Hi all,
I have read many posts on this forum and learned a lot about the theory of calibration (thanks all) and am now playing with my setup to see what I can improve picture wise.

Samsung LE40M86BD LCD TV
HTPC via DVI > HDMI using Widows Media Player for video

HCFR - Latest version (2.01?)
HCFR Calibration DVD
SpyderTV Colorimeter

So far...

I have read the HDTVtest review of the set and noted the likely optimal presets to obtain closest to rec709 color gamut (namely Colour Space set to WIDE and xvYCC ON).

Calibrated the set using SpyderTV software/hardware (colour presets, brightness and contrast only, colour and tint unavailable). Spyder suggested the WARM2 preset (as did the HDTVtest panel).

I ran a greyscale using HCFR and the SpyderTV probe and also ran the primaries to see how they lie on the CIE diagram (set to rec709 in HCFR prefs).

I noticed that the HCFR calibration DVD offers two types of references, full screens (supposedly for projectors) and for TV, etc it has window references (i.e. just a small rectangle in the middle of the screen). For comparison I ran the primaries on both full screen and windowed and compared the results.

Screenshot:

http://i91.photobucket.com/albums/k309/dackjaniels/WindowvsFullScreenMeasures.jpg

The larger white triangle was plotted using the full screen patterns and the 'inner' white triangle was using the windowed patterns. I notice quite a difference in the results, which brings me to my first question...

1. Which types should I be using for my LCD TV? Full screen or windowed?

2. Secondly, is the HCFR calibration DVD of lesser or greater quality than say, DVE? I have DVE but don't recall if it has IRE in increments of 10).

3. Thirdly, and I only just reaised this!, I am using a DVD (in HTPC) to display the reference images yet am using the rec709 HD CIE standard in HCFR. Should I therefore be using a HD source for the test images rather than playing a DVD in Media Player? I don't actually have a HD drive in my HTPC but play a lot of HD files from the HDD (.mkv files). I have download a HD version of the cal disc that a helpful forum member created, should I try to use that instead?

Thanks for taking the time to read this, look forward to your replies :)

Regards,
Steve

1. You should aways use the windowed pattern. It displays a more intensive field and makes it easier for the meter to read what's being displayed.
2. I have both DVE and the HD Calibration disc created by the folks on this forum. The HD Cal disc displays exactly the same grays and colors as DVE. The HD one is preferable, though, since the menu is layed out in a way that is built for HCFR whereas with DVE you have to kind of hunt for the right windows to display.
3. If you're going to calibrate to 709, then you need a device that's going to send a 709 signal to your tv. More than likely the HTPC isn't translating from 601 to 709 when it is sending the information.

I use the HD Cal disc for my calibrations and it works great. I have both the HDDVD and Blu-ray versions. Both work well. I'd suggest one of those if at all possible.

NOTE: What you calibrate for should depend on what you'll be watching on your TV. If you're wanting to calibrate for mostly HD material, then you definitely want to calibrate to 709. However, if you're only watching SD material then you might as well calibrate to 601 and then you can always recalibrate to 709 later when you get some HD material to watch.

Thanks for the informative reply nicholc2.

Just to clarify, if I am able to find a way of playing the HD Cal 'disk' from my HDD through Media Player (in the absence of a HDDVD or BD player in my HTPC) will the signal to my TV then be 709, because the source is HD video?

I have a PS3 for BD playback but enjoy a fair amount of HD video on my HTPC as well so just assumed rec 709 was best for me.

Another quick Q, I have tried different TV presets to get the primaries as close as possible to the CIE reference points. Can they be moved any closer through calibration using user menu controls? Beyond the usual brightness, contrast, tint, etc.I can only see geryscale controls (RGB cuts/gains) and 'My Colour Control' with 4 sliders named Pink, Green, Blue and White.

Thanks in advance :)

Since you have a PS3, download the AVCHD version of the AVS HD calibration disc from this forum:

http://www.avsforum.com/avs-vb/showthread.php?t=948496

You can use that on your PS3 and you'll be calibrating at 709. That's actually what I use. Either that or the HDDVD version depending on which the client has. Both are small enough to be burned to a regular DVD, but play in the format they are supposed to when inserted into the PS3/HDDVD players.

Yes, the grayscale controls are to fine tune the grayscale to get it as close as possible to the reference. Since you have HCFR, play around with it some. (record your defaults first) The cuts affect the low end of the gray scale and the gains affect the high end of the gray scale. They do interact, though, so you'll have to go by trial and error.

The color control sliders I believe affect saturation. In order to test this, put up the 50% saturation pattern from the AVS HD disc and move the slider and take measurements and see if it changes. I've seen mixed reviews on here as to whether this actually changes anything. I'm actually supposed to calibrate one soon, so I'll see for myself.

Hope this helps.

Steve,

A set's primaries are normally fixed unless the display has a CMS. The Samsung that you have most likely does not. I have a big screen Samsung LCD and have found that Color Space Auto and turning xvYCC Off yields the closest points to the 709 spec. Since the primary points are fixed, your measurements are showing you the native color gamut of the set - in theory, this should not change whether you're feeding it a SD (dvd) or a HD signal. Typically though it is best to calibrate using the source that you will use the most. So, if you have a Blu-ray player, using that device is probably best. As for your user controls, color affects the brightness of all the primaries, tint/hue affects the position of your secondaries along the lines that define your color gamut. The white balance controls affect your grayscale tracking and the MyColor Control doesn't seem to do much but results may be different depending on the Samsung set in question as well as the input being used. I would leave them alone (especially white which just alters your grayscale across the entire intensity range) but once you've gotten the rest of your settings optimized, you can always experiment with the test patterns to see what blue and green do on your set.

If you haven't read through it yet, I'd take a look at Tom Huffman's excellent calibration howto thread. It is a sticky thread at the top.

hope this helps,


--tom

Thankyou both for your help. I have downloaded the AVCHD Cal disk and will try it soon, just trying to get the HTPC input as close as I can now that I've started, then I'll move on to the PS3.

I also read Tom's guide and downloaded his cal DVD image to compare results against the HCFR disk, using window patterns on both. I noticed what I think is a significant difference and suspect it is due to the different window size used. Any comments?
Tom's disk has a window size of approximately one third screen width whereas the HCFR window is barely large enough to cover the sensor on my SpyderTV. On my 40" LCD the HCFR windows are about two inches high by 3.5" wide! Is this as it should be?

Anyway, thanks again. About to try my third greyscale run and will post a few results later on, no doubt with a few questions ;)

Regards,
Steve

No, that isn't right. It should be larger than that. Use the AVCHD disk and you'll be good to go. I would use the PS3 first and then doublecheck with the other if you wish. But from my experience, the PS3 is spot on.

Thanks for comfirming that, it did seem a rather small window to me :)

Still with HTPC at the mo' but and just tried altering settings for Colour Space and xvYCC as per Tom's suggestions.

Screenshot below shows CIE with 4 different combinations of settings.

http://i91.photobucket.com/albums/k309/dackjaniels/CIEComparison.jpg

Top right are Tom's suggested settings and look the best to me.
Bottom right are my original choices and secondaries look closer but I assume this is because I have calibrated colour and tint with these settings (using SpyderTV)?

What do you think? Go with Tom's settings and recal basic settings and run greyscale again?

Incidentally, I performed the above measurements with the suspect HCFR disc so will post another screenshot using the Cal disc from Tom Huffmans thread...

Here it is...

http://i91.photobucket.com/albums/k309/dackjaniels/CIEComparisonTOMSDVD.jpg

Seems readings using Tom's disc are consistently different and I assume more accurate so will go with that for HTPC and use the AVCHD disc for PS3.

Thanks,
Steve

Steve, the upper right looks the most accurate to me as far as the primaries color points go. It looks fairly close to my Samsung. But the secondaries do look a bit odd especially yellow. One thing about the secondaries is that if you haven't optimized your white point i.e. gray scale then your secondaries will shift. So, I would definitely do your grayscale/gamma optimization before going back and re-doing the color/tint controls in order to get your secondaries better positioned. On my Samsung, Movie mode with the Warm2 color temperature setting was the closest to D65. But also keep in mind that brightness/contrast will affect your grayscale tracking as well. It is essentially an iterative dance back and forth as you find the right set of settings so you get good grayscale tracking/gamma as well as an image that is bright enough for you in your viewing conditions.

hope this helps,


--tom

Thanks Tom. Will try your suggestions out and let you know how I get on.

Steve

I actually calibrated a Samsung LCD last night (LN-T4661). Turned out great. This is the second one of these I have done. If I were in the market for an LCD, I'd definitely look into getting one. :-)

Here are the results:

http://i233.photobucket.com/albums/ee111/lishbabe13/HT%20Pics/LN-T4661Calibration.jpg

Looks good to me! :)

Wish I could get my gamma curve that flat, not sure its possible on my set though.

I basically restarted my calibration from defaults using the settings you suggested.
After the first greyscale run it (the greyscale) looks quite good though not yet perfect. The gamma however, seems very low and I'm not sure how to raise it up. Any suggestions?

Also what total light output would you aim for in a light controlled room? (Y value of 100% Grey). I've read that around 100cd/m2 (35 ft/l) is about right so I guess mine is a little low at 70 odd cd/m2. It was actually spot on with backlight on 2 (or maybe 3) but I lowered it to 1 hoping to increase the low gamma (not sure if this was correct thinking on my part).

Would appreciate any comments you have and am glad of your ongoing support ;)

I posted the details here as I believe its the correct thread to post calibrations in...

http://www.avsforum.com/avs-vb/showthread.php?p=13501291&posted=1#post13501291

Thanks,
Steve

The lower contrast could be what's causing your lower gamma too. With his TV, I started out by getting the gray as close as possible, going back and adjusting brightness and contrast (brightness with test pattern, contrast to 41ft/l). Doing this I still had issues with the gamma curve. His LCD actually has a gamma setting which I utilized to get it close, then I adjusted the brightness and contrast to get it where it ended up. Remember gamma is just another way of expressing the luminance. So brightness and contrast directly affect it.

I actually had to compromise on the brightness end. Raising the brightness anymore took the gamma too low, but lowering it made the 10% and 20% too high, so I had to find the sweet spot and get it to where you see above.

If your set has a gamma setting, try changing that and see what you get. Also, try 41ft/l instead of 35. Anywhere from 35-45 is ok IMO. His actually ended up being around 45 or so if I remember correctly, but that was ok because it made the gamma flat. Plus his set is in a room that has a lot of windows, so a little brighter isn't going to hurt. :-) In the end, you want to get the gamma as flat as possible, so if that means you have to have the contrast a little higher sobeit. I wouldn't go lower than 35, though, as you'll lose alot of detail in the picture if you do.

Thanks for the advice nicholc2. If I understand you correctly I should try raising the contrast control until 100% grey has a Y of 41ft/l and use a pattern to set brightness. At present my contrast control is on 96 of 100 and backlight is on 2 in a zero to ten range. Raising contrast to 100 will not make much difference so the only way I can increase light output and achieve 41ft/l is to raise the backlight a notch or so. Is this correct? I've been trying to keep the backlight as low as possible as my set is prone to clouding.

Steve

You should be able to compensate for that some with the brightness control. I found on the one I did that had the same affect initially, but when I turned up the backlight, I turned down brightness and it was good. Like most things, though, you'll probably have to come to a compromise with it. No set it perfect and some limitations just have to be worked around. So long as your gamma is tracking good, you might have to go with a lower 100% IRE out of necessity. Do you sacrifice blacks or do you sacrifice whites? That's sometimes a hard call and one in the end that will have to be up to you. Hopefully with yours you can increase the backlight and then lower the brightness and it will be ok. :-)

Great thanks nicholc2, will let you know how I get on :)