AVS HD 709 - Blu-ray & MP4 Calibration - Page 2 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #31 of 4106 Old 12-04-2007, 07:39 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Fringing at sharp b/w edges is the typical sign of an overly aggressive sharpness setting. The pattern from avia pro in this post would be nice, it'll show de-interlace artifacts as well.
zoyd is online now  
Sponsored Links
Advertisement
 
post #32 of 4106 Old 12-04-2007, 07:55 AM - Thread Starter
AVS Special Member
 
alluringreality's Avatar
 
Join Date: Jul 2006
Posts: 3,204
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 40
I take it fringing is what I was calling edge enhancement halos. Is that correct or is there more to it?

Can you explain, what do the mhz markings mean on the pattern you linked?

As far as I'm concerned this disk is rather simple so long as we don't mess up the levels and a few other things in going from image to video. Areas like de-interlace artifacts though, I have no idea, so how would that show up in that image?
alluringreality is offline  
post #33 of 4106 Old 12-04-2007, 11:13 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by alluringreality View Post

I take it fringing is what I was calling edge enhancement halos. Is that correct or is there more to it?

Can you explain, what do the mhz markings mean on the pattern you linked?

As far as I'm concerned this disk is rather simple so long as we don't mess up the levels and a few other things in going from image to video. Areas like de-interlace artifacts though, I have no idea, so how would that show up in that image?

yes, edge enhancement produces halos, fringing, ringing, whatever you want to call it near sharp contrast edges. You can use a high spatial frequency pattern to test for that and display resolution at the same time. The frequency burst patterns do this or you can have just one multiburst pattern, like a bulls-eye, that will test vertical and horizontal resolution at the same time. The 6.75 MHz label on that image only applies if you are testing a 480i signal path. You would just change the label to match the appropriate supported frequency for your HD modes.
zoyd is online now  
post #34 of 4106 Old 12-04-2007, 11:34 AM
AVS Special Member
 
jvincent's Avatar
 
Join Date: Jan 2003
Location: Ottawa, Canada
Posts: 3,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by hwjohn View Post
I don't know how much we can really do about that except double check the pattern, and it has already been triple checked for the most part.

Could you quantify how far off it is? That may at least give users an idea of how far off the filter based method is (I would guess it would also vary from filter to filter). If it is off less than 5% or so, I would call that pretty darn good.
I don't think there's a problem with the pattern. Here's an more detailed explanation of my setup.

I'm using the filters that come with the original AVIA disk.

I have my tint control set to 32, the mid point, and set the saturation using the blue filter.

Looking at the CIE chart in HCFR I have the primary and secondary points as close to the references as possible. Since none of them actually fall on the primaries, an inherent limitation of the plasma's gamut, I have adjusted the primaries/secondaries so that the lines extending through the measured points will hit the reference points, if that makes sense.

When I use the filters blue is essentially dead on.

To get the green bars to match I have to move tint by about 7 clicks and for red it is about 9 clicks. Total range on tint is 0 to 64.

How much of this is D2 error vs filter error, I have no idea. I do know that the picture looks better than ever. I've watched a couple of HD-DVDs since calibrating and all I can say is WOW.

I've attached my HCFR data if it helps.

 

XboxHD.zip 2.0830078125k . file
Attached Files
File Type: zip XboxHD.zip (2.1 KB, 43 views)
jvincent is offline  
post #35 of 4106 Old 12-04-2007, 11:38 AM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

yes, edge enhancement produces halos, fringing, ringing, whatever you want to call it near sharp contrast edges. You can use a high spatial frequency pattern to test for that and display resolution at the same time. The frequency burst patterns do this or you can have just one multiburst pattern, like a bulls-eye, that will test vertical and horizontal resolution at the same time. The 6.75 MHz label on that image only applies if you are testing a 480i signal path. You would just change the label to match the appropriate supported frequency for your HD modes.

Zoyd,
This is definetely a stupid question... but what does 6.75 MHz refer to? I'm assuming it is something to do with the bandwidth required to display that pattern accurately, but that is the one thing I never understood in test patterns.

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #36 of 4106 Old 12-04-2007, 11:51 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by hwjohn View Post

Zoyd,
This is definetely a stupid question... but what does 6.75 MHz refer to? I'm assuming it is something to do with the bandwidth required to display that pattern accurately, but that is the one thing I never understood in test patterns.

Right about the bandwidth, 480i signal paths should nominally support a frequency of 525 lines/image at 30 frames/sec (actually it's 262.5 lines/image at 60 fields/sec)=6.4392 MHz If there is a low-pass filter or other distortion in the path the image will blur, that's why you typically see mutliple images at various frequencies you can estimate where the roll-off is.
zoyd is online now  
post #37 of 4106 Old 12-04-2007, 11:54 AM
AVS Special Member
 
jvincent's Avatar
 
Join Date: Jan 2003
Location: Ottawa, Canada
Posts: 3,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by hwjohn View Post

Zoyd,
This is definetely a stupid question... but what does 6.75 MHz refer to? I'm assuming it is something to do with the bandwidth required to display that pattern accurately, but that is the one thing I never understood in test patterns.

The 6.75MHz is essentially a square wave pattern, i.e. alternating black and white lines of single pixel width for SD-DVD.

It's a throwback to analog TVs where the actual analog bandwidth of the TV/player wasn't always enough to fully resolve the pattern.

EDIT: zoyd beat me to it.
jvincent is offline  
post #38 of 4106 Old 12-04-2007, 12:05 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by jvincent View Post

I have my tint control set to 32, the mid point, and set the saturation using the blue filter.

If the primaries/secondaries were done with 100% patterns your blue luminance is very close to spec (only -4.5%) so it looks like the blue filter is doing a good job in setting luminance but due to the chromiticity error blue dE=22. Red also has good luminance (+5.4%) but dE for red is 20., again because of the chromiticity error. green is low (-27.6%) but has a fairly good dE=11.1, so you've compensated for green oversaturation with a lower than spec luminance.
zoyd is online now  
post #39 of 4106 Old 12-04-2007, 12:06 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by jvincent View Post

The 6.75MHz is essentially a square wave pattern, i.e. alternating black and white lines of single pixel width for SD-DVD.

It's a throwback to analog TVs where the actual analog bandwidth of the TV/player wasn't always enough to fully resolve the pattern.

EDIT: zoyd beat me to it.

Ok, so black and white lines that are two pixels wide each would require half the bandwidth, and would be labeled 6.75/2?

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #40 of 4106 Old 12-04-2007, 12:09 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by jvincent View Post

I don't think there's a problem with the pattern. Here's an more detailed explanation of my setup.

I'm using the filters that come with the original AVIA disk.

I have my tint control set to 32, the mid point, and set the saturation using the blue filter.

Looking at the CIE chart in HCFR I have the primary and secondary points as close to the references as possible. Since none of them actually fall on the primaries, an inherent limitation of the plasma's gamut, I have adjusted the primaries/secondaries so that the lines extending through the measured points will hit the reference points, if that makes sense.

When I use the filters blue is essentially dead on.

To get the green bars to match I have to move tint by about 7 clicks and for red it is about 9 clicks. Total range on tint is 0 to 64.

How much of this is D2 error vs filter error, I have no idea. I do know that the picture looks better than ever. I've watched a couple of HD-DVDs since calibrating and all I can say is WOW.

I've attached my HCFR data if it helps.

What functions does your CMS allow for? Adjust primaries? Adjust color decoding for each color individually?

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #41 of 4106 Old 12-04-2007, 12:13 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by hwjohn View Post

Ok, so black and white lines that are two pixels wide each would require half the bandwidth, and would be labeled 6.75/2?

no, same bandwidth, whether the pixel (or voltage) level is black or white doesn't make any difference, it still has to be sampled. Don't confuse spatial frequency at the display with sample frequency.
zoyd is online now  
post #42 of 4106 Old 12-04-2007, 12:20 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

no, same bandwidth, whether the pixel (or voltage) level is black or white doesn't make any difference, it still has to be sampled. Don't confuse spatial frequency at the display with sample frequency.

If there is no correlation between spatial frequency at the display and the bandwidth, then how do we use test patterns with different spatial frequencies to determine where the bandwidth rolls off? I must be very confused on something.

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #43 of 4106 Old 12-04-2007, 12:30 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by hwjohn View Post

If there is no correlation between spatial frequency at the display and the bandwidth, then how do we use test patterns with different spatial frequencies to determine where the bandwidth rolls off? I must be very confused on something.

There is a correlation, it's just not in the way you were thinking. A test pattern with just separated line pairs will produce a 6.44 MHz signal after encoding and being sent through the signal path, and be just resolved on the display in 480i mode. Likewise a just separated line pair will produce a 12.88 MHz signal through the signal path and be just resolved on the display in 480p mode, etc.
zoyd is online now  
post #44 of 4106 Old 12-04-2007, 12:36 PM
AVS Special Member
 
jvincent's Avatar
 
Join Date: Jan 2003
Location: Ottawa, Canada
Posts: 3,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by zoyd View Post

If the primaries/secondaries were done with 100% patterns your blue luminance is very close to spec (only -4.5%) so it looks like the blue filter is doing a good job in setting luminance but due to the chromiticity error blue dE=22. Red also has good luminance (+5.4%) but dE for red is 20., again because of the chromiticity error. green is low (-27.6%) but has a fairly good dE=11.1, so you've compensated for green oversaturation with a lower than spec luminance.

Now that you mention it I can't remember whether I save the measurement with the 75% or 100% pattern. When I was going through the disc I tried both.
jvincent is offline  
post #45 of 4106 Old 12-04-2007, 12:36 PM
AVS Special Member
 
dr1394's Avatar
 
Join Date: May 2002
Location: Mizar 5
Posts: 3,184
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 22
The correct equation is:

NTSC digital video

(1000/1001 * 30 frames/sec) * 525 lines/frame * 858 samples/line = 13,500,000 samples/second

or

PAL digital video

25 frames/sec * 625 lines/frame * 864 samples/line = 13,500,000 samples/sec

By adjusting the size of the horizontal blanking, it's possible to have a common sampling rate for PAL and NTSC.

6.75 MHz is the frequency of the square wave formed by a black sample followed by a white sample.

For HD the sample rate is either 74.25 MHz or 1.001/74.25 MHz (74.176 MHz). Again, the horizontal blanking is adjusted to have a common sample rate.

30 frames/second * 1125 lines/frame * 2200 samples/line = 74,250,000 samples/second
25 frames/second * 1125 lines/frame * 2640 samples/line = 74,250,000 samples/second
24 frames/second * 1125 lines/frame * 2750 samples/line = 74,250,000 samples/second

60 frames/second * 750 lines/frame * 1650 samples/line = 74,250,000 samples/second
50 frames/second * 750 lines/frame * 1980 samples/line = 74,250,000 samples/second
30 frames/second * 750 lines/frame * 3300 samples/line = 74,250,000 samples/second
25 frames/second * 750 lines/frame * 3960 samples/line = 74,250,000 samples/second
24 frames/second * 750 lines/frame * 4125 samples/line = 74,250,000 samples/second

So the maximum frequency for HD is 37.125 MHz.

Ron

HD MPEG-2 Test Patterns http://www.w6rz.net
dr1394 is offline  
post #46 of 4106 Old 12-04-2007, 12:36 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

There is a correlation, it's just not in the way you were thinking. A test pattern with just separated line pairs will produce a 6.44 MHz signal after encoding and being sent through the signal path, and be just resolved on the display in 480i mode. Likewise a just separated line pair will produce a 12.88 MHz signal through the signal path and be just resolved on the display in 480p mode, etc.

Still confused. What is the difference between
Quote:


A test pattern with just separated line pairs

and
Quote:


Likewise a just separated line pair

.

Is there an example somewhere that shows how to calculate the required bandwidth for a certain resolution? Maybe that would help me.

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #47 of 4106 Old 12-04-2007, 12:41 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by hwjohn View Post

Still confused.

See Ron's calcs (I was using simple guestimates). There is (hopefully) no difference in the original test pattern and what gets displayed on your screen (isn't that the whole point of calibration). The difference is only the rate at which the information is sent from one end to the other, the higher the information rate the higher your (potential) resolution will be. I was also not using quite the right terminology, for 480i the sampling frequency required is 13.5 MHz while the nyquist frequency is 6.75 MHz. The bandwidth is usually kept somewhat below the nyquist frequency to avoid distortion near nyquist.
zoyd is online now  
post #48 of 4106 Old 12-04-2007, 12:42 PM
AVS Special Member
 
jvincent's Avatar
 
Join Date: Jan 2003
Location: Ottawa, Canada
Posts: 3,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by hwjohn View Post

What functions does your CMS allow for? Adjust primaries? Adjust color decoding for each color individually?

I can adjust both primaries and secondaries.

The controls allow me to move all six of them "along the triangle" as it were.

In other words, the red control allows me to bias it towards yellow or magenta. Similarly the cyan allows me to bias towards blue or green and so on for each primary or secondary. Adjustable range is 0 to 61 with 32 being the mid-point.

The CMS on the NEC is the same as on the Pioneer Elite plasmas (since they are essentially the same unit) and as like the Elites they it is available in the user menu and is programmable for each input.
jvincent is offline  
post #49 of 4106 Old 12-04-2007, 12:46 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
Quote:
Originally Posted by jvincent View Post

Now that you mention it I can't remember whether I save the measurement with the 75% or 100% pattern. When I was going through the disc I tried both.

Probably 100% because if I use 75% your dE numbers go sky high and I'm sure you would have noticed it.
zoyd is online now  
post #50 of 4106 Old 12-04-2007, 01:22 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by dr1394 View Post

The correct equation is:

NTSC digital video

(1000/1001 * 30 frames/sec) * 525 lines/frame * 858 samples/line = 13,500,000 samples/second

or

PAL digital video

25 frames/sec * 625 lines/frame * 864 samples/line = 13,500,000 samples/sec

By adjusting the size of the horizontal blanking, it's possible to have a common sampling rate for PAL and NTSC.

6.75 MHz is the frequency of the square wave formed by a black sample followed by a white sample.

For HD the sample rate is either 74.25 MHz or 1.001/74.25 MHz (74.176 MHz). Again, the horizontal blanking is adjusted to have a common sample rate.

30 frames/second * 1125 lines/frame * 2200 samples/line = 74,250,000 samples/second
25 frames/second * 1125 lines/frame * 2640 samples/line = 74,250,000 samples/second
24 frames/second * 1125 lines/frame * 2750 samples/line = 74,250,000 samples/second

60 frames/second * 750 lines/frame * 1650 samples/line = 74,250,000 samples/second
50 frames/second * 750 lines/frame * 1980 samples/line = 74,250,000 samples/second
30 frames/second * 750 lines/frame * 3300 samples/line = 74,250,000 samples/second
25 frames/second * 750 lines/frame * 3960 samples/line = 74,250,000 samples/second
24 frames/second * 750 lines/frame * 4125 samples/line = 74,250,000 samples/second

So the maximum frequency for HD is 37.125 MHz.

Ron

Thanks for the explanation Ron. I have never thought about this, so I'm still trying to digest it.

As the frame rate and lines/frame become slower/lower, the samples/line increase. Are there any advantages to this?

I am still not understanding how a vertical pattern consisting of a 2 pixel white line followed by a 2 pixel black line doesn't half the bandwidth needed to represent it. It is my understanding that a pattern like this (let each letter represent one pixel, W for white, B for black):

WB
WB
WB
WB
WB

Requires 6.75MHz to fully resolve, assuming 480i. Suppose our TV's max bandwidth was half that, or 3.375 MHz (sounds like something I built). The above pattern would be completely blurred to gray. However, we would be able to represent a pattern like this:

WWBB
WWBB
WWBB
WWBB
WWBB
WWBB

Isn't this the theory behind the test patterns we see? As the pattern becomes more "coarse," the associated bandwidth decreases. What am I missing?

By the way, have you looked at the disc yet? What did you think?

Thanks to anyone who can help set me straight here.

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #51 of 4106 Old 12-04-2007, 01:31 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,024
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Liked: 425
hwjohn,

I completely misread your original post, I thought you were saying that alternating single (not double) pixel width lines required half the 6.75 MHz bandwidth. Sorry for the ensuing confusion.

Another note, I think the chroma signal is usually sampled at half the luma rate so color difference signals can only support half the resolution that the pure luma signal supports. How that affects real-world resolution in images I don't know.
zoyd is online now  
post #52 of 4106 Old 12-04-2007, 01:39 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

hwjohn,

I completely misread your original post, I thought you were saying that alternating single (not double) pixel width lines required half the 6.75 MHz bandwidth. Sorry for the ensuing confusion.

Another note, I think the chroma signal is usually sampled at half the luma rate so color difference signals can only support half the resolution that the pure luma signal supports.

It's my fault; I have a tendency to confuse everyone and start arguments all at the same time. Probably a result of my southern "redneck" grammar. That part of my personality fits in a lot better with my other (and unfortunately more expensive) hobby: guns

I think you are right on the chroma signal; it is a consequence of the 4:2:0 subsampling used for DVDs.

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #53 of 4106 Old 12-04-2007, 01:49 PM
AVS Special Member
 
dr1394's Avatar
 
Join Date: May 2002
Location: Mizar 5
Posts: 3,184
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 22
Quote:
Originally Posted by hwjohn View Post

Thanks for the explanation Ron. I have never thought about this, so I'm still trying to digest it.

As the frame rate and lines/frame become slower/lower, the samples/line increase. Are there any advantages to this?

I am still not understanding how a vertical pattern consisting of a 2 pixel white line followed by a 2 pixel black line doesn't half the bandwidth needed to represent it. It is my understanding that a pattern like this (let each letter represent one pixel, W for white, B for black):

WB
WB
WB
WB
WB

Requires 6.75MHz to fully resolve, assuming 480i. Suppose our TV's max bandwidth was half that, or 3.375 MHz (sounds like something I built). The above pattern would be completely blurred to gray. However, we would be able to represent a pattern like this:

WWBB
WWBB
WWBB
WWBB
WWBB
WWBB

Isn't this the theory behind the test patterns we see? As the pattern becomes more "coarse," the associated bandwidth decreases. What am I missing?

By the way, have you looked at the disc yet? What did you think?

Thanks to anyone who can help set me straight here.

You have it correct. That's why I have this pattern on my website.



It's just 4, 3, 2, and 1 pixel pairs which would be 9.28125, 12.375, 18.5625 and 37.125 MHz (all divided by 1.001 for 29.97 fps).

BTW, on the numbers I posted previously, remember that it's just the number of blanking pixels that are changing. For all the HD rates, there's still an 1920x1080 or 1280x720 image inside of the blanking.

Ron

HD MPEG-2 Test Patterns http://www.w6rz.net
dr1394 is offline  
post #54 of 4106 Old 12-04-2007, 02:06 PM
AVS Special Member
 
dr1394's Avatar
 
Join Date: May 2002
Location: Mizar 5
Posts: 3,184
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 22
BTW, the alternating pixel patterns are usually called "patches" since they are square waves. "Bursts" are made with sine waves. Here's some low frequency bursts that show the sine wave (especially the one in the upper right hand corner which is 0.5 MHz).



Ron

HD MPEG-2 Test Patterns http://www.w6rz.net
dr1394 is offline  
post #55 of 4106 Old 12-04-2007, 02:26 PM - Thread Starter
AVS Special Member
 
alluringreality's Avatar
 
Join Date: Jul 2006
Posts: 3,204
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 40
Quote:
Originally Posted by dr1394 View Post

BTW, on the numbers I posted previously, remember that it's just the number of blanking pixels that are changing. For all the HD rates, there's still an 1920x1080 or 1280x720 image inside of the blanking.


What are blanking pixels?
Okay... so it's video timing related. Does it relate to creating test patterns though? From the discussion I don't see how it fits in.

The post #54. Would there be a practical use for that pattern?
alluringreality is offline  
post #56 of 4106 Old 12-04-2007, 03:24 PM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by alluringreality View Post

What are blanking pixels?
Okay... so it's video timing related. Does it relate to creating test patterns though? From the discussion I don't see how it fits in.

The post #54. Would there be a practical use for that pattern?

It doesn't relate directly I guess... they were just trying to steer me in the right direction on understanding the burst patterns, which in turn does have at least a little something to do with creating test patterns since we need a sharpness pattern.

The blanking signal occurs during the time between screen refreshes. On CRTs, it is the time that it takes for the guns to get back from the right side to the left, or the bottom back up to the top. By fudging it a bit and sticking more "fake" pixels in the blanking signal, each standard (NTSC or PAL) can have the same sample rate. Correct me if that is incorrect Ron, Zoyd.

I'll let Ron comment on the practical use of the patterns since they are his. He has probably thought of a lot more uses for them than I have.

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #57 of 4106 Old 12-05-2007, 08:12 AM
AVS Special Member
 
dr1394's Avatar
 
Join Date: May 2002
Location: Mizar 5
Posts: 3,184
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 22
Quote:
Originally Posted by alluringreality View Post

What are blanking pixels?
Okay... so it's video timing related. Does it relate to creating test patterns though? From the discussion I don't see how it fits in.

It's how you relate pixels to samples per second (expressed in MHz). If you know the sample rate, then you can construct a digital sine wave of any desired frequency with a little code. Like so for a 1 MHz burst:
Code:
                f = ((2 * 3.141592653 * 1000000) / (74250000 / 1.001));
                for(i = 0; i < 460; i++)  {
                        value = sin(f * (double)i);
                        value = (value * 219 / 2) + (219 / 2) + 16;
                        y = (unsigned char)value;
                        buffer422[index++] = 0x80;
                        buffer422[index++] = y;
                }
Quote:
Originally Posted by alluringreality View Post

The post #54. Would there be a practical use for that pattern?

It's actually more useful for an oscilloscope or waveform monitor. See this post in the GetGray thread:

http://www.avsforum.com/avs-vb/showt...43#post6692743

Also, you can have a burst at any frequency, but you can only have patches at 1, 1/2, 1/3, 1/4, etc. of the sampling frequency. If your display can't resolve full resolution, the 1x patch will be gone, but the 1/2x patch will most likely be present. You don't know where your display is cutting off. It could be anywhere between 20 and 37.125 MHz. With bursts, you can estimate the cutoff frequency (although it's a bit difficult by eye). Here's a burst pattern starting with 20 MHz in the upper left hand corner and ending with 35 MHz in the lower right hand corner (incrementing by 1 MHz).



Ron

HD MPEG-2 Test Patterns http://www.w6rz.net
dr1394 is offline  
post #58 of 4106 Old 12-05-2007, 08:22 AM - Thread Starter
AVS Special Member
 
alluringreality's Avatar
 
Join Date: Jul 2006
Posts: 3,204
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 40
Thanks again for the feedback. So far I haven't read anything that might indicate too many issues with creating a sharpness pattern the same way we created the other patterns. I might just use autocad to create some different shapes and export them as BMPs to use in images. The thing I notice with 'enhancement' type controls is that they tend to make diagonal lines blocky and autocad is the easiest way I can think to create something at all similar to some of the different shapes the Sony pattern offers to look at.
alluringreality is offline  
post #59 of 4106 Old 12-05-2007, 08:49 AM
AVS Special Member
 
hwjohn's Avatar
 
Join Date: Oct 2006
Location: Down by the river
Posts: 1,018
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by dr1394 View Post

It's how you relate pixels to samples per second (expressed in MHz). If you know the sample rate, then you can construct a digital sine wave of any desired frequency with a little code. Like so for a 1 MHz burst:
Code:
                f = ((2 * 3.141592653 * 1000000) / (74250000 / 1.001));
                for(i = 0; i < 460; i++)  {
                        value = sin(f * (double)i);
                        value = (value * 219 / 2) + (219 / 2) + 16;
                        y = (unsigned char)value;
                        buffer422[index++] = 0x80;
                        buffer422[index++] = y;
                }
It's actually more useful for an oscilloscope or waveform monitor. See this post in the GetGray thread:

http://www.avsforum.com/avs-vb/showt...43#post6692743

Also, you can have a burst at any frequency, but you can only have patches at 1, 1/2, 1/3, 1/4, etc. of the sampling frequency. If your display can't resolve full resolution, the 1x patch will be gone, but the 1/2x patch will most likely be present. You don't know where your display is cutting off. It could be anywhere between 20 and 37.125 MHz. With bursts, you can estimate the cutoff frequency (although it's a bit difficult by eye). Here's a burst pattern starting with 20 MHz in the upper left hand corner and ending with 35 MHz in the lower right hand corner (incrementing by 1 MHz).



Ron


At the risk of alluringreality killing me for further hijack of this thread


Code:
f = ((2 * 3.141592653 * 1000000) / (74250000 / 1.001));
Pretty sure gives you the distance along x-axis between samples (for lack of a better term).
Code:
for(i = 0; i < 460; i++)  {
What significance does 459 have?

Code:
value = sin(f * (double)i);
Find value of 1MHz sin wave at a specific point along x-axis...

Code:
value = (value * 219 / 2) + (219 / 2) + 16;
This looks like some type of bias operation to shift the entire curve. What significance do 219 and 16 have (I'm guessing 16 has to do with video level black)?
Code:
y = (unsigned char)value;
                        buffer422[index++] = 0x80;
                        buffer422[index++] = y;
}

... and this is where my deficiency in C comes in... why 0x80 the buffer?

AVS HD 709 - Free calibration disks
The 2007 Patriots: 18 -1
Tom who?
"...the small of napalm in the evening breeze, as I crouch behind a shopping cart in the parking lot..." - The Mall Ninja
hwjohn is offline  
post #60 of 4106 Old 12-05-2007, 09:48 AM
AVS Special Member
 
dr1394's Avatar
 
Join Date: May 2002
Location: Mizar 5
Posts: 3,184
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 22
The burst is 460 pixels wide. The sin() function returns a number from -1 to 1, so it needs to be converted to 16 to 235. 219 is (235 - 16). 0x80 is the Cb or Cr chroma pixel (black and white pattern).

Ron

HD MPEG-2 Test Patterns http://www.w6rz.net
dr1394 is offline  
Reply Display Calibration

Tags
Lcd Hdtv , Led Hdtv , lg bp530
Gear in this thread - bp530 by PriceGrabber.com

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off