MPEG4 Picture Quality

[pcdoc]

As others here, I am thinking about giving up my 2 HR10s for 2 HR20s. The only thing holding me back is the degrading picture quailty of DirecTV over the last few years. For those who have made the switch, is the MPEG4 HD picture quality on the HR20 better or the same as MPEG2 on the HR10?
Thanks.

[JimSpence]

I can't answer specifically. But, I think it is too soon, since DirecTV is still getting their ducks in a row.

My estimate is that PQ should get better.

[joed32]

We will have a better idea in about three weeks, til then nobody knows.

[fasTLane]

Locals are not in mpeg4 now?

[Redux]

We will have a better idea in a couple of weeks, til then nobody knows.We have already had reports that MPEG4 HiDef locals are just as good as OTA. I expect to hear more of this kind of report.

The source must be considered. Yes, intended.

[IOTP]

We have already had reports that MPEG4 HiDef locals are just as good as OTA. I expect to hear more of this kind of report.

The source must be considered. Yes, intended.

Speaking of someone that has MPEG-4 HD locals in Phoenix let me share this:

Watching 15 KNXV I have noticed some pixelation in the local nightly news in HD.

I think it's from the uplink to DTV from the local tv provider.

Since all the primetime shows are in "repeat" it's hard to tell. I havent noticed pixelation in other ABC broadcasted events. (eg football, etc)

I pressed hard a couple of weeks ago when I had my HR20 and new dish installed for the install tech to LEAVE the OTA in place. He wasn't happy, but did it nonetheless.

I can flip from 15 KNXV MPEG-4 to 15-1 OTA and notice that OTA receives no pixelation.

Time will tell.

I am still waiting, like we all are for additional HD channel. That is why I upgraded to the HR20 and new MPEG-4 dish. Having been on the HD bandwagon for 2+ years now.

Local news on channels 3 (independent), 12 (NBC), and 15 (ABC) broadcast all local news in HD.

Someone also told me, that the OTA locals broadcasted over most cities are not-compressed, whereas MPEG-4 obviously are.

Bring it on, more HD channels that is.

Hope this helps.

[tadrow]

I was looking for artifacting or anything annoying when watching a football game in HD on an MPEG4 local channel. I didn't see anything that I would have been able to say was worse than what was being broadcast OTA. I was pleasantly surprised with what I saw. I watch on a 100 inch screen, so when things are a little ugly for most people, they are really ugly for me.

[Billy66]

I think, I believe that the OTA is better. If I put them on and switch back and forth, I can convinve myself that the OTA picture is better. When I look in dark color areas I think I can see greater pixelation. I guess I do.

However, after I flip back and forth several times and get mixed up, I can't guess 100% accurately if I'm watching the OTA local or the MPEG4 local.

I'll concede that I'm not a videophile, but that's been my experience.

[ken erickson]

Speaking of someone that has MPEG-4 HD locals in Phoenix let me share this:

Watching 15 KNXV I have noticed some pixelation in the local nightly news in HD.

I think it's from the uplink to DTV from the local tv provider.

Since all the primetime shows are in "repeat" it's hard to tell. I havent noticed pixelation in other ABC broadcasted events. (eg football, etc)

I pressed hard a couple of weeks ago when I had my HR20 and new dish installed for the install tech to LEAVE the OTA in place. He wasn't happy, but did it nonetheless.

I can flip from 15 KNXV MPEG-4 to 15-1 OTA and notice that OTA receives no pixelation.

Time will tell.

I am still waiting, like we all are for additional HD channel. That is why I upgraded to the HR20 and new MPEG-4 dish. Having been on the HD bandwagon for 2+ years now.

Local news on channels 3 (independent), 12 (NBC), and 15 (ABC) broadcast all local news in HD.

Someone also told me, that the OTA locals broadcasted over most cities are not-compressed, whereas MPEG-4 obviously are.

Bring it on, more HD channels that is.

Hope this helps.
I've changed over to the HR-20 two weeks ago and so far both the OTA and MPEG4 locals look about the same. I haven't had a chance to watch any HD games yet (missed the Raiders and the 49'er games so far). Hope get to see some games this holiday weekend.

[TyroneShoes]

I can't answer specifically...My estimate is that PQ should get better.
What in the world could you possibly be basing that on?

Certainly not the laws of physics, which dictate that any time you chain dissimilar digital compression algorithms together, such as M2 to M4 (which affects every OTA channel and current sat channel delivered by satellite), the rounding errors can only increase and artifacting increases as a direct result.

For digital video of any kind, the PQ ceiling is absolutely fixed at encode to M2. There are no exceptions. Anything done to it later on regarding re-encoding has no opportunity to improve PQ, and every opportunity to degrade it.

I'm not saying that the PQ will get noticeably worse, but don't expect it to get any better. DTV is increasing payload rates 30%, and M4 will accept a 30% increase over M2 with identical levels of bit starving, so PQ should be not noticeably different, and this seems to jive with what foks are seeing.

On the other hand, if they also abandon 1280x1080 for 1920x1080, PQ could get marginally better for some channels in regards to actual resolution (IOW, there would be less FURTHER degradation than they apply now under "HD Lite"), but that is not directly related to the move to M4, and does not release it from the shackles of chained algorithms which will still be in play.

So I'm afraid your "estimate" really has no merit.

[TyroneShoes]

Speaking of someone that has MPEG-4 HD locals in Phoenix let me share this:

Watching 15 KNXV I have noticed some pixelation in the local nightly news in HD.

I think it's from the uplink to DTV from the local tv provider.

Since all the primetime shows are in "repeat" it's hard to tell. I havent noticed pixelation in other ABC broadcasted events. (eg football, etc)

I pressed hard a couple of weeks ago when I had my HR20 and new dish installed for the install tech to LEAVE the OTA in place. He wasn't happy, but did it nonetheless.

I can flip from 15 KNXV MPEG-4 to 15-1 OTA and notice that OTA receives no pixelation.

Time will tell.

I am still waiting, like we all are for additional HD channel. That is why I upgraded to the HR20 and new MPEG-4 dish. Having been on the HD bandwagon for 2+ years now.

Local news on channels 3 (independent), 12 (NBC), and 15 (ABC) broadcast all local news in HD.

Someone also told me, that the OTA locals broadcasted over most cities are not-compressed, whereas MPEG-4 obviously are.

Bring it on, more HD channels that is.

Hope this helps.

I hope you resolve this issue, and I think you sharing it here is helpful, but everyone should keep in mind that what this experience is relating is purely a reception problem. This has no bearing on PQ as is commonly understood from an engineering standpoint. It is a common misconception that pixellation related to bitstarving due to poor reception is a PQ issue, but it really isn't, even though a picture that tears occasionally can be thought of in lay terms to have less "quality" than one that does not (and is probably why folks sometimes get this confused with PQ).

My best guess is that if your dish were aimed properly and you could verify a clear path to the sats, that you could fix the pixellation problem you are experiencing. IOW, reaim the dish, and see if that doesn't help.

If there actually were a pixellation problem related to actually delivering the M4 channels (rather than merely a random reception problem) then DTV would be in very deep doody with this program, undeniably.

As someone geographically close, I have a vested interest in the outcome (PM me with your results if you get a chance) yet I don't receive M4 channels as of yet, and hope to never receive them. It could also be a problem with DTV's reception OTA locally of these channels before being fibered to their uplink site. If so, all spot Phoenix viewers would have the same problem and reaiming would not help, so I would also try to verify whether your friends and neighbors have the same issue if you can.

[IOTP]

My best guess is that if your dish were aimed properly and you could verify a clear path to the sats, that you could fix the pixellation problem you are experiencing. IOW, reaim the dish, and see if that doesn't help. If there actually were a pixellation problem related to actually delivering the M4 channels (rather than merely a random reception problem) then DTV would be in very deep doody with this program, undeniably.

Again, only noticed it on the Local HD news. I'll try and watch more ABC and see if I can see any sort of pattern.

[Indiana627]

I switched from an HR10 to an HR20 in June. It's my very strong opinion that MPEG4 is better PQ than MPEG2. My MPEG4 locals look just as good to me as they do OTA.

The only direct channel comparison I can offer is YES. With my HR10, I would get YES-HD games on channel 95 from time to time and they never looked that good. Now I get the MPEG4 version of YES-HD and the games simply look amazing. Same with Mets games on SNY-HD.

When I swapped receivers I made no other changes to my setup - I'm even using the same HDMI cable.

To make long story short, I'd advise anyone who'd listen to not let the fear of MPEG4 PQ being worse than MPEG2 PQ stop them from going to the HR20.

[IOTP]

The only direct channel comparison I can offer is YES. With my HR10, I would get YES-HD games on channel 95 from time to time and they never looked that good. Now I get the MPEG4 version of YES-HD and the games simply look amazing. Same with Mets games on SNY-HD.

When I swapped receivers I made no other changes to my setup - I'm even using the same HDMI cable.

Are the channels in 94 / 95 area MPEG4? I was wondering when you receive your YES-HD or SNY-HD, are these channels MPEG-4 content?

Wondering because I would like to have my FOX SPORTS ARIZONA (649) in HD if possible. Only HOME games broadcast on KTVK are in HD, FSN AZ are SD and it's just not the same.

Threads and people say the regional FSNs and other regional sports will "eventually" be HD. The future is now.

It's like I say to myself, crap, "it's not in HD."

[TyroneShoes]

...Someone also told me, that the OTA locals broadcasted over most cities are not-compressed, whereas MPEG-4 obviously are...(I wanted to address this point separately)

All digital video that is delivered to consumers is compressed to one degree or another. This includes DVDs, OTA, M4, M2, Blu-Ray, or what-have-you. The degree of compression is what can make a difference in how good the PQ is, as can the quality of the original analog video whenever the source is not digital. Compression can range from having no visible degradation to having significant visible degradation.

DBS quality is typically about the same for any good-quality source, whether M2 or M2 converted to M4, generally speaking.

Also, FSA and other FOX regional sports channels are all actively converting to HD and should be finished this year. Whether DTV carries them or not will be a separate issue. KTVK did not renew their D-backs contract, so all of those games will move to FSA.

[patsfan123]

I upgraded to the HR20 a few weeks ago (I still have my HR10-250 in a different room).

The MPEG4 HD channels definitely look better than the MPEG 2 channels. I don't have OTA anymore so I can't compare to that (except from memory), but between MPEG2 and MPEG 4 it is crystal clear on my 50" Pioneer plasma that MPEG4 is better.

Which is good news for all of us in the long run.

[pcdoc]

I upgraded to the HR20 a few weeks ago (I still have my HR10-250 in a different room).

The MPEG4 HD channels definitely look better than the MPEG 2 channels. I don't have OTA anymore so I can't compare to that (except from memory), but between MPEG2 and MPEG 4 it is crystal clear on my 50" Pioneer plasma that MPEG4 is better.

Which is good news for all of us in the long run.

Thanks, Patsfan. That clinched my decision. I call D* and just told them about my many random reboots lately and they are sending me an HR20 free, as I have the replacement plan on my account. Also, as others have mentioned, I keep the HR-10.

Everyone, thanks for your opinions.

[milominderbinder]

Last winter you could still tell the difference between OTA and HR20 MPEG4. Not anymore, at least not in the Chicago DMA.

I am able to split screen the OTA antenna and the HR20 side-by-side on a new Mitsubishi 73".

I have had friends and family compare. You just cannot tell the difference.

I have done the same results comparing the OTA antenna and the H20.

Again, no one can tell the difference.

The only reason to avoid the H20 is you will be miserable because it is not a DVR.

The real comparison is not MPEG4 to MPG2. The real comparison is HD to SD.

We are in our final countdown to 70 MPEG4's in September.

And then 100 by years end. And then 150 by year end 2008.

The D10 satellite is up and ahead of schedule.

Vinyl to CD, VHS to DVD, Black & White to Color, SD to HD.

When big changes come some ride the wave and have a blast. Some are swept behind.

You have invested roughly a gazillion dollars in your HD so far, right?

If you have an HR10, you should be able to get the $19 price and free installation still.

If you don't like the HR20, suspend it and you are out only the $19 to try it. Or hedge your bet by using both the HR10 and HR20 for just the $4.99 extra mirroring fee each month. Or sell the HR10 on eBay for $180 and make a profit on the HR20.

The HR20 has been reliable for months now. The new VOD beta is wonderful.

Can you explain it again...what is the the reason to not try the HR20?

- Craig

[Ein]

From the horse's mouth:
http://www.tvweek.com/news/2007/03/directv_hd_additions_wont_redu.php

PQ will not change with Mpeg-4.

[Redux]

From the horse's mouth:
http://www.tvweek.com/news/2007/03/directv_hd_additions_wont_redu.php

PQ will not change with Mpeg-4."[Robert] Mercer [DirecTV spokesman] said there's nothing lacking in DirecTV's current quality and said that quality would not change despite the addition of extra channels."

If you can say with a straight face that there's nothing lacking in DirecTV's current quality, you can say just about anything.

But unless they get greedy with the compression level and/or stack compressions there should be no decline. And even then, there are many with eyesight such that they won't notice. A lot of money could have been saved in home sound systems over the years if people would simply get their hearing tested first.

[fasTLane]

I upgraded to the HR20 a few weeks ago (I still have my HR10-250 in a different room).

The MPEG4 HD channels definitely look better than the MPEG 2 channels. I don't have OTA anymore so I can't compare to that (except from memory), but between MPEG2 and MPEG 4 it is crystal clear on my 50" Pioneer plasma that MPEG4 is better.

Which is good news for all of us in the long run.Which HD channels are available in both compression schemes?

[Indiana627]

Are the channels in 94 / 95 area MPEG4? I was wondering when you receive your YES-HD or SNY-HD, are these channels MPEG-4 content?

Wondering because I would like to have my FOX SPORTS ARIZONA (649) in HD if possible. Only HOME games broadcast on KTVK are in HD, FSN AZ are SD and it's just not the same.

Threads and people say the regional FSNs and other regional sports will "eventually" be HD. The future is now.

It's like I say to myself, crap, "it's not in HD."
Channels 94/95 are MPEG2 since you can get them on the HR10. Once I switched to the HR20, I began getting YES-HD and SNY-HD MPEG4 on channels 622 and 625 (yes same channel numbers as the SD versions). The HD games on 622 and 625 are much, much better than the same game on 94/95.

For your particular RSN, I can't say because I'm not familiar with it. All I know is I have no problem saying D* MPEG4 PQ is better than D* MPEG2 PQ.

[sluciani]

I think, I believe that the OTA is better. If I put them on and switch back and forth, I can convinve myself that the OTA picture is better. When I look in dark color areas I think I can see greater pixelation. I guess I do.

However, after I flip back and forth several times and get mixed up, I can't guess 100% accurately if I'm watching the OTA local or the MPEG4 local.

I'll concede that I'm not a videophile, but that's been my experience.I live about 20-25 miles from the Empire State Building in NY, so my OTA roof antenna pulls in the same feed that D* uses for the East Coast feed. I can tell you that sitting 8'-10' away from a 50" Fujitsu plasma (768p), there is virtually no visual difference between D*'s MPEG-4 and MPEG-2 OTA on the same channels.

Where I do see a bit of a difference is in the smoothness of cruise control operations. Unfortunately MPEG-4 trickplay is still not quite as smooth as MPEG-2 trickplay. I'm not sure if D* can make it better through programming, if faster decoding hardware is required or if MPEG-4 in general, because of the higher compression ratios, will never ever have as smooth trickplay as MPEG-2.

At any rate, MPEG-4 trickplay is really not an issue, but just not as good as MPEG-2, IMO. /steve

[milominderbinder]

It will cost you $19 for the box and $4.99 a month to have both the HR10 and HR20.

You get the best of both worlds. Plus you can record 4 programs instead of 2.

Plus you would have 50 extra hours of HD storage.

I think the modern day sage Homer Simpson would equate this to donuts...

It is like asking, "Do I want 70 donuts next month or 10?"

It will only cost you $4.99 per month to have 70 donuts instead of 10.

Why are you talking about donuts instead of eating them? :)

I think Homer would say:

"Donuts good..."

Tell me again. Why you are not checking out the HR20 when it will only cost you $20?

Is the point of a receiver the receiver or what it can receive?

- Craig

[fasTLane]

...and the selling goes on

[sluciani]

Which HD channels are available in both compression schemes?I get ABC, CBS, FOX and NBC three ways on the HR20: MPEG-2 "HD-Lite" (channels currently in the 80's), MPEG-4 and MPEG-2 OTA. /s

[SMWinnie]

(as truncated)
[A]ny time you chain dissimilar digital compression algorithms together, such as M2 to M4 (which affects every OTA channel and current sat channel delivered by satellite), the rounding errors can only increase and artifacting increases as a direct result.

For digital video of any kind, the PQ ceiling is absolutely fixed at encode to M2.Clearly true for re-encodes of OTA signals, but is there any chance that D* might be able to get superior (higher bitrate) starting material?

F'rinstance, the local ABC affiliate runs a lower bitrate on their ABC-HD feed in order to carry CW on a subcarrier. The local PBS affiliate has floated plans to run 3-4 subchannels as PBS makes the programming available. (NBC has a 24/7 weather service.) If D* could get the full 19.2mbps signal prior to the bitrate reduction, then the nonzero-degraded MPEG4 re-encode might be competitive with the reduced bitrate OTA signal. (When last I checked, that was about 13.6mbps for the ABC-HD feed on WLAJ-DT.) Splitting off the full bitrate signal for D* and the MSOs at the local studio is expensive to retrofit but less expensive to build into the station's next studio upgrade.

And the assumed 19.2mbps upper limit and MPEG2-to-MPEG4 transcode are driven by a changable current production and distribution practice. If I recall correctly, HBO announced in June that they were moving their whole lineup to 8.0Mbps MPEG4 encodes.

Now, since MPEG2 is hardcoded into ATSC, that doesn't extend to OTA broadcasts...unless the committee amends the ATSC standards to allow MPEG4. Candidate standard (http://www.atsc.org/standards/cs_documents/CS-TSG-659r2.pdf)

[rcliff]

Certainly not the laws of physics, which dictate that any time you chain dissimilar digital compression algorithms together, such as M2 to M4 (which affects every OTA channel and current sat channel delivered by satellite), the rounding errors can only increase and artifacting increases as a direct result.Yes, but this would be true only if they they were sending an unmodified stream. You're ignoring the fact that D* absolutely butchers the mpeg2 signal they are sending down now. If they re-encode with mpeg4 at higher quality than they currently do with mpeg2, picture quality should improve over what we see now.

[sluciani]

Yes, but this would be true only if they they were sending an unmodified stream. You're ignoring the fact that D* absolutely butchers the mpeg2 signal they are sending down now. If they re-encode with mpeg4 at higher quality than they currently do with mpeg2, picture quality should improve over what we see now.And it does. As I said in a previous post, I'm happy to report that normal playback of MPEG-4 is virtually indistinguishable from the MPEG-2 OTA it is encoded from. :) /steve

[videojanitor]

Here's what I have found (if it matters to anyone): In comparing the same program from three sources: OTA, MPEG-4 HD Local, and MPEG-2 HD DNS, there are definite differences to my eyes.

OTA is best, hands down.

MPEG-4 HD Local looks very close to OTA *most* of the time, but there are some things that it can't handle -- most difficult for it is a simple dissolve (mix) especially from say a 100% white screen to almost anything else. It falls apart, with obvious blocking. This could be due to re-encoding, or an insufficient bit rate to handle this material. Either way, the flaws are there if you know where to look, but likely would go unnoticed by most viewers.

MPEG-2 HD DNS is at the bottom. It has a much higher number of compression artifacts, which can be seen even during scenes with minimal motion. The edges of everything look "busy" or "noisy."

Bottom line, MPEG-4 should provide an improvement over MPEG-2 for the upcoming channels, assuming that they don't go nuts and allocate fewer bits than the HD locals are currently allowed.

[pendragn]

Which MPG4 channels are you guys watching to do the comparisons? I didn't know any of them were online yet.

tk

[videojanitor]

My comparison was based on an MPEG-4 LOCAL channel. I compared that to the same station via OTA, and also against KCBS (channel 81, MPEG-2).

[bigpuma]

Which MPG4 channels are you guys watching to do the comparisons? I didn't know any of them were online yet.

tk

I get Fox, ABC, CBS, and NBC in HD from DirecTV in Mpeg-4 format. I can't tell the difference between those and OTA.

[rcliff]

Bottom line, MPEG-4 should provide an improvement over MPEG-2 for the upcoming channels, assuming that they don't go nuts and allocate fewer bits than the HD locals are currently allowed.A big assumption. D* knows that people are making buying decisions based on the current mpeg4 locals and are in full sales mode so they're probably putting their best foot forward. What they roll out with the HD new channels is yet to be seen but historically no one knows how to lower the bar better than D*.

[TyroneShoes]

...MPEG-4 trickplay is still not quite as smooth as MPEG-2 trickplay. I'm not sure if D* can make it better through programming, if faster decoding hardware is required or if MPEG-4 in general, because of the higher compression ratios, will never ever have as smooth trickplay as MPEG-2...
There is probably a good reason for that. Trick-play features are hampered by the GOP length. M2 has a 15-frame GOP typically when delivered to consumers, while AVC M4 is capable of GOP lengths up to 200 frames, which is one of the big reasons it can compress more efficiently for a similar PQ delivery.

You can get an idea of how this comes into play by trying a reverse frame advance on a HR10, which jumps in increments of 15 frames, precisely because of the 15-frame GOP length.

As PVRs advance, I would expect them to cope with these issues better over time, but I doubt a OS up rev to current platforms could have much impact on that. We can hope, of course.

[TyroneShoes]

Clearly true for re-encodes of OTA signals, but is there any chance that D* might be able to get superior (higher bitrate) starting material?

F'rinstance, the local ABC affiliate runs a lower bitrate on their ABC-HD feed in order to carry CW on a subcarrier. The local PBS affiliate has floated plans to run 3-4 subchannels as PBS makes the programming available. (NBC has a 24/7 weather service.) If D* could get the full 19.2mbps signal prior to the bitrate reduction, then the nonzero-degraded MPEG4 re-encode might be competitive with the reduced bitrate OTA signal. (When last I checked, that was about 13.6mbps for the ABC-HD feed on WLAJ-DT.) Splitting off the full bitrate signal for D* and the MSOs at the local studio is expensive to retrofit but less expensive to build into the station's next studio upgrade.

And the assumed 19.2mbps upper limit and MPEG2-to-MPEG4 transcode are driven by a changable current production and distribution practice. If I recall correctly, HBO announced in June that they were moving their whole lineup to 8.0Mbps MPEG4 encodes.

Now, since MPEG2 is hardcoded into ATSC, that doesn't extend to OTA broadcasts...unless the committee amends the ATSC standards to allow MPEG4. Candidate standard (http://www.atsc.org/standards/cs_documents/CS-TSG-659r2.pdf)I guess there is always a chance, SW, but here is how it works currently, which is that the HD signal from the local station is most always picked off OTA first, meaning whatever configuration is set for 8VSB will set the bit rate for the signal eventually converted to M4. That means that the PQ can't get better than OTA, but could get worse.

Stations I am affiliated with wish to provide a direct fiber feed to DTV to increase reliability (if the x-mitter goes down, DTV customers still get the signal) but that is probably going to happen only after the 2-17-09 cutoff, and will not affect the encoding. Typically, when an HD signal is fibered in that manner, it is already encoded as M2, and delivered as SMPTE310 or ASI to the vendor, and therefore identical in PQ to what is broadcast and subject to the same limitations, which is why DTV receiving the locals OTA works just as well as that scheme would (except for the reliability factor).

Should the station have the opportunity to send HD with less compression, which is very unlikely due to the cost of the fiber path itself, which is based on bit rate, they likely will see no reason to double-encode for non-OTA subs, because there would be very little payoff for the added expense. The station is also bound by the quality of the incoming HD network feed, which typically arrives as an already-encoded ASI signal.

When a station encodes the local HD signal or combines an existing network HD signal into the SMPTE310 signal sent to their transmitter, the bit rates for secondary channels are set there. Each does not have its own 19.34 mb to play with. It would be very difficult if not expensive to send those signals with less compression to DTV or cable vendors.

[TyroneShoes]

Yes, but this would be true only if they they were sending an unmodified stream. You're ignoring the fact that D* absolutely butchers the mpeg2 signal they are sending down now. If they re-encode with mpeg4 at higher quality than they currently do with mpeg2, picture quality should improve over what we see now.
You're free to just keep telling yourself that, but that would only mean that you are delusional regarding how this really works in the real world. It's just not as simple as you seem to assume, and such conclusions can't be drawn when wholly without scientific basis. You're also ignoring the fact that the picture quality CEILING is SET IN STONE by the original encoding, meaning that the PQ can only get worse as it is transported further down the line, and never better.

Also, M4 has by definition a less-accurate encoding of the original digital video than M2, since it throws away a significantly-larger portion of the original information. In no way does it have "better" PQ than M2.

There is only one way that M4 DELIVERY can transport better quality than M2 delivery, which is if it uses less-severe bit rate restrictions, and that would only be noticeable if they were severely restricting the bandwidth of the original M2 signal today, which they are not. "severely butchering" is an unfair characterization, as we have provided actual proof right here on this forum that the differences are minimal.

Regardless, DTV has already stated that they will encode 30% more information, which is balanced out by the fact that their flavor of M4 will accept 30% more information for an "equivalent" level of post-degradation from the transport process. They spent hundreds of millions to bring us more channels, and not to bring us the same number of channels only with marginally better in PQ.

Even if that were their goal, converting to M4 would not help them realize it. Using a more-efficient algorithm or adding more available bits makes absolutely no positive contribution to PQ when there are enough bits to fully encode the signal already. If you take identical signals previously properly-encoded at 15 mbps and transport one at 15 mbps and transport the other at 1500 mbps, both will look absolutely identical at decode, because both will decode only all of the original 15 mbps of bits of information, and not one bit more (since no more information existed to be encoded in the first place).

Now, when you take either of those signals and re-encode it with a different algorithm (M4) before delivery to the customer, the laws of physics require that rounding errors and artifacts can only increase, meaning PQ only has one path to take. It either stays nearly the same, or it gets worse, depending upon engineering encoding-parameter choices. And it is physically impossible for it to get better.

Bottom line, nothing in DTV's plans do or can improve original PQ, because within the realms of practicality and budget restraints, that would be impossible.

[milominderbinder]

Well regardless of quality, 80 Channels of MPEG4 are being tested right now:

80 HD channels currently in testing mode (http://www.tivocommunity.com/tivo-vb/showthread.php?p=5460180#post5460180)

- Craig

[TyroneShoes]

Here's what I have found (if it matters to anyone): In comparing the same program from three sources: OTA, MPEG-4 HD Local, and MPEG-2 HD DNS, there are definite differences to my eyes.

OTA is best, hands down.

MPEG-4 HD Local looks very close to OTA *most* of the time, but there are some things that it can't handle -- most difficult for it is a simple dissolve (mix) especially from say a 100% white screen to almost anything else. It falls apart, with obvious blocking. This could be due to re-encoding, or an insufficient bit rate to handle this material. Either way, the flaws are there if you know where to look, but likely would go unnoticed by most viewers.

MPEG-2 HD DNS is at the bottom. It has a much higher number of compression artifacts, which can be seen even during scenes with minimal motion. The edges of everything look "busy" or "noisy."... This has absolutely not been my experience. In my tests (which are backed up by jpegs posted earlier on this very forum) I compared recordings of identical frames of FOX network OTA (KSAZ-DT Phoenix) with those from FOX network DNS (KTTV-DT Los Angeles). Since both local stations use MPEG splicers and therefore contribute nothing either positive or negative to network PQ, this provides true empirical evidence of what is different after processing by DTV, since that is the only changed variable.

The results, repeated on 3 different broadcasts, were very definitive that for static images, absolutely no difference could be seen, which is what we might expect since there is typically plenty of bits to go around for static images. For images with high motion, however, the affects of the post-compression in DTV delivery were temporally obvious, IOW resulting in pixellation that was only obvious for a frame or so, but obvious nonetheless to trained eyes such as those of broadcast engineers like yourself, which is exactly what you describe above for DTV M4 delivery.

Maybe it's time to repeat that experiment. It would be interesting if both of us could post new images supporting our conclusions.

[rjnerd]

Remember, jpeg is itself a lossy compression scheme. If you want an honest comparison, you will have to use raw images.

[rcliff]

You're free to just keep telling yourself that, but that would only mean that you are delusional regarding how this really works in the real world. It's just not as simple as you seem to assume, and such conclusions can't be drawn when wholly without scientific basis. You're also ignoring the fact that the picture quality CEILING is SET IN STONE by the original encoding, meaning that the PQ can only get worse as it is transported further down the line, and never better.

Also, M4 has by definition a less-accurate encoding of the original digital video than M2, since it throws away a significantly-larger portion of the original information. In no way does it have "better" PQ than M2.

There is only one way that M4 DELIVERY can transport better quality than M2 delivery, which is if it uses less-severe bit rate restrictions, and that would only be noticeable if they were severely restricting the bandwidth of the original M2 signal today, which they are not. "severely butchering" is an unfair characterization, as we have provided actual proof right here on this forum that the differences are minimal.

Regardless, DTV has already stated that they will encode 30% more information, which is balanced out by the fact that their flavor of M4 will accept 30% more information for an "equivalent" level of post-degradation from the transport process. They spent hundreds of millions to bring us more channels, and not to bring us the same number of channels only with marginally better in PQ.

Even if that were their goal, converting to M4 would not help them realize it. Using a more-efficient algorithm or adding more available bits makes absolutely no positive contribution to PQ when there are enough bits to fully encode the signal already. If you take identical signals previously properly-encoded at 15 mbps and transport one at 15 mbps and transport the other at 1500 mbps, both will look absolutely identical at decode, because both will decode only all of the original 15 mbps of bits of information, and not one bit more (since no more information existed to be encoded in the first place).

Now, when you take either of those signals and re-encode it with a different algorithm (M4) before delivery to the customer, the laws of physics require that rounding errors and artifacts can only increase, meaning PQ only has one path to take. It either stays nearly the same, or it gets worse, depending upon engineering encoding-parameter choices. And it is physically impossible for it to get better.

Bottom line, nothing in DTV's plans do or can improve original PQ, because within the realms of practicality and budget restraints, that would be impossible.Delusional??? I think not. What you're saying would only be true if they re-encoded their already low bandwidth HD-lite mpeg2 stream. This would not be true if they went back to the original source material then encoded as mpeg4. Regarding D*'s current mpeg2 HD, it is indeed severely butchered and I find some material almost unwatchable on my HD10-250 vs Cablevision or OTA on my Series 3. Yes, rencoding garbage as mpeg4 will only result in worse garbage but going back to the original high quality source stream and encoding as a high qaulity mpeg4 stream could easily yield PQ that surpasses the junk that D* currently calls HD.

[videojanitor]

This has absolutely not been my experience. [...]
Maybe it's time to repeat that experiment. It would be interesting if both of us could post new images supporting our conclusions.

I'm confused as to how we disagree. I said OTA is best, MPEG-4 as currently delivered by D* is worse (though the "average" viewer can't see the difference), and MPEG-2 as currently delivered by D* is at the bottom. Do you disagree with this??

[sluciani]

I'm confused as to how we disagree. I said OTA is best, MPEG-4 as currently delivered by D* is worse (though the "average" viewer can't see the difference), and MPEG-2 as currently delivered by D* is at the bottom. Do you disagree with this??When I first started watching MPEG-4 locals back in March, I noticed pixelization during the 1-2 second fades to and from black on scene changes and commercial breaks, similar to what you reported previously, VJ. Since then, D* has improved the encoding, at least for my NY MPEG-4 locals, so much so that I no longer see any difference between MPEG-4 and OTA MPEG-2, viewing from about 8-10 feet away on a 768p 50" Fujitsu panel during normal playback, and I'm a pretty critical viewer. I do still see differences in trickplay smoothness, however. /steve

[ayrton911]

I get ABC DNS from Los Angeles on 87, only one I get. Would I also get the MPEG 4 version, or is that only for locals?

[thedeak08]

I live in the Indianapolis area and they have NBC, FOX, and ABC in mpeg4 HD right now. I honestly can't tell a difference between the OTA and the mpeg4 from the satellite. They both look great, and it's on a 67' Samsung DLP TV, where you can notice things more with a big TV. Bottom line for me is....I will take the satellite over OTA just because I don't want one of my shows I'm watching to cut out when using OTA, which does happen. So I will sacrifice just a little PQ if there really is any.

[newsposter]

Bottom line for me is....I will take the satellite over OTA just because I don't want one of my shows I'm watching to cut out when using OTA, which does happen.

You must have a terrain problem or something. The one thing about OTA i love is that when the birds are out, the OTA is always working. Can't remember the last time i missed a show on ota.

[fasTLane]

My OTA-HD in Atlanta has been just fine.
Having doubts the HR20 will show all these channels anytime soon so, for locals, the semi-tivo is out of the loop.


:eek: oops!

[Jebberwocky!]

color me impressed with the qualiy of the mpeg4 reception. Even my Wife made a comment today shortly after my install and she never makes comments about picture quality.

[Ein]

color me impressed with the qualiy of the mpeg4 reception. Even my Wife made a comment today shortly after my install and she never makes comments about picture quality.


There is an old Chinese saying: "A new latrine always smells better for the first few days" :)

[TyroneShoes]

Delusional??? I think not. What you're saying would only be true if they re-encoded their already low bandwidth HD-lite mpeg2 stream. This would not be true if they went back to the original source material then encoded as mpeg4...Yes, rencoding garbage as mpeg4 will only result in worse garbage but going back to the original high quality source stream and encoding as a high qaulity mpeg4 stream could easily yield PQ that surpasses the junk that D* currently calls HD.I'm not disputing anything you are saying here, except the part about whether you are delusional or not. I'm also not saying there's anything wrong with not understanding the fine points of something (investment banking makes my eyes glaze over), but there might be if you refuse to try to make even the most minimal effort to understand, and instead prefer to remain uninformed and disseminate information that isn't relevant as if it were relevant, which only makes the issue harder to understand for everyone else. There is no question that you are apparently indeed deluded into thinking that what you state above is possible for DTV. In theory, it is. In the real world of DBS delivery, it really just isn't. I'm sorry that DTV or anyone else can't make it possible for you, but you're just going to have to accept that.

As far as anyone beyond the post-production phase for the production houses that supply HD to television networks, stations, to cable and satellite networks, and to cable and satellite vendors such as DTV, the original material is already in MPEG-2 format. Most content is compressed even before it can be edited into a program. This means that DTV's only source for LIL HD channels is content that is already MPEG-2 encoded and also severely reduced in bit rate by SMPTE310 restrictions, and the same applies to regional sports channels and HD cable networks that DTV turns around. This also means they must re-encode it and suffer the inevitable second-generation rounding errors if they want to squeeze it into M4 distribution. There really just is no practical way around that. I wish that there were. The "original high-quality source" as far as DTV is concerned, is already highly-compressed using MPEG-2. Raw uncompressed digital video is just not available to them, or to anybody else.

The HD Lite issue is not a factor. HD Lite does not increase artifacts, it only reduces potential resolution, and only of original 1920x1080 content (which is rare) and only a fraction of the time (when the content actually achieves that level of resolution). There's a huge difference between potential resolution and achieved resolution, and there's also a huge difference between artifact issues and resolution issues, which are two wholly different things. HD Lite actually reduces artifacting for equivalent payload rates, which is exactly why they do it in the first place.

None of this is opinion or theory. It is all factual. Choose not to accept it or not to believe it at your own risk.

[Jebberwocky!]

There is an old Chinese saying: "A new latrine always smells better for the first few days" :)


I usually break them in rather quickly. :D

[generalpatton71]

The HD Lite issue is not a factor. HD Lite does not increase artifacts, it only reduces potential resolution, and only of original 1920x1080 content (which is rare) and only a fraction of the time (when the content actually achieves that level of resolution). There's a huge difference between potential resolution and achieved resolution, and there's also a huge difference between artifact issues and resolution issues, which are two wholly different things. HD Lite actually reduces artifacting for equivalent payload rates, which is exactly why they do it in the first place.

None of this is opinion or theory. It is all factual. Choose not to accept it or not to believe it at your own risk.

Sorry I got to disagree. While many people point to the res when referring to HD Lite it's the bitrate that makes HD Lite truly HD Lite. I agree that reducing res reduces artifacts, but I don't agree with your definition of HD Lite. I have a 720p set and I can tell you I shouldn't be able to tell the difference between HBO at 1280x1080i with D* and my local CBS at full 1080i because my set can only display 1280x720, but I can because the bitrate is so low on HBO with D*.

[rcliff]

I'm not disputing anything you are saying here, except the part about whether you are delusional or not. I'm also not saying there's anything wrong with not understanding the fine points of something (investment banking makes my eyes glaze over), but there might be if you refuse to try to make even the most minimal effort to understand, and instead prefer to remain uninformed and disseminate information that isn't relevant as if it were relevant, which only makes the issue harder to understand for everyone else. There is no question that you are apparently indeed deluded into thinking that what you state above is possible for DTV. In theory, it is. In the real world of DBS delivery, it really just isn't. I'm sorry that DTV or anyone else can't make it possible for you, but you're just going to have to accept that. The name calling is simply not needed here and much of what you say is simply not true regardless of how many times you say it. And as far as making this issue more difficult for everyone to understand, it is you that continues to incorrectly refer to DirecTV as "DTV" instead of D*. FYI, in case you are unaware, "DTV" is the acronym for Digital Television not DirecTV.

[pendragn]

And as far as making this issue more difficult for everyone to understand, it is you that continues to incorrectly refer to DirecTV as "DTV" instead of D*. FYI, in case you are unaware, "DTV" is the acronym for Digital Television not DirecTV.
I'm not taking Tyrone's side, but I think enough people around here know that DTV can be shorthand for "DirecTV" and that combined with the context of the sentence usually makes it easy to know which is being used. In all honesty I don't see a lot of people talk about Digital TV, I see DTV used for DirecTV more.

tk

[TyroneShoes]

Sorry I got to disagree. While many people point to the res when referring to HD Lite it's the bitrate that makes HD Lite truly HD Lite. I agree that reducing res reduces artifacts, but I don't agree with your definition of HD Lite. I have a 720p set and I can tell you I shouldn't be able to tell the difference between HBO at 1280x1080i with D* and my local CBS at full 1080i because my set can only display 1280x720, but I can because the bitrate is so low on HBO with D*."HD Lite" is only a colloquial term so has no true definition. But the term arose right here on these forums, where the generally accepted definition refers to 1280x1080 resolution used by DTV, and not to reduced bit rates, which is really a separate manner.

There is certainly no intention to confuse folks, at least not on my part, but I'm happy to apologize if that has backfired. On the contrary, I hoped by staying within the locally-accepted definition, that we could actually clear up some confusion, such as confusing bit rate issues with resolution issues.

Likewise, the accepted definition of "DTV" on this forum is "DirecTV", and not "digital TV", which in the industry is referred to as "DT". If folks are confused by that, I can't really help them. The terms "D*" and "E*" for DirecTV and EchoStar got co-opted long ago, and folks became confused and started using "D*" to refer to "DISH Network", so many of us abandoned that term in the interests of less confusion, as well, for the common "DTV". When in Rome....

[TyroneShoes]

The name calling is simply not needed here and much of what you say is simply not true regardless of how many times you say it... Hmmm. Sounds "simply" like "I'm rubber, you're glue...."

I'm afraid the only thing I will be "simply" apologizing for to you for is that I really have nothing to apologize for. I will certainly not be goaded into apologizing for something I did not do. No one here called you a "name", at least not yet. I did refer to you as both "delusional" and "deluded", which is what IMHO, you most certainly are. That is not calling you a name. If I had called you a name, there really would be no doubt about it, but I have no reason to sink to the level of name-calling, and I have not. Intelligent folks can typically understand the distinction, which is not really all that fine. I have no beef with you or anyone else. Other than your intransigent position on something you apparently know very little about, you seem like a regular, even possibly a likable person. I'm trying to keep an open mind about that and am willing to give you every chance to convince me of that. So let's try to move beyond all of this.

A delusion, according to my dictionary, is "an idiosyncratic belief or impression that is firmly maintained despite being contradicted by what is generally accepted as reality or rational argument..."

If that doesn't describe the state you find yourself in, I don't know what possibly could, because I could not have described it anywhere near that elegantly.

That does not refer to your person in a derogatory manner and is not meant as a slight in any way, only as a statement of how I perceive your point of view regarding this subject, so please try not to think of it in that way. The human condition is that we are all deluded on occasion, and that we eventually, or hopefully, attain enlightenment. There's nothing wrong with being human or being subjected to human processes. And if someone actually does understand the issue, that doesn't make them any better than you, just better-informed. No one is looking down on you or pointing fingers. The only shame would be in refusing to evolve beyond a deluded state. I am only trying to help you attain enlightenment.

I've been wrong before and I will be wrong in the future, most likely. Again, there is no shame in being wrong, only in holding to beliefs out of stubborness or spite. But I am just not wrong about this, as these principles are part of the bedrock of modern broadcast engineering.

You can even look it up if you don't believe me or are actually motivated to learn something beyond what you think you might know, although I really don't care and won't entertain further childishness regarding the subject. A good place to start might be "The Standard Handbook of Video and Television Engineering" by Whittaker and Benson, which is widely held as the final-word bible of broadcast engineers, and which spells out in plain english beyond any doubt just how right I am. You don't have to commit the principles to memory as do those of us who make their living abiding by them. All you have to do is "simply" read it.