Jump to content


Welcome to DBSTalk


Sign In 

Create Account
Welcome to DBSTalk. Our community covers all aspects of video delivery solutions including: Direct Broadcast Satellite (DBS), Cable Television, and Internet Protocol Television (IPTV). We also have forums to discuss popular television programs, home theater equipment, and internet streaming service providers. Members of our community include experts who can help you solve technical problems, industry professionals, company representatives, and novices who are here to learn.

Like most online communities you must register to view or post in our community. Sign-up is a free and simple process that requires minimal information. Be a part of our community by signing in or creating an account. The Digital Bit Stream starts here!
  • Reply to existing topics or start a discussion of your own
  • Subscribe to topics and forums and get email updates
  • Send private personal messages (PM) to other forum members
  • Customize your profile page and make new friends
 
Guest Message by DevFuse

Photo
- - - - -

A 2010 compendium of D* DVR bugs


  • Please log in to reply
194 replies to this topic

#161 OFFLINE   Steve

Steve

    Hall Of Fame

  • DBSTalk Club
  • 22,429 posts
Joined: Aug 22, 2006

Posted 27 May 2010 - 08:32 AM

Does a DVR track frame rates or does it treat it simply as data? [...]

Video is encoded to a frame rate, but I believe trickplay always needs to land on a "key" frame, which is the first frame in a sequence of frames that whatever compression is being used determines is most important to encode with as little loss of quality as possible.

That's why it sometimes appears after stopping FFX1, which has no autocorrection applied to it, that the video still jumps backwards before resuming play. I believe that happens when the closet key frame to the stopping point is before the resume point, and the HR needs to "back-up" to find it. Otherwise, if it always jumped to the next keyframe, there would be a risk of a dropped word of dialog.

MPEG-2 by it's nature is encoded with more key frames than MPEG-4, which is why MPEG-2 file sizes are larger. As a result, tho, MPEG-2 trickplay is smoother.

I think trickplay is an area where the HR's were being unfairly compared to the DirecTiVo, because the HR10 only had to deal with MPEG-2 files. Comparing MPEG-2 to MPEG-4 trickplay is really comparing apples to oranges.
/steve

...Ads Help To Support This SIte...

#162 OFFLINE   Syzygy

Syzygy

    Hall Of Fame

  • Topic Starter
  • Registered
  • 1,306 posts
Joined: Dec 05, 2005

Posted 27 May 2010 - 03:45 PM

MPEG-2 by it's nature is encoded with more key frames than MPEG-4, which is why MPEG-2 file sizes are larger. As a result, tho, MPEG-2 trickplay is smoother.

I think trickplay is an area where the HR's were being unfairly compared to the DirecTiVo, because the HR10 only had to deal with MPEG-2 files. Comparing MPEG-2 to MPEG-4 trickplay is really comparing apples to oranges.

As I recall, trick play on my HR21 behaved the same with the rare MPEG-2 recording as it did with MPEG-4.

Are there any MPEG-2 channels left for an HR2x to experiment with? T101, say?
Frank TiVangelist since Aug 1999, HD since Dec 2002, DirecTV since Aug 2004, DECA/MRV since Nov 2010
HR24-100 > HDMI > Pio 5020FD plasma
HR24-100 > HDMI > Sony KV30HS420 tube

DirecTV & Dish HD Channel Spreadsheet
DirecTV & Dish HD Channels Listed by Package

#163 OFFLINE   LameLefty

LameLefty

    Hall Of Fame

  • Registered
  • 12,043 posts
Joined: Sep 28, 2006

Posted 27 May 2010 - 03:48 PM

As I recall, trick play on my HR21 behaved the same with the rare MPEG-2 recording as it did with MPEG-4.

Are there any MPEG-2 channels left for an HR2x to experiment with? T101, say?


Record a local OTA.

"Human sacrifice, dogs and cats living together... mass hysteria!"
Directv since 1997
Will Work for Beer


#164 OFFLINE   Steve

Steve

    Hall Of Fame

  • DBSTalk Club
  • 22,429 posts
Joined: Aug 22, 2006

Posted 27 May 2010 - 04:33 PM

As I recall, trick play on my HR21 behaved the same with the rare MPEG-2 recording as it did with MPEG-4 [...]

You haven't watched enough MPEG-2 then. :) As Lefty says, try an OTA recording. Trickplay is smooth as silk. I record NFL games OTA whenever possible for just that reason.
/steve

#165 OFFLINE   drpjr

drpjr

    Icon

  • Registered
  • 722 posts
Joined: Nov 23, 2007

Posted 27 May 2010 - 07:57 PM

Well, we're lost together. We can blow up a house in Berzerkistan from a console in California, but we can't measure 30 seconds accurately on a DVR? Oh well, I really didn't mean to start an argument, just commenting on an ongoing MINOR annoyance, but now I'm with you, befuddled again. :lol:

Rich


Befuddled? Can I get an Amen Brother?:D This is something I would never have noticed on my own. Good catch by the way. It reminds me of the picture intermittantly flashing through the screensaver, it doesn't effect function and it doesn't bother me in the least, but I want to know why it does it. After reading Steve's explanation I feel that I have a pretty good grasp of the concepts involved. I just keep coming up with a bunch of "yeah...but" questions.:lol:
HR34/HR24-500/Mits WD73C8
HR20-100 HR24-500//Samsung UN55F6350AF
SWiM Network w/MRV
AT&T 2Wire Gateway Modem/Router

#166 OFFLINE   Mike Bertelson

Mike Bertelson

    6EQUJ5 WOW!

  • Moderators
  • 13,824 posts
Joined: Jan 24, 2007

Posted 27 May 2010 - 08:42 PM

Well, we're lost together. We can blow up a house in Berzerkistan from a console in California, but we can't measure 30 seconds accurately on a DVR? Oh well, I really didn't mean to start an argument, just commenting on an ongoing MINOR annoyance, but now I'm with you, befuddled again. :lol:

Rich

I’m guessing here, but I think it has little to do with "can't" and more to do with the amount of work for so little a benefit. How much coding and how many more modules would you need to add to the firmware to get exactly 30 seconds vs. 30-33 seconds? Is it really worth it to do all that work when, for 99%, of users, the extra 1-3 seconds isn’t really going to matter? IMHO, I would say no as I feel there are much more important things to spend time and lines of code on.

My 2¢ FWIW. :)

Mike

µß
Since it costs 2.4¢ to produce a penny, my 2¢ worth is really 4.8¢ worth.  That 4.8¢ is my own and not the 4.8¢ of DIRECTV, Dish, or anyone else for that matter.


#167 OFFLINE   bonscott87

bonscott87

    Cutting Edge: ECHELON '07

  • Registered
  • 9,809 posts
Joined: Jan 21, 2003

Posted 27 May 2010 - 08:43 PM

As I recall, trick play on my HR21 behaved the same with the rare MPEG-2 recording as it did with MPEG-4.

Are there any MPEG-2 channels left for an HR2x to experiment with? T101, say?


Yep, MPEG2 on an HR2x is just as smooth as the DirecTivo. Trickplay "choppiness" is due to MPEG4. Almost a guarantee that this new DirecTivo will have the same issue. Just a nature of the technology.

#168 OFFLINE   Syzygy

Syzygy

    Hall Of Fame

  • Topic Starter
  • Registered
  • 1,306 posts
Joined: Dec 05, 2005

Posted 27 May 2010 - 09:03 PM

Are there any MPEG-2 channels left for an HR2x to experiment with? T101, say?

Record a local OTA.

As Lefty says, try an OTA recording.

Sorry, I can't. Not all of us have OTA tuners (HR20 or AM21). ;)
Frank TiVangelist since Aug 1999, HD since Dec 2002, DirecTV since Aug 2004, DECA/MRV since Nov 2010
HR24-100 > HDMI > Pio 5020FD plasma
HR24-100 > HDMI > Sony KV30HS420 tube

DirecTV & Dish HD Channel Spreadsheet
DirecTV & Dish HD Channels Listed by Package

#169 OFFLINE   Syzygy

Syzygy

    Hall Of Fame

  • Topic Starter
  • Registered
  • 1,306 posts
Joined: Dec 05, 2005

Posted 27 May 2010 - 09:04 PM

Yep, MPEG2 on an HR2x is just as smooth as the DirecTivo. Trickplay "choppiness" is due to MPEG4. Almost a guarantee that this new DirecTivo will have the same issue. Just a nature of the technology.

OK, I believe. Too bad, tho.
Frank TiVangelist since Aug 1999, HD since Dec 2002, DirecTV since Aug 2004, DECA/MRV since Nov 2010
HR24-100 > HDMI > Pio 5020FD plasma
HR24-100 > HDMI > Sony KV30HS420 tube

DirecTV & Dish HD Channel Spreadsheet
DirecTV & Dish HD Channels Listed by Package

#170 OFFLINE   Syzygy

Syzygy

    Hall Of Fame

  • Topic Starter
  • Registered
  • 1,306 posts
Joined: Dec 05, 2005

Posted 27 May 2010 - 09:12 PM

I switched to 30SLIP for testing and found that "30-second slip" always slips between 32 and 33 seconds. I used the HDNet test pattern [HD] to time it.

I switched back to 30SKIP and (again with the HDNet test pattern) found that one "30-second skip" always skips between 31 and 32 seconds, while a group of skips covers less time than expected; each successive skip tacked onto the chain of skips goes forward about 28 seconds.

Conclusions:
1. SLIP's extra 2 or 3 seconds (5 or 6 seconds in another report) is mainly due to slugglish response (which seems to affect each SLIP in a chain of SLIPs).
2. SKIP's nearly-correct timing (for a single keypress) is the result of combining a programmed amount of around 28 seconds with the somewhat variable slugglishness. However, succeeding SKIPs in a chain are shorter because the slugglishness (latency) has little or no effect on them.
3. The average duration of a "30-second" SKIP can provide a rough estimate of a particular unit's latency.

Edited by Syzygy, 28 May 2010 - 04:17 PM.

Frank TiVangelist since Aug 1999, HD since Dec 2002, DirecTV since Aug 2004, DECA/MRV since Nov 2010
HR24-100 > HDMI > Pio 5020FD plasma
HR24-100 > HDMI > Sony KV30HS420 tube

DirecTV & Dish HD Channel Spreadsheet
DirecTV & Dish HD Channels Listed by Package

#171 OFFLINE   Rich

Rich

    Hall Of Fame

  • DBSTalk Club
  • 20,192 posts
  • LocationPiscataway, NJ
Joined: Feb 22, 2007

Posted 28 May 2010 - 07:33 AM

Video is encoded to a frame rate, but I believe trickplay always needs to land on a "key" frame, which is the first frame in a sequence of frames that whatever compression is being used determines is most important to encode with as little loss of quality as possible.

That's why it sometimes appears after stopping FFX1, which has no autocorrection applied to it, that the video still jumps backwards before resuming play. I believe that happens when the closet key frame to the stopping point is before the resume point, and the HR needs to "back-up" to find it. Otherwise, if it always jumped to the next keyframe, there would be a risk of a dropped word of dialog.

MPEG-2 by it's nature is encoded with more key frames than MPEG-4, which is why MPEG-2 file sizes are larger. As a result, tho, MPEG-2 trickplay is smoother.

I think trickplay is an area where the HR's were being unfairly compared to the DirecTiVo, because the HR10 only had to deal with MPEG-2 files. Comparing MPEG-2 to MPEG-4 trickplay is really comparing apples to oranges.


That does make it much clearer. I've been getting some long jump backs that look like auto-correction and I thought that was due to having the Ethernet MRV setup. Sometimes it jumps back a minute or two. Rather disconcerting when you don't know what's causing it. I feel better knowing it's normal. Normal I can deal with. :)

Rich

#172 OFFLINE   Rich

Rich

    Hall Of Fame

  • DBSTalk Club
  • 20,192 posts
  • LocationPiscataway, NJ
Joined: Feb 22, 2007

Posted 28 May 2010 - 07:44 AM

Befuddled? Can I get an Amen Brother?:D This is something I would never have noticed on my own. Good catch by the way. It reminds me of the picture intermittantly flashing through the screensaver, it doesn't effect function and it doesn't bother me in the least, but I want to know why it does it. After reading Steve's explanation I feel that I have a pretty good grasp of the concepts involved. I just keep coming up with a bunch of "yeah...but" questions.:lol:


Sometime in Steve's life he must have learned that to reach people with information that they can comprehend you have to get down to a level that they can understand. His posts have always been clear and (relatively) easy to understand.

If you want everyone to understand the message, you must get down to a fifth or sixth grade level. For the younger generations who don't read I don't know how far down you'd have to go. That's not an insult aimed at younger folks, that's an indictment of our school systems. Just go to any school and ask the students if they've read Moby Dick. I did this many times while my son (who never read it) was growing up and found one person who had read it. And he was from Sierra Leone. But, they've all seen the movies. :)

Rich

#173 OFFLINE   Rich

Rich

    Hall Of Fame

  • DBSTalk Club
  • 20,192 posts
  • LocationPiscataway, NJ
Joined: Feb 22, 2007

Posted 28 May 2010 - 07:48 AM

I’m guessing here, but I think it has little to do with "can't" and more to do with the amount of work for so little a benefit. How much coding and how many more modules would you need to add to the firmware to get exactly 30 seconds vs. 30-33 seconds? Is it really worth it to do all that work when, for 99%, of users, the extra 1-3 seconds isn’t really going to matter? IMHO, I would say no as I feel there are much more important things to spend time and lines of code on.

My 2¢ FWIW. :)

Mike


I can't really think of anything but fights where you'd notice the inaccuracy of the slip and, as I said, it's really a minor annoyance. And it's easily correctable with a quick jump back. If the TS hadn't started this thread, I would never have commented on it.

Rich

#174 OFFLINE   Steve

Steve

    Hall Of Fame

  • DBSTalk Club
  • 22,429 posts
Joined: Aug 22, 2006

Posted 28 May 2010 - 08:08 AM

That does make it much clearer. I've been getting some long jump backs that look like auto-correction and I thought that was due to having the Ethernet MRV setup. Sometimes it jumps back a minute or two. Rather disconcerting when you don't know what's causing it. I feel better knowing it's normal. Normal I can deal with. :)

I do want to say that what I was talking about (finding a key frame) only applies to FFX1, where there isn't any intentional autocorrection applied, AFAIK, and searching for a key frame sometimes takes you backwards a second or two.

If autocorrection seems too long on FFX2 or FFX3, that may be either because your reflexes are above the average user for which they calculated what the correction time should be, or the s/w release you're on may be beta or still needs to be tuned better for local vs. MRV autocorrect. I generally don't see autocorrect issues, because I'm a 30-skipper. I use that to get past MRV commercials and it works flawlessly for me.

Edited by Steve, 28 May 2010 - 08:19 AM.

/steve

#175 OFFLINE   Steve

Steve

    Hall Of Fame

  • DBSTalk Club
  • 22,429 posts
Joined: Aug 22, 2006

Posted 28 May 2010 - 08:17 AM

I switched back to 30SKIP and (again with the HDNet test pattern) found that one "30-second skip" always skips between 31 and 32 seconds, while a group of skips covers less time than expected; each successive skip tacked onto the chain of skips goes forward about 28 seconds [...]

This may be by design, to compensate for "looking" time in between SKIPS. If you're like me, I stack 4-5 right away, and then pause to see where I'm at before clicking SKIP again.
/steve

#176 OFFLINE   Rich

Rich

    Hall Of Fame

  • DBSTalk Club
  • 20,192 posts
  • LocationPiscataway, NJ
Joined: Feb 22, 2007

Posted 28 May 2010 - 09:14 AM

I do want to say that what I was talking about (finding a key frame) only applies to FFX1, where there isn't any intentional autocorrection applied, AFAIK, and searching for a key frame sometimes takes you backwards a second or two.


Ah, even clearer. I do use FFX2, 3 and 4 for various reasons and they all react the same way.

If autocorrection seems too long on FFX2 or FFX3, that may be either because your reflexes are above the average user for which they calculated what the correction time should be, or the s/w release you're on may be beta or still needs to be tuned better for local vs. MRV autocorrect. I generally don't see autocorrect issues, because I'm a 30-skipper. I use that to get past MRV commercials and it works flawlessly for me.


I'm not on beta and I usually am on MRV. I do usually use the HR that the program was recorded on. I also do go to "Local" if I'm gonna watch a couple programs on a particular HR. I think that works better than leaving the MRV on. Might be my imagination. The only reason I use the slip is that I was missing too many relevant commercials with the skip. Such as new shows, that sort of thing.

I can usually stop the FFx right on the place that I want to be, but it goes back and I have to slip forward. Not a big deal. Something that's been happening to me long before MRV came on line. Does seem to have gotten a tad worse lately tho.

Rich

#177 OFFLINE   drpjr

drpjr

    Icon

  • Registered
  • 722 posts
Joined: Nov 23, 2007

Posted 28 May 2010 - 11:44 AM

I’m guessing here, but I think it has little to do with "can't" and more to do with the amount of work for so little a benefit. How much coding and how many more modules would you need to add to the firmware to get exactly 30 seconds vs. 30-33 seconds? Is it really worth it to do all that work when, for 99%, of users, the extra 1-3 seconds isn’t really going to matter? IMHO, I would say no as I feel there are much more important things to spend time and lines of code on.

My 2¢ FWIW. :)

Mike


I agree completely. I am one of the 99% that thinks it really doesn't matter. And I would bet most of the remaining 1% only care about accuracy where my interest lies in the mechanics of the function. I realize this type of esoteric discussion has very limited appeal but I am thankful there are some who will share their knowledge. Thanks Mike
HR34/HR24-500/Mits WD73C8
HR20-100 HR24-500//Samsung UN55F6350AF
SWiM Network w/MRV
AT&T 2Wire Gateway Modem/Router

#178 OFFLINE   Rich

Rich

    Hall Of Fame

  • DBSTalk Club
  • 20,192 posts
  • LocationPiscataway, NJ
Joined: Feb 22, 2007

Posted 28 May 2010 - 11:56 AM

I agree completely. I am one of the 99% that thinks it really doesn't matter. And I would bet most of the remaining 1% only care about accuracy where my interest lies in the mechanics of the function. I realize this type of esoteric discussion has very limited appeal but I am thankful there are some who will share their knowledge. Thanks Mike


I never thought that post I wrote would stir up so many posts. :lol:

Rich

#179 OFFLINE   drpjr

drpjr

    Icon

  • Registered
  • 722 posts
Joined: Nov 23, 2007

Posted 28 May 2010 - 12:30 PM

Video is encoded to a frame rate, but I believe trickplay always needs to land on a "key" frame, which is the first frame in a sequence of frames that whatever compression is being used determines is most important to encode with as little loss of quality as possible.

If I'm understanding correctly a more static picture= less encoding= less "key" frames= less accuracy in trickplay and the opposite for a more motion picture. If so that would fit Rich's observation while watching talking heads between boxing rounds.

If autocorrection seems too long on FFX2 or FFX3, that may be either because your reflexes are above the average user for which they calculated what the correction time should be, or the s/w release you're on may be beta or still needs to be tuned better for local vs. MRV autocorrect. I generally don't see autocorrect issues, because I'm a 30-skipper. I use that to get past MRV commercials and it works flawlessly for me.


I think your observation of reflexes is a good one. No matter what s/w version, beta, local server, client or whatever autocorrection is always too short for my wife and too long for me. I am convinced the only constant is reflex time. Steve thanks for all the easily digestable info.
HR34/HR24-500/Mits WD73C8
HR20-100 HR24-500//Samsung UN55F6350AF
SWiM Network w/MRV
AT&T 2Wire Gateway Modem/Router

#180 OFFLINE   Rich

Rich

    Hall Of Fame

  • DBSTalk Club
  • 20,192 posts
  • LocationPiscataway, NJ
Joined: Feb 22, 2007

Posted 30 May 2010 - 07:41 AM

This may be by design, to compensate for "looking" time in between SKIPS. If you're like me, I stack 4-5 right away, and then pause to see where I'm at before clicking SKIP again.


Steve, after reading your posts about the skip, I considered how often you've pointed me in the right direction. And I also remembered that some time in the past they upgraded the skip function. So I tried it. And now I'm changing all my HRs to skip. So much faster now it's almost like using a TiVo again. My thanx for making me reconsider. Another paradigm shot in the butt! :lol:

Rich




spam firewall