Official JVC RS20 / HD750 Calibration and CMS thread (NEW FIRMWARE V1.1) - Page 29

Thanks
The Ramps are only on the HD-DVD DVE disk --not on the Blue Ray Basics. I am supprised that you do not see Color ramp compression at, say 8.

KT

Yes. This is just a confusion in terminology. Rec. 601 refers to a color decoding matrix used to convert back and forth between YCbCr and RGB that is used in SMPTE 170M and EBU, but it does not officially specify any particular gamut/white point. However, the SMPTE-C gamut/white point is implicit.

It is my understanding that REC 601 defines the color encoding and decoding matrix for Std Definition (along with REC 709 for HD).

The Color Gaumet used for Std Definition and HD is a different matter ..with REC 709 specified for HD and SMPTE-C for Std Definition.

So, assuming I have it right, the correct color Gaumet for HD is SMPTE-C. That said, I think many (including me - until recently) ignore this separation between encode/decode and Color Gaumet and refer to the SMPTE-C Gaumet as REC 601.

Edit--I just now see Tom has answered your question - i will defer to his answer.

KT

Manni.
I have tried both of your settings with quite good success. The color behavior is the best I have seen on my particular RS20 - -so far - of settings shared here on this forum. The flesh tones are the best I have seen and compared to THX (my only reference) your new settings are the only ones that duplicate the darkness/lightness of various levels of Blue (all other settings darken intense blue colors in the source -for example the intense blues in the background of the new Tonight show).

However I have a problem that is common to both of your settings and that of Ricks recent post #809 involving weak color levels with the color control set at 0. As it is the only method available to me -- and knowing now that the color filter method for setting color/tint and judging color decoders is in error -- I have no choice but to use DVE and the old make -do method of setting color level using the blue filter and Color Bars.

Here is the problem: Using the above method I now am getting very high color level settings ----- 15 for your REC709 and 20 for your SMPTE-C color Gaumet setups (and 20 for Ricks ISF-type setup). This is very unusual. The consequences of these settings is that the color ramps (HD-DVD DVE -720p section) are going significantly into saturation at the brightest end of the ramps. I can correct this by dropping the Contrast from 8 to 0. But this cuts the overall brightness of the image quite a bit. (edit note--I find that the color is probably too high at 15-20 and better at 0 on real program sources -- suggesting that the filters are wrong.)

So something is wrong here -- most likely it is the use of the color filters for setting the color level with the RS20.
To find out if this is the case - I wonder if you could run a test on your projector for me using your settings for REC 709:
Use your DVE basics HD disk and go to the NTSC color bars (or just use any NTSC color bars) and using the DVE Basics blue filter --set the color to 0 and see if the filters show the color level is too low - and then see if raising the color level to 15-20 then looks correct through the filter. (We are just comparing our 2 projector behavior here--not testing the validity of the use of the blue filters)

If the filters behave on your system as they do on mine--my problem is in using the filters to set the color level. I will drop that method and my problem goes away. But if you still find the color level is correct at about 0 and way off at 15 -20 there has to be something wrong or different with my projector or setup. If you can find time to do this it would resolve my main problem area, and would be very much appreciated.

Thanks
KT

Yes it does indeed help.
But we need to clarify the description of behavior of the "white" bit at the bottom of the blue bar. That blue bar has beneath it a segment of white. Viewed through the blue filter -when the color level is at 0 that segment of white should be a brighter blue (rather than darker) than the blue bar on top. We both should see that small bottom segment darken and match the blue bar above as we go to color levels of 15-20. Does this clarification sound correct?

So -if so - I can now say that the color compression issues I have seen are explained: 1. If I raise the color to 15 incorrectly -I am pressing the bright end of the color ramps into saturation. But I can eliminate the color saturation by lowering my Enhanced input contrast level to the incorrect setting of 0.

2. Or - If on the either hand I set the contrast to 8 correctly, and then drop the color level to 0 - I can eliminate the saturation problem. The latter case appears to be the correct one and I will ignore the filter test that says the color is too low at 0.

Later, I will get back to you on the relative performance of your 2 setting with new HD Movie sources. Initially the differences between your 2 settings are not as much as I expected.

Also, could you indicate what meter and software you used to develop your 2 settings. I have just received my Eye-one D2 meter and Calman software for the beginning of some calibration attempts here.

Thanks

KT

OK, Thanks for the clarification.

The D2 is as far as I want to go into calibration attempts. I realize the results may be hit or miss but it should at least result in a better Gray scale than has resulted doing it all by eye (over 750 hours now on lamp). In any case - I consider this all fun stuff to experiment with.

Again, thanks for your help,

KT

Hmm. Some interesting tweaks suggested in the last two pages. To confirm the latest thinking of how to setup the projector:
- Use Standard HDMI setting on the source and Enhanced on the projector, and adjust the contrast/brightness to suit. What advantage is there to doing this - isn't it better to use the settings correctly?
- Use SMPTE-C (rec 601 SD gamut) for most HD disks because most HD disks are mastered in rec 601 except for the odd disk like Transformers that uses the proper rec 709 HD gamut? Again this does not sound right??
- Recommended sharpness settings of 10 and detail enhancement of 35.

Manny, no sarcasm intended, I was just looking to confirm the impressions I was getting on the thread. I was posing questions not looking to 'trash'.

Thanks for the details about enhanced - that makes sense now.

SD REC601 is interesting. Is it the case that most HD/Blu-rays are mastered to REC601? Is it only because the monitors used for the mastering are SMPTE-C?

I like the higher sharpening/detail enhancement. I have a large screen so a bit more sharpening does help the image and it seems to work pretty well on the RS20 by not introducing visible artifacts. I can see a slight amount of ringing with 10/25 with a test pattern but it is not visible on film material, and it adds to the crispness of the image + more 3D depth.

I agree with you that skin tones with rec709 are too hot, I often find lips and cheeks too pink on the RS20 and never had that with my previous projector Canon SX50 which had accurate colours. I watched Ronin last night with your rec601 settings and could not tell a difference when switching back & forward between your 601 and 709 settings, but I also did not see the hot cheek effect either.

Keep up the experimenting!

Manni

I am interested in your observations about using SMPTE C. I have seen posts from well respected contributers that support what you say

As you are in the UK are you testing with PAL or NTSC discs? The Pal gamut is very similar to HD so if you are testing with PAL discs this would imply that they were also mastered incorrectly!

[quote=Manni01;16687364]@deanbob: sorry for the slight overreaction
As far as I understand, it's more a case of most blu-rays being still mastered to rec-709 with colorists using old monitors calibrated to smpte-c. The situation is clear as mud, some saying that colorists are able to interpolate what they see on these monitors to make sure it produces rec-709 material, some saying it isn't necessarily the case. The only thing I know is that I am struggling to find blu-rays with skin tones that look better with the rec709 calibration than with the smpte-c calibration...

What you are saying about screen size re sharpness/detail enhancement is very interesting and makes complete sense. I have never seen an improvement (just, like you, some ringing on patterns, not on real material), probably because my screen is "only" 88" diag (16/9). I understand now why some here, probably with a bigger screen, can see an improvement when they pump up these settings.]

Hi Manni
I would like to make 2 points about the question of sharpness settings.

1. Often I see the concept stated that the sharpness settings and perceived sharpness of images is related to screen size.

With the exception of perhaps the grain size of the elements in the sprays applied to the screen- the screen size should have no effect on the perceived resolution or the settings for image sharpness. Assuming the viewers are seated so that the same included visual angle on the retina remains constant the only thing that really changes with screen size is how far away one sits from the screen. So I do not believe screen size has anything to do with the various sharpness setting preferences (but - the included visual angle will have a great deal to do with these preferences).

Having said that I use an 76 inch wide Stewart Studiotech 130 screen and sit with the eye at about 9.7 to 10 feet away from the screen. I am one of those who likes my images sharp, and so as I have posted before I like the sharpness of the highest quality images at about 35 and the detail the same. At these levels, on the best sources, I do not see any outlining (if I do I will back the sharpness down) and I will not see any ringing simply because in my tests the sharpness control on the RS 20 at any level does not introduce any ringing of the video signal.

This I think - is all a personal preference issue and I realize I am in the minority here. For example - we know GRogers who obviously has an excellent eye -- prefers setting of 0,0 which probably supports the general preferences on this thread.

2. The key to sharp images I believe is to focus on the pixels themselves (which I am sure you are doing) and not on images. I think it is possible that some folks are setting their focus on just images --and if so they are they will not achieve a high level of focus. If this is so, then it seems likely these viewers will not be able to see much difference between settings of 0 and say 35 , and so may end up choosing the safer low settings near 0 because they can not see any increase in sharpness at the higher settings.

It is critical to focus on the sharpness of the individual pixels and not on the sharpness of images.

KT

The suggestion about the red focus was for the RS2. It shouldn't be necessary on the RS20, which has much better color correction in the lens. To get a red (green or blue) grid, you can use an AccuPel HDG-4000 and turn on only the color you want (shameless plug ), or use a white line grid and then temporarily use the RSx convergence adjustment to offset the red, green, and blue colors a couple of pixels for focusing purposes.

I don't agree with that.
If the lens focuses on red, out of focus and fringing will affect green and blue.
The best is to balance in the middle of the spectrum: green.

Focusing on green pattern (or image) is not very acurate.
Focusing on red is difficult (blurr feeling)
Focusing on blue is quite impossible (similar to night vision)

The best way is to use a B&W image.
Our vision get 3 times more details in a B&W image compared to the same image in color.
The visual sensitivity for focusing is improved.
Focus, fringing and convergence errors are balanced in one go.

This suggestion is also valid for setting the sharpness...(hot subject)

Here is the recommended method for setting sharpness of an image in Photoshop:
- enlarge the image to the desired size (effective pixels)
- switch from RGB to Lab color system
- turn off a and b chanels
- The L chanel displays a B&W image with the maximum of spacial information
- adjust sharpness to your taste
- switch back to RGB
- save the image
note: the sharpness correction should be done at the very final stage.

notes:
- iris full open will reduce depth of field and increase focus setting precision
- zooming does not keep the focus

I could not tell the difference when switching between rec709 and rec601 although I was going back and forward between the two colour settings in the advanced colour setting menu option and was thinking maybe the colours don't get changed until saving the settings (or putting two colour settings in different user settings and going back & forward between user setting on the remote). However the only colour issues I have is the occassional pink cheeks or lips, everything else looks OK to me colour wise (like you I use skin tones to tell if there is a colour issue). I will try watching another movie with your rec601 settings again, because the pink lips effect is off putting - in one movie I actually had orange lips but I think that might have been the source.

Regarding sharpness, I have a 135'' (3.5m) wide screen and sit about 11 feet from the screen, which is a little closer than usually recommended. This gives me the big screen effect which to me is the aim of a home theatre, but it does mean that source quality and sharpness become more important. A blu-ray source with a higher sharpness setting easily beats the local cineplex for picture quality.

Another trick for focusing correctly is to use a PC (I use a HTPC so this is easy). I throw up a text file with a screen full of 'a' characters and find that very useful for focusing the screen, better than the usual video calibration disks. I find my RS20 has consistent focus across the screen with minimal chromatic abberations, basically the slight focus differences and chromatic errors are not noticable at my viewing distance for PC text, so have no impact on video sources which don't need the same clarity as PC text. Also if you have slightly inconsistent focus in different areas of the screen, especially the edges (which is normal for lenses in this price range), get the focus in the middle of the screen correct.

Dang, if that sucker ever falls over, you'll all be crushed to death !
You must live in the Houston Astrodome.

Something got lost in translation when I translated metric to imperial for you folks Stateside, the rest of the world uses metric so I thought I'd use the old world measurements but slipped up.

Gregr, you wrote wrote in your review that it is possible to adjust the focus using a pattern other than the internal green pattern. Can you please tell me how to do this with an externally generated pattern?

A second question regarding gamma. What reference equation do you use - ITU or power function?

Thanks

I just tried your new settings for REC 709 and SMPTE-C and thought I would give you some immediate feedback.

On my projector I am seeing an obvious improvement in flesh tones - more natural. These are the most natural to date.
And I find that there is now a considerable difference between the 2 settings.

I do indeed find all the movies (limited to DirecTV so far), as suggested by GregR, look much better with the SMPTE-C settings (I find REC 709 has too much color and is a bit red/pink on flesh tones). HDTV shows such as Conan O'Brion are not obviously better with REC 709 relative to SMPTE-C but are better than both the earlier versions .
Overall (for my projector), and especially with SMPTE-C, very nice improvements .

KT

Manni,
What is the reason you ended up with a new set of settings (for example, you ended up with significant changes in magenta for SMPTE-C ? What happened that gave you that result on this last calibration?

Were you just trying a recalibration attempt, or were you kind of forcing some things to achieve some intended result - or was it the result of a different meter technique...etc..?

KT