CurtPalme.com Home Theater Forum

[Spanky Ham]

David,
Your best bet is to call Craig directly and discuss what you would like him to do. He has a kid on the way or it might already have arrived, so he might be a little tied up.

[stridsvognen]

[stridsvognen]

[David1]

I tweaked the focus and got the rest of the vertical resolution so its super tight now.
The biggest flaw I am seeing in the picture is streaking anybody have any input on that ?
Does a G90 have a streaking problem?

I do have a drablet which I had off for all my tests but turned on does that streak .

Thanks

David

[CIR Engineering]

[CIR Engineering]

[CIR Engineering]

Kurt,

I just looked at the newest Moome IFB cards on my theater G90. I don't see any of the problems that I remember you talking about in this thread. I looked at the Lumagen cross-hatch test pattern and it looks normal to me. I looked at SMTPE 1:1 and it looks pretty decent to me there as well. There might be slightly more attenuation on the newest Moome cards compared to the v2's but it's not terribly attenuated.

Also keep in mind that on the G90 you will always see some attenuation. The attenuation has unusual characteristics though in that the amount of attenuation will not increase much as you go from 1080p 60Hz, 72Hz, and even 96Hz. The attenuation basically remains visibly about the same on the screen even as refresh rate is increased.

Let me know if there were other patterns that you wanted me to look at.

I'm going to try an experiment with the SM test patterns if I can figure it out...

You may also want to have a look at that pattern with your Oppo connected directly to the Moome board and with 12-bit color enabled. This upsampling may provide more data and make the burst look better (though it may not .

craigr

[CIR Engineering]

Basically Kurt there is no industry standard way to upsample the 4:2:0 on the BD to output with HD HDMI. Because of this all results will be different.

I took eight photos of the chroma burst of the SM test disc. However, when I unloaded my camera, I only had seven shots so hopefully these are in the correct order. I think I forgot to shoot the chroma burst with the Oppo BDP-105D in YCbCr 4:2:2.

You can download full resolution copies of the photos in one zip file or look at the down sampled versions here.

Here is the zip file:
http://www.cir-engineering.com/bin/chroma/chroma_burst.zip

So here is what I think each photo is:

Moome IFB-FULLHD V3 & Oppo BDP105D YCbCr 422


Moome IFB-FULLHD V3 & Oppo BDP105D YCbCr 444


Moome IFB-FULLHD V3 & Oppo BDP105D RGB


Moome IFB-FULLHD V3 & DUNE Base 3.0 YCbCr 444


Moome IFB-FULLHD V2 & DUNE Base 3.0 YCbCr 444


Moome IFB-FULLHD V2 & Oppo BDP105D RGB


Moome IFB-FULLHD V2 & Oppo BDP105D YCbCr 444


Frankly, with the projector calibrated to either Moome board I can't tell a difference with real world video between the two Moome boards, except maybe in the rarest of images.

craigr

[stridsvognen]

There is a very nice manual to the Spears & Munsil disc, explaining how the croma multiburst have to look like.

And i expect the Radiance to pass 4:2:2 if its set to 4:2:2 in and out, so i get the same croma result with and without the Radiance.

I sure dont agree that its not visible, sure it depends the display you use, but a good quality display i find it very important.

Can you try shoot a picture of the green tube face while running the croma multiburst, 422 direct out of the oppo, and using the very last moome card.

[CIR Engineering]

Kurt, what do you think you are supposed to see in the green tube with the S&M chroma burst pattern?

craigr

[stridsvognen]

[Spanky Ham]

Here is the full quote from Stacey's website:

[CIR Engineering]

Kurt,

My analysis is below, but just so you know, this is the last time I am going to look at the SM chroma burst pattern for you. The pattern violates Nyquist and is not showing anything of value in terms of real world image quality. The only thing the SM chroma burst pattern is good for is evaluating switches, receivers, signal boosters, or other devices that are NOT supposed to be processing video in order to make sure that they are not processing the signal. Distortion in this pattern does not indicate an error in the Radiance processing nor does it signify any sort of peaking or bandwidth restrictions in the G90 or Moome cards. This also applies to the Moome boards as well because they are converting digital data to analog which will most likely distort the pattern.

The Radiance's job is to process video and as such will distort the pattern. Lumagen is capable of forcing the pattern to look correct with a firmware update, but this improvement in the test pattern will not translate to any improvement in real world video. Essentially, the pattern will be made to look correct as a special case, which will help the test pattern, but not actual image quality.



Looking at the SM chroma burst on my PC would suggest that there are only 18 parts green in the 1:1 part of the chroma burst (about 1 IRE above black). Furthermore, neither the red nor the blue are above even 50% stimulus in the 1:1 portion of the SM chroma pattern and are at higher stimulus in other parts of the test pattern. So in the lower bandwidth parts of the pattern the lines are brighter because they are at a higher stimulus.

On the Oppo 105D directly to the Moome v3 I see perfect resolution looking into the tube on all three tubes for RGB and YCbCr444. For YCbCr422 there is some vertical attenuation in green. This is on Oppo HDMI output 1, on output 2 there may be totally different results as the processing is much different. As some proof that the information is still present in the signal even though it doesn't show in the test pattern, when running through the Radiance with the same configuration I see restoration of the vertical and full attenuation in the horizontal.

Due to there only being three steps above black in the SM pattern for green 1:1, I would not be surprised if after passing through any processing there is no response in green 1:1 on the chroma burst. The Lumagen is there to alter the signal and make corrections so what you put in will not be what you get out. The Radiance is converting all incoming signals to RGB.

Also realize that when you look at the Oppo set to 1080p 60Hz, the Oppo is processing. When you look at the Oppo in source direct 24Hz, then most of the Oppo processing is disabled. So comparing an Oppo outputting 1080p 60Hz direct to the Moome is not the same comparison as looking at the Oppo in source direct going through the Radiance. This is not an apples to apples test.

Here is the Oppo BDP-105D on HDMI output 2 showing 1080p 60Hz on the G90 through the Moome newest v3 batch.


Here is the Oppo BDP-105D on HDMI output 2 showing 1080p 24Hz in source direct mode through the Moome newest v3 batch.


I would suggest spending time evaluating and tweaking areas of the image signal that will actually translate into real world performance improvements. But if you want to continue to harp on the SM chroma burst pattern then have fun on your journey down the rabbit hole.

Regards,

craigr

[Nashou66]

Your right about the 24hz out from an oppo. I forgot to mention in my screen shot the oppo is 1080p Source direct 24hz, and the radiance also is outputting 24hz the TV-one C2-2250 was set to 96hz in my screen shot.

nashou

[Spanky Ham]

[stridsvognen]

I wonder.

We/ i like bandwidth to resolve the luma 1:1, and it for sure makes a huge difference.

When comes to the croma who already is ½ the resolution of whats in the luma channel, we dont care, or its not important to the image if we cant resolve the finest pattern, who must be a 2;2 resolution.

In other words, a format with full color resolution/ a disc with native 4:4:4 will look the same as a disc with 4:2:0.?

For the Radiance, why is it not a helthy goal to have perfect color space conversion, where nothing is lost.?

Im ok with that some dont see a difference no matter how the croma is handled around the chain, and think its a joke to test and find the most perfect path, i guess its a bit the same as when some have a CRT with lousy bandwidth, they also dont seems to share my enthusiasm about full bandwidth to resolve whatever you trow into the CRT.

For now im quite happy how my OPPO 103D HDMI2 out 4:2:2 1080P60hz looks on my Marquee, and i for sure see a difference on movie material changing to HDMI 1 out of the oppo, or putting the Radiance into the chain.

I have looked at different video processors, HDMI and RGBHV out, DVDO VP50 Crystalio 2 Lumagen HDQ Calibre Vantage, Radiance XS. Different players as the Oppo 83/ 103D Denon 4010 and A1UD. And i have 6 moome cards and a Hdfury3

I also use my JVC x7, Sony LCD screen, G90 and my Marquee to compare different kind of performance. That way ill be able to know what part of the chain do what.

And i agree that they dont perform the same, some are perfects, others close, and some far from.

Wonder how low we can put the color resolution before its visible to everyone.

Ill keep hunting for better performance wherever i can find it, and right now the croma is where i focus, and i cant use the G90 to evaluate that, as it will never be able to show any useful difference in movie material, it simply dont have the bandwidth, and it has way to much noise who washes out the image. Just as any other standard CRT i have seen.

[stridsvognen]

[gjaky]

Kurt in your eyes you have about 7 million receptors to percieve colors, and have about 75 to 150 million receptors to percieve brightness, as Craig already mentioned the human eye have much less resolution in color than in brightnes, as you can now see the difference is in the factor of 10!

[stridsvognen]

[gjaky]

You hopelessly drifting towards the world of HTPCs...