I will look at the manual again. Thank you again for the help

I'm back to say that while I'm still not completely confident while using the program, with many hours of trial and error, I do feel like I am getting better results. I have been able to correct errors that were always eluding me elsewhere. The key to this is knowing your television, or whatever it is that you're needing to calibrate, and secondly, read those graphs. Having multiple graphs really helps me to feel better about my calibration because I can see, even though RGB Balance may look good, we still have EOTF Dif, and clipping to look at, which is wonderful! Now I just need to get into the 3DLut area to fully make this complete for my calibration. The graphs are so important to the process here.

That's great news!
Keep at it, and as before, do ask any questions.

Steve

I do have a few questions. If I push too far with RGB on my TV, a Sony X950G, I do get weird things going on. Changing color in white. I've noticed this in UI's of programs, as well as actual movie watching. I've narrowed it down to going too far out on RGB. My problem comes with me trying to get luminance values where they need to be on 5 and 10%. The RGB values need to be pushed out in order to get luminance right. I have to use High in 2 point in order to ease RGB values at top end for 20 point grayscale, but I can't use low in 2 point because that also causes weirdness to happen, like veins in skin? You can see things like veins poking through skin, it's really odd. If there is any way to help me take the guess work out of that equation, I would be appreciative, because that would help me to get the better calibration. I'm getting good results now, but I'm loosening RGB controls on the low end, trying to ease the numbers down there, but I still fill it's not perfect. My other question has to do with EOTF Dif. I can get RGB all on the line except for 5% and 100%. They start to go off above or below the line in those spots. Is there any way to get those on the line? I've tried doing lower controls to get there, like 10% and 95, 90 percent, but did not succeed fully there.

With RGB Gain/Bias control it all depends on how they have been implemented in the specific TV's CMS.
Usually, Gain is high (whites) and Bias low (black).
And when adjusting Gain the rule of thumb is you reduce the value, not increase, or you will potentially cause clipping.
(And Green is usually not adjusted, other than to lower peak luma.)
When adjusting Bias the rule of thumb is you increase the value, not reduce, or you will potentially cause crushing.

But, with low-lights most displays have a colour tint in the blacks.
That is inherent with most LCD screens for example - a blue bias.
As that is inherent it can't be reduced near black.

This is all explained in the Manual Calibration Guide.

Steve

I read that with the manual, and that I should do low 1st to avoid errors in high, but something is odd with this TV where when I 1st got it, the low was on the line, but dropped after time to below the line. Every time I've tried doing the low end in 2 point, I've gotten very weird results. I've used recommendations elsewhere to only use high in 2 point and that oddness has gone away. The only thing I'm seeing now is if I go too far with RGB controls in 20 point I'll see things like changing colors in white, or sometimes when a camera pans, I'll see changing color in the sky, things like that. The more I try to reduce those numbers, the less noticeable it is.

The key is getting to know what the CMS controls on any given TV actually do...
No two manufacturers have the same CMS controls, and some are better/worse than others.
JVC for example have six axis RGBCMY controls, and any display/projector with six axis controls is an immediate giveaway that the CMS will be bad.
(Secondary CMY colours are always a direct calculation from the primary RGB colours, so should never need their own controls.)

The problem is, there is no single set of rules for manual display adjustment...

Steve

Steve
Yeah, it's just trial and error I guess. I just wish I knew how to gain complete control and know exactly what to shoot for on this thing. I do still have the RGB Separation that I need to check out, but as Ted said, that could be tricky since I have to move contrast all the way to 100 in order for the 20 grayscale to align properly during calibration. The biggest thing that worries me at this time is moving the numbers too high on the low end, or anywhere really, on the 20 point grayscale, and still not knowing exactly how far is too far.

Alright, I wanted to show my results after grayscale. I'm getting terrible results on grayscale after calibrating. I know I'm missing knowledge somewhere to fix this. The grayscale ramp has black streaks in the picture. I'm not sure if you can see it well enough. Also, before calibration the ramp is smooth from black to white. It isn't smooth anymore after calibration. I am unsure how to fix it on the ramp. I know the different controls add or subtract colors throughout the ramp, but I'm not sure of the method used to go in and get those satisfactory results, while also keeping my numbers looking good in the program.

As the graphs show grey scale dE values that are tiny, the issue is rather weird.
I would profile with a greater number of points in the grey scale, to see if there are issues 'between' the presently measured points.
(I fixed the link to your images too - you tried to link to 'pages' not images - the url must end in .png, .jpg, etc... it is easier to use the 'Attach Files' option within this forum too.)

Steve

I am working on grayscale until I can get everything matched up completely on the line from top to bottom. Is there something else I should be trying to do instead of this? The results are horrible, and I've done it so many times now at this point. I sometimes get results that are better than others, but none of my results are anywhere near perfect.

Please do as I suggested.
Measure the grey scale with a LOT more points, to see if there are measurable variations.
The present measurements are JUST the CMS control points, so there may be issues between them you are not seeing.
As the control points have very low dE values, they themselves cannot be improved more.

Steve