this, you could easily end up with essentially a very large paper weight (and that is a guarantee if you didn't write down all settings before hand).

I have an older 27 inch Toshiba CRT (model 27A32) that after calibration with HCFR and an i1D2 had a horrible red push and blue pull (~+40% for red and ~-20% blue compared to the optimal brightness values given the primaries) in addition to the yellow point being extremely leaning toward red. This set is sort of retired and thus could be expendable, so, I began to wonder whether or not the red push "feature" was simply a configurable option stored in the EEPROM but not settable from any GUI interface (the user and service menus). Since I work in the software/computer field, it seemed logical to assume that the red push (to counter the bluish grayscale) was simply an option programmed in. I've had the service manual for this set since I bought it and I knew there was a 're-initialize the EEPROM' option. So, after double checking that I had written down all service menu variable/value pairs for each input type (luckily the service mode is fairly basic with only a few options being different across input types), I decided to reset. That part worked - the tv took a few seconds and then shut itself off and on reboot, all the settings were back to initial values with the user controls at max. After re-setting everything back to the values I had written down, I then re-calibrated with HCFR and the i1. Here was what I found:

Grayscale was the same - no changes to service menu settings.
Brightness/Contrast were slightly different. My luminance at 100% was decreased by about 10 cd/m^2 and the gamma response at the 5% and 10% points was a bit different but better. Slight changes to the contrast/brightness controls brought everything back to pretty much where it had been before but with gamma a bit better.
Color - my hunch was right. The red push was completely gone and yellow was much closer to it's optimal point. The primaries did not change positions. In fact, the brightness values compared to the optimal targets given the primaries were all +/- 5% and the secondaries were better as well. It only took one or two ticks to optimize both color and tint.

So, what to conclude - I'm not sure other than it seems that it would not have been hard for Toshiba to include a "Movie" mode on this display that turned the red push off and gave more accurate color since it does seem like there is a byte in the EEPROM which says 'red push' on or off and that it was turned on during factory calibration. I should also note that I have no idea what other configurable things got reset back to the real defaults yet are not configurable from the GUI service menu. Given that contrast changed slightly, best guess is that a few other things did get changed.

cheers,


--tom

Interesting...what model Toshiba is it?

Interesting...what model Toshiba is it?

tsanga,

I have edited the post to include the model but was hesitant because I did not want people trying this. I only did it because I didn't particularly care if I messed up since it is a cheap expendable CRT. :)

cheers,


--tom

tsanga,

I have edited the post to include the model but was hesitant because I did not want people trying this. I only did it because I didn't particularly care if I messed up since it is a cheap expendable CRT. :)

cheers,


--tom

I asked because I have one too. :) 24AF45

@thomasl,

how do you measure red push? How deos it sho on the charts?

thanx.

@thomasl,

how do you measure red push? How deos it sho on the charts?

thanx.

mlaun,

you measure the brightness values for the individual colors and then compare them to their target values - which is a percentage of your brightness of the same intensity white reading. For example, a 75% red window should be ~21% as bright as a 75% white window. If you haven't already, read over Kal's calibration guide as well as Tom Huffman's guide (sticky post at the top of this forum). They are both excellent resources to learn about the basics. Kal's guide can be found here:

http://www.avsforum.com/avs-vb/showthread.php?t=1029594

hope this helps,


--tom

you measure the brightness values for the individual colors and then compare them to their target values - which is a percentage of your brightness of the same intensity white reading. For example, a 75% red window should be ~21% as bright as a 100% white window.

I thought the measured Y for red should be ~21% of the measured Y for white, given the same intensity? i.e. either 75% window for both, or 100% for both.

If this is wrong, then I have a major hole in my understanding.

I thought the measured Y for red should be ~21% of the measured Y for white, given the same intensity? i.e. either 75% window for both, or 100% for both.

No, you're not wrong - you've got it exactly right - I had a typo. That's what I get for not re-reading my sentences. I've fixed the original post. Thanks for pointing it out.

cheers,


--tom

thanx thomas,

i do do know about the '21%' on Y. But how is red push going to manifest itself on the reading?

in the HCFR chart thread i posted readings from my ps3 via DVI. One with upscaling the other without. Without upscaling i have visible red push although the cie diagramms are quite similar.

Which difference in the values would explain the red-push.

btw: no mention of red push in the grey scale for dummies thread.


here again my chc files.

i do do know about the '21%' on Y. But how is red push going to manifest itself on the reading?

Mlaun, red push is simply a common term for red being overly bright. So, instead of a 75% red window pattern measuring 21 percent of a 75% white window pattern, it'll measure 25 or 30 percent (or more) while your other colors' measurements are ok. Without dedicated color decoder/brightness controls, the only "fix" is to back off the color control and make a compromise somewhere.

hope this helps,


--tom

thanx. helps me understand.

In my non upscaled data , red is too luminous (bright) at lower saturations with a shift to magenta while blue and green havd a red shift at lower saturations. Guess that explains the visual red push.

Recall posting to an older (but '08) AVS thread here covering red push that I'd added a 1-dB line attenuator (from Parts Express) to my year-2000 CRT RPTV. It's just a packaged RCA-plug-type resistor inserted between the Pr input (YPbPr) cable and Pr input. That trims back red push for some displays, and some years back on another TV-tech board there's a long series of posts on how to fix some Mitsubishi models this way. Haven't tried them, but 2-dB or higher attenuators, or adjustable ones, are available for worse red push cases when service adjustments--the best fix--aren't available.

Noticed red push early with my display, using the Avia test DVD color decoder test, but have always been able to get satisfactory flesh tones with user controls. It's a close call, but decided to keep my 1-dB Pr attenuator in place because it seems to make certain blue hues, rarely observed, less purple--presumably because it's knocking down red slightly. Numerous posts here and elsewhere make the important distinction between grayscale imbalance and color decoder errors. -- John