Here is what I see through both Cable and OTA. I see it through two Tivo HD boxes and a standard QAM tuner.

From Tivos (one cable and one OTA):

http://www.boringgames.com/HD2.JPG
http://www.boringgames.com/HD4.JPG

One from QAM:

http://www.boringgames.com/HD3.JPG


The only common thing between the Tivos and QAM/ATSC setup is the type of cable (Belden 2.25ghz Quad Shielded RG6) and compression fittings.


Be glad to listen to any suggestions on what could be wrong.

I would suggest you provide more information about your setup. Hardware and interconnect specifics as a start.

Certainly need display details, but your 1st/last shots seem to have a screen overlay--those dispersed bluish blotches in the 1st--currently being discussed in some plasma-display-forum threads (http://www.avsforum.com/avs-vb/showthread.php?t=1113122). Some there attribute this to poor anti-reflective coatings applied to certain plasma screens. -- John

Here is what I see through both Cable and OTA. I see it through two Tivo HD boxes and a standard QAM tuner.

From Tivos (one cable and one OTA):

...

One from QAM:

...

The only common thing between the Tivos and QAM/ATSC setup is the type of cable (Belden 2.25ghz Quad Shielded RG6) and compression fittings.


Be glad to listen to any suggestions on what could be wrong.

So, it can't be the TiVo, since it happens on QAM.
It can't be the cable co., since it happens OTA.

It is some odd macroblocking type thing, but from the first screenshot (is that Dexter?) it seems odd, because I've never seen color specific macroblocking.

Does this happen on all content all the time?

Does it ever happen on SD?

DVD? Blu-Ray?

The cable comes in the house to a three way splitter. The three way splitter goes to the two Tivo HDs and to the cable modem. This is all the Belden 2.25 ghz swept coax.

One Tivo HD is connected to a Mitsubishi Rear Projection TV via component cables to the 1080I input. And the other Tivo HD is connected to a 37" Sharp Aquos LCD via HDMI. Both TVs show the same quality of picture as above. The same show recorded on the two Tivos look the same. And I have used Avia and Digital Video Essentials. Sharpness is all the way down and everything is adjusted as best as can be.

I connected the OTA antenna to one Tivo and recorded the same show via OTA and Cable and they are pretty much identical.

The third picture there is my OTA antenna hooked up to a Silicon Dust HD Homerun tuner. It was recorded on a Vista Media Center PC and it looks the same played back via Vista Media Center.


I started this thread originally in the Tulsa Reception thread and was getting responses that they don't see the problems so I brought it over here.

So, it can't be the TiVo, since it happens on QAM.
It can't be the cable co., since it happens OTA.

It is some odd macroblocking type thing, but from the first screenshot (is that Dexter?) it seems odd, because I've never seen color specific macroblocking.

Does this happen on all content all the time?

Does it ever happen on SD?

DVD? Blu-Ray?

That is NCIS on CBS as is the second picture. The Third was from Chuck on ABC. It happens on NBC, USA and other channels.

Generally HD looks ok, but there are spots in every show that look like this. People have brought up compression and panning issues, but I still go back to not finding anyone that complains about the same issues/quality.. But yes, the same stuff happens during show intros every time, every week so there is consistency.

SD does not, SD capture of one of the same pics:

http://www.boringgames.com/SD2.JPG


DVD Progressive Scan player looks fine and is connected to a 480p input on the TV. No Blu-Ray yet.

The Sharp Aquos has many HDMI inputs from game systems. They look amazing on it. No issues there.

That is NCIS on CBS as is the second picture. The Third was from Chuck on ABC.
Chuck is on NBC - just for the sake of keeping track of what stations you're monitoring...

You are not receiving QAM digital if you only have an antenna connected to your HDhr you are receiving OTA digital.
What graphics card are you using and what driver release for it are you using.
When you say you are connected to the 1080i input on your Mits does that mean 1080i/60 using the component interface or 1080i/60 using a DVIor HDMI interface?
When you say a 480p interface on the TV do you mean 480p over VGA, 480p over compoent. or 480p over DVI/HDMI?
What is the resolution of the source broadcasts shown and if HD resoloution are the programs "filmed" with SD or HD resolution cameras?

Chuck is on NBC - just for the sake of keeping track of what stations you're monitoring...

Yes, NBC. Wow, I don't watch anything on ABC normally...

You are not receiving QAM digital if you only have an antenna connected to your HDhr you are receiving OTA digital.

Sorry, meant ATSC

What graphics card are you using and what driver release for it are you using.

9800GTX+ with 182.08 WQHL drivers under Vista x64

When you say you are connected to the 1080i input on your Mits does that mean 1080i/60 using the component interface or 1080i/60 using a DVIor HDMI interface?

1080i component inteface. This TV has no HDMI or DVI inputs.

The Sharp LCD is all HDMI 1080p

When you say a 480p interface on the TV do you mean 480p over VGA, 480p over compoent. or 480p over DVI/HDMI?

480p over Component

What is the resolution of the source broadcasts shown and if HD resoloution are the programs "filmed" with SD or HD resolution cameras?

I have no idea where to find that info

ABC, FOX, and ESPN-HD are broadcast in 720p the rest are broadcast in 1080i/60.
If you are watching a "HD" resolution 16:9 aspect ratio progama that is "filmed" with a SD camera and not a HD widescreen camera then the video will be of poorer quality and will have blasc/grey/or blue side pillers added to the actual 4:3 video content.
What is the native resolution of your Mits if it is a digital technology TV?
Or if it is a CRT technology TV does if have 9" guns and capable of scanning at either 480i or 1080i?
Have you also tried 1080i/60 with the Sharp in addition to 1080p/60 or 1080p/24?

Looks to me like a lot of those frames are right at a scene transition. Interframe compression algorithms like MPEG2 rely on encoding a small number of full frames, and then storing the rest of the frames as the difference between the full frame and itself. For this to work best, the encoder needs to be able to detect when there's a major change in the picture and compensate by using a full frame rather than trying to figure out all the differences. Some encoders do this better than others, and some don't seem to try it at all. When it doesn't happen, the resulting picture can look really bad for a few frames until the encoder finally does insert a new full frame.

As for the frame with the explosion, those are simply compression artifacts. Fire and moving water are some of the hardest things to encode because there's a large amount of change from frame to frame, and the motion is random and can't accurately be tracked. Nothing you can do on your end will fix overcompression by your local stations.

AFAIK part of the "frame" header in MPEG compreed video contains the information stating if the following compressed data next is a full frame from a new scene, a full frame from the current scene or a change of the data from the currennt frame.
Therfore the OP should not be seing problems on scene changes.

No. A properly configured encoder with sufficiently advanced capabilities will insert a full frame on a scene change, but there are a whole lot of encoders (probably the majority in use) that don't have the capability or have it disabled.

I am surprised since I have never heard of this problem occuring with the MPEG2 decoders avaialle with hardware acceleration in recent generation graphic chips so I assumed that any MPEG2 software decoders would also decode any MPEG2 properly encoded content properly.

It's not a decoder problem. The same poor quality is present to everyone watching that particular source; most people just don't notice since it's only visible for a couple frames.

So it is a problem with the encoder being used by the source since it apparently does not indicatd when the scene changes.
Toy4x4,
What channels do you have the problem on. If all broadcast channels do you also have the problem with cable only HD content channels?

It's the broadcast channels and cable channels. Some channels are better than others but I can't really think of a channel that I watch regularly that does not have some problem. I would say Discovery, History and other non-broadcast ones are best but they have their moments.

When you use the antenna are you also using it through the splitter as you are when using cable?
I am surprised that there is not a single combination of direct cable or direct antenna connection to either your PC or your Tivo that works with either of your TVs using either 720p or 1080i.

How long ago did this start? I would hate to think that this is the way it has always been. If it has just started, go through the process of elimination. Run your cable directly to your box and then to the TV. If this just started and since you have the same problem on multiple TV's, I would start with the splitter. You say that your DVD is fine and I am guessing that is connected directly to your TV without any other interference. Try doing a direct connect. Something inbetween has let you down.

No. A properly configured encoder with sufficiently advanced capabilities will insert a full frame on a scene change, but there are a whole lot of encoders (probably the majority in use) that don't have the capability or have it disabled.

Hi there

I know very little about MPEG, but a quick check of some HD recordings indicate that a scene change without a full MPEG frame is not sufficient to cause these artifacts. I can use the VideoReDo MPEG editor to display the I, B or P frame type. Sometimes the scene change happens with an I frame, sometimes with a B or P frame; in the 2 dozen scene changes checked, no artifacts were seen.

Perhaps a scene change without an I frame _and_ too much compression can cause these artifacts.

I do recall seeing a similar artifact the last time I did a edit, but it may take a while for me to find it.

Regards

Hi there

I know very little about MPEG, but a quick check of some HD recordings indicate that a scene change without a full MPEG frame is not sufficient to cause these artifacts. I can use the VideoReDo MPEG editor to display the I, B or P frame type. Sometimes the scene change happens with an I frame, sometimes with a B or P frame; in the 2 dozen scene changes checked, no artifacts were seen.

Perhaps a scene change without an I frame _and_ too much compression can cause these artifacts.

I do recall seeing a similar artifact the last time I did a edit, but it may take a while for me to find it.

Regards
There are certainly other factors involved, including the overall bitrate of the stream, the availability of extra bits in the buffer, the quantization matrix in use, how good the encoder is, etc. One thing to keep in mind is that you're in LA, and all the network stations are running fairly new encoders that are much more efficient than the older ones that are probably still in use in a smaller market like say #52 Tulsa where the OP is located.

Hi there

In a panning scene from an old (19Nov2007) episode of Chuck, an object in the middle of the P frame is smeared by horizontal lines. This is probably due to a 3:2 pull-down issue (assuming Chuck is shot on film or 24fps video)?

Then about 20 frames later the pan ends with a cut to the next scene. The B frame that precedes the cut has two blobs, a blue blob and a red blob, that line up perfectly with objects in the next B frame of the new scene. Is there a way to determine if this is an encode or decode error?

Regards