1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

HR20 ; Live TV vs Buffered TV

Discussion in 'DIRECTV HD DVR/Receiver Discussion' started by Supervolcano, Mar 8, 2007.

Thread Status:
Not open for further replies.
  1. Mar 8, 2007 #1 of 33

    Supervolcano Hall Of Fame

    Jan 23, 2007
    If there's been a deep discussion about this, forgive me, but let's go deeeeeep into this one gang!!

    No comments are unwelcome!!

    Milo, brott, tibber, and I were having a discussion the other day and I think it's time that everyone starts talking, thinking, and brainstorming about this one as a big group.

    Milo's survey points to watching "Live TV" as being a potential place where the unexpected seems to occur the most. Sometimes minor things and sometimes major things. Anywhere from total lockups, trickplay issues, audio drops and crackling, pixilating, to even stranger phenomenons like finding pixilation in a recording when you previously watched the live broadcast and everything was fine. Only God knows what other possible little issues people have seen surrounding "Live TV" non-trickplayed.

    Over the years I used to have 2 tivo's (1 hughes sd-dvr80 and 1 HR10-250) plus I currently have another Hughes SD-DVR80 and ALL of them occasionally suffered from "audio studdering" now and then. Recordings for me never suffered, just live tv.

    The simple answer for me, without ever reading a forum thread about it back then, was to simply hit the tivo's 6 second back button one time and all my audio problems would magically disappear. It became a habit for me which I still have.

    So then last night tibber and I got to talking in chat about this a little more with a couple other guys and tibber mentioned something I didn't know since i don't have 2 receivers side by side anymore .... He said that when you switch channel on HR20, you are watching "truely live tv", but if you trickplay back and then forward, then your suddenly NOT watching live tv anymore (and let's leave out that 30 sec slip bug because that's not what we're talking about here) .... this proven (he said) by comparing the broadcast to a H20 non-dvr.

    So now we have 3 types of watchable tv.
    "Truely Live TV"
    "End of Buffered Live TV"
    "Buffered TV"

    In the 5 weeks I've had the HR20, I've watched a LOT of live tv on it, but I still find myself addicted to that 6 sec back (and pause button too) so I'm rarely watching "truely live tv" even when I fast forward back to what i thought was live tv .... what I now call "buffered tv" and "end of buffered live tv"!!!

    Maybe I just got the luck of the draw and got a great receiver, could very well be because i have rarely had problems, but I CAN SAY that the few times I've had to reboot from lockups or whatnot .... The receiver was in fact watching "truely live tv".

    Then I got to thinking...
    None my tivo's EVER showed "truely live tv"!!! They always had a built in buffer as my broadcast was always about 2 seconds or so behind my old non-dvr's. I know this because I had 4 tv's (3 receivers and an ota feed) in my old living room FOR YEARS and I often had all of them going.

    Theory Time!!!
    1. Could a lot of the HR20's problems be tied right to this one .... dare i say .... BOTTLENECK of trying to do to much work all at the same time simply to get your DVR's and your NON-DVR's playing a perfectly simulcasted broadcast?
    2. If there was a built-in predefined full-time 2 second buffer, could this help the processor and memory to keep up with what it needs to do to deliver us a perfect seamless signal/decryption full-time?
    3. Wouldn't this senario allow for one set of functions to be running full-time as opposed to (speculating here, don't sue me if this isn't the case) possibly different sets of functions as we switch from "truely live tv" over the "end of buffered tv" ... because if this is the case, any programmer knows software jumps take more processor time the further away the functions are from each other inside the code?
    The object here is "efficiency".
    The more efficient the receiver runs, the more stable it can run AND the better our chances get for more complex tasks being implemented like for example ....dare I say .... DLB?
    • Is it purely fixable code bugs that are the problem?
    • Is the processor starved for speed?
    • Is there not enough memory in the unit?
    • Are there too many memory leaks?
    Remember guys, I'm brainstorming and theorizing.
    Feel free to correct me and/or build upon me.
    Any takers?
  2. Mar 8, 2007 #2 of 33

    dallascontractor DBSTalk Club Member

    Jan 9, 2007
    Are you sure the deal between ota and the sat you are kind on talking about is not sat to earth delay?
    I was watch Mavericks Tuesday on 643 on two different TV's. The main TV I would back up to see plays and FF. I was noticing that the TV in the office that had not been backed up was about 2-3 seconds in front in show than the TV in Den that was backed up, never could get it back to same speed without changing channel and then going back. FF in any spped and 30 second foward to not get back to live TV without changing channels.
  3. Mar 8, 2007 #3 of 33

    jasonblair Icon

    Sep 5, 2006
    i just got my HR20 and H20 to upgrade from my H10 and HTL-HD. (I moved far enough out that I couldn't get HD locals OTA.)

    I was REALLY surprised to see that the TV with my HR20 was perfectly timed to my TV with the H20. I always thought DVRs added a buffer, and therefore the image lagged from non-DVRs.

    Somehow the HR20 doesn't lag. Supervolcano might have a point... but, I really like that the TV's are synched. Sometimes I'm doing work that requires I move back and forth between rooms, and it's nice if there's no "echo" when I'm watching the same show on both TVs.
  4. Mar 8, 2007 #4 of 33

    mtnagel Hall Of Fame

    Sep 18, 2006
    I always knew that the jump to live thing didn't really take you to live, but what about if you do that and then FF more? I always figured that did bring you to truly live tv. Maybe not? So the only way to be at truly live tv is through a channel change?
  5. Mar 8, 2007 #5 of 33

    paulman182 Hall Of Fame

    Aug 4, 2006
    Earth-to-sat-to-earth trip is roughly a quarter-second.
    There are unknown delays in D*'s encoding/decoding process.
  6. Mar 8, 2007 #6 of 33

    Supervolcano Hall Of Fame

    Jan 23, 2007
    OTA vs. Satellite plays no part in this discussion.
    Sorry i included the word OTA as being used on one of my tv's.
    I was simply talking about my setup.

    OTA is always going to get to you faster than satellite, not only because of the reason you stated, but also because of the time it takes for the signal to get from the original broadcaster to D* and the time it takes for D* to do it's magic with the signal in their house.

    Precisely what tibber told me.
    That's what I'm referring to.

    I agree it's kind of a nice "feature", but at what "cost" does that feature come?

    Can performance and stability be improved if we sacrifice this ability like tivo did?

    If so, I would vote for better performance and stability any day of the week.

    According to tibber, and now dallascontractor, apparently that answer is YES, a channel change is the only way back to truely live tv.
  7. Mar 8, 2007 #7 of 33

    Supervolcano Hall Of Fame

    Jan 23, 2007
    Let me add one more perverbial monkey wrench to all this which might enlighten some people.

    Have you ever been sitting at truely live tv (as described above) after a channel change and you decide to press rewind or 6 second back or the 15 minute skip to tick backwards and the receiver suddenly freezes on a frame for a couple seconds before it begins the trickplay function?

    Have you ever been in "buffered tv" mode already and the rewind functions don't seem to have the same "lag" that they did when it was in "truely live tv" mode?

    I believe this is a direct result of moving from "truely live tv" into "buffered tv" mode.

    It's like the receiver is requiring time to get off of live tv and switch the set of FUNCTIONS being used and/or other things in memory before it can begin processing your button request.

    This is often reported as a "bug" in the issues thread when I believe this is simply the way it's designed to operate.

    If we were always in the "end of buffered tv" instead of "truely live tv" upon a channel change, this delay before the button press would probably disappear .... which would result in better performance and stability for the display of rewinding frames.

    More food for thought....
  8. Mar 8, 2007 #8 of 33

    cygnusloop Hall Of Fame

    Jan 26, 2007
    SV- You are definitely on to something here. There is no question that the unit is flaky when chasing close to live, and when watching "live" whatever that means. I'll even go you one better. All my unwatchable recordings (which were numerous, and are now non-existent since about x12a) happened when the HR20 was in what you are calling "Truely Live TV" as opposed to buffered or end of buffered. My theory on the unwatchables always was that it was some kind of switching issue.

    I agree with you that "live" for my DVR40 DTiVo is always what you are calling "end of buffered TV". I also agree that getting the logic and the data handling right in this area is THE key to making DLB's a reality on the HR20. As tibber has eloquently pointed out, ALL the hardware and system capability is in place for DLB's. I would also posit that once you get the data handling robust enough to handle DLB's, alot of the other flakiness will go away as well (negative time, having to RW to get to the start of the show, IKD's, unwatchables). I think all of these bugs are (were) related to the issue that you have raised here, SV. D* has done a good job of patching most of these bugs, but I wonder if the root cause has truly been addressed. (The "root cause" that I am trying to describe here is poor data handing, flaky switching, and the HR20 seemingly not knowing exactly where it is in time on occasion). I think that even the new 30sec slip bug is a symptom that grows out of this same root cause.

    Here's a question: Does the system behave the same when bonking up against live TV when you are chasing a currently recording program? (i.e. green time bar vs. orange time bar)?

    Anyway, good thread! Let's kick this one around for a while.
  9. Mar 8, 2007 #9 of 33

    HarleyD Hall Of Fame

    Aug 31, 2006
    OK, I officially have a headache now.
  10. cygnusloop

    cygnusloop Hall Of Fame

    Jan 26, 2007
    suck it up, Harley! :lol:
  11. SuperTech1

    SuperTech1 Icon

    Jan 8, 2007
    This whole discussion reminds of CD players, satellite radio, etc.
    Remember the first portable or car CD players? One little bump and you lost audio. At first they tried shock/isolation which helped but then some bright person decided - Hey why not buffer the audio stream? You're listening a few seconds behind but then any bumps or interruptions are absorbed by the buffer (up to a point). Then after the interruption you refill the buffer.
    When I listen to my satellite radio and go through a tunnel or under a long overpass I don't immediately lose the sound. Why? It's buffered of course!
    Who needs live!!:)
  12. Coffey77

    Coffey77 Cutting Edge: ECHELON '07

    Nov 12, 2006
    This is a great brainstorm Super. I'm marking it so I can read up on it a bit more. Great work so far from what I've read. I'll add my hay-penny when I get a chance. :up: Was it discussed that this is maybe how they're getting around the "legal" issues? A buffer to a buffer...
  13. Supervolcano

    Supervolcano Hall Of Fame

    Jan 23, 2007
    Honestly, I think that is the exact reason why the HR20 "was designed" this way.

    But I feel it was the wrong people who made the decision to implement this.

    I have a feeling all the corporate suits were sitting around a table complaining to each other about how NON-DVR's and DVR's don't sync up.

    I have a feeling some corporate suit walked into the head programmer's office and said "Fix this thing so my HR20 picture will be timed to display the exact same as the H20".

    I have a feeling the head programmer looked at the corporate suit and said "Huh, are you frickin high on drugs? Tivo buffered this stuff for a reason. There's just too much going on to do this in realtime reliably".

    I have a feeling the corporate suit responded with "I don't care what you think. Just do it"!!

    To D*'s engineers & programmers:
    If this is what happened, and the suits forced you or talked you into doing this, please print out this thread and take it to the next board meeting and/or the president's desk. Request that they rethink what they asked you to do.
  14. Koz

    Koz Icon

    Sep 16, 2006
    Great brain-storming SV! I think you've pretty much said everything I can think of. I also agree with your theory on someone too high up asking for the HR20 and H20 to be in sync. Either that or some engineer proposed it and is now having a heckuva time backing up his promise.

    Anyway, they need to take a hard look at the future of the product. Can they fight through the current live tv arrangement and get it to work? Or should they cut their losses and go to buffered "live" tv? I'll let them make that call. (I'm sure we all would.) But I wouldn't be hurt if they sacrificed the current live tv in return for improved performance and reliability.

    The HR20 engineers seem to be pushing the right buttons the last couple months and I'm sure they'll get this one figured out as well.
  15. machavez00

    machavez00 Hall Of Fame

    Nov 2, 2006
    I agree this is what is behind the lack of DLB on the HR20/R15. Is "buffered live tv" patentable and if so does TiVo hold the patent?
  16. Coffey77

    Coffey77 Cutting Edge: ECHELON '07

    Nov 12, 2006
    If I were to look at this from a different angle...

    What if one of the programmers from D* noticed how lagged the Tivo was thought that, if they could make a DVR that does not suffer from Satellite delay-delay it might make people happier and could be a great idea. One of my complaints of the Tivo was the huge lag between it and "LiveTV". (There was never hope of it matching the voice of the Bears...) The concept of having a DVR which was also capable of "normal" viewing is actually quite brilliant and something of the future. Sure, it's not perfect - yet - but that is probably one of the tricky parts of "fixing" the HR20. One little screw-up could cause the whole system to become unstable.

    Don't forget what the main goal is right now of the HR20... Stability!
  17. Stuart Sweet

    Stuart Sweet The Shadow Knows!

    Jun 18, 2006
    I have definitely seen timing issues between my R10 DirecTivo and HR20. I will have do some research on this because I can never recall a time when HR20 was actually ahead of Tivo on Live TV. Perhaps that's because I'm a remote hog and can't keep away from the trickplay long enough.

    As to the "suit-related" source of Live TV being truly live, that's a theory. Here's another one: If you had to buffer 2 seconds of Live TV before showing a picture, then it would take 2 seconds longer to show the picture. Channel change times have already been an issue with the HR20 and I doubt any of us would put up with longer ones.

    In my experience, tibber has been a reliable source for all sorts of information. If he says Live TV is truly live, I believe him, but I have no way to test it, as my D10 is in the garage stored as a backup unit. All I'm using currently are DVRs. It does make you wonder though, if the holdup in channel changes ISN'T buffer related, then what is it? I have Native OFF, so that's not it. SD to SD channel changes are slower than a D10 and if you're just switching to two different live feeds, what's causing that if it's NOT buffering?
  18. Supervolcano

    Supervolcano Hall Of Fame

    Jan 23, 2007
    I've never looked into the patents, but I seriously doubt it. If it is patentable, and was patented by tivo, then Windows Media Player, RealPlayer, QuickTime, and all the other video rendering programs on your PC are violating that patent. Every single video streaming software on the PC buffers your video.

    Like I said in Post #6...

    I agree it's kind of a nice "feature", but at what "cost" does that feature come?

    Can performance and stability be improved if we sacrifice this ability?

    If so, I would vote for better performance and stability any day of the week.

    Some of the "time involved" with native off channel changes is as follows:

    Software has to interact with the hardware:
    Receiver has to decide what LNB & transponder is broadcasting the channel.
    Receiver has to tell multiswitch to tune in that LNB.
    Receiver has to tell LNB to tune in that transponder.
    Receiver has to begin decoding the signal for that channel.

    Smartcard has to interact with the receiver:
    Smartcard has to verify the channel is included with what the customer pays for.
    Smartcard has to verify the channel is ok with the parental controls.
    Smartcard has to verify the program is ok with the parental controls.
    Smartcard has to verify the program is not blacked out in your region.
    And I'm sure there are several more verifications on top of these.

    Native on obviously is a whole different ball of wax so we won't compare those apples to these oranges since the receiver has to interact with the tv.

    OK, I just ran some tests using channels 501-508 on my HR20 vs SD-DVR80 vs Sony-B55 (the B55 non-tivo has gotta be close to a decade old).

    The Sony-B55 Non-Tivo changed channels in an average of 1.8 seconds.
    The SD-DVR80 Tivo changed channels in an average of 2.0 seconds.
    The HR20 Divo changed channels in an average of 2.1 seconds (native off).

    (But channel changing is not an exact science and can vary greatly from channel to channel because of the way smartcards interact with the receiver which I described earlier.)

    Without using any trickplays:
    The audio/video of the SD-DVR80 is a few MICROSECONDS behind the HR20.
    The audio/video of the HR20 is maybe 0.10 to 0.20 seconds behind the Sony-B55.

    Pressing Rewind thru 5 minutes then Fast Forward back to live tv:
    The audio/video of the SD-DVR80 is back to exactly where it should be, maybe .10 to 0.20 seconds behind Sony-B55.
    The audio/video of the HR20 is anywhere from 6 to 16 FULL SECONDS behind the Sony-B55.
    :eek2: :eek2: :eek2: :eek2:

    Sooooooo ... might have to redefine my theories!!!
    These tests put a kink in them, for sure.
  19. Stuart Sweet

    Stuart Sweet The Shadow Knows!

    Jun 18, 2006
    So maybe I'm just dealing with my own perception that channel changes on the D10 were faster, just because everything else in the UI was fairly speedy. That is some great data gathering BTW, supervolcano.
  20. macEarl

    macEarl Godfather

    Jan 1, 2007
    Nah - all they'd have to say is buffering a TV signal and it's unique. I don't believe that TiVo's patented it, but it's confusing to read their patents.

    Great thread, SV! FWIW, I've always been frustrated to be at a point where the show just went into commercial, I didn't hear the last line, skipped back, then skipped ahead and I'm back to the same point in the commercial. I'm in buffer at that point when before, I was in live. Thanks! I always just assumed that "Trickplay" was broken that way and nobuddy cared to discuss the obvious.

    (PS - To call buffer/record playback control "Trickplay" instead of something really clever like PlaybackControl always said to me that no one mature was working that end of it.... I hope whoever invented that name gets their lunch stolen from the corp fridge. Really. And their Tupperware, too.)
Thread Status:
Not open for further replies.

Share This Page