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

KTLA 5 Los Angeles and DVR disc usage

Discussion in 'DIRECTV HD DVR/Receiver Discussion' started by gphvid, Sep 25, 2011.

  1. gphvid

    gphvid Godfather

    498
    0
    Jun 19, 2007
    Thousand...
    Hi, I've noticed that apparently when I record KTLA HD on my DVR, it uses more disk space than a similar show recorded on another channel. Nikita, for instance, will take about 3% whereas Alphas will take maybe 1% or so. Even the other locals don't use as much as KTLA does. KTTV 11 runs Fringe and that uses 1% or so despite them being 720p.

    I'm curious why this happens. If KTLA is running MPEG4 like everyone else, should it use the same amount of space the other channels are using and not more?

    It's not a big deal, but something I noticed. The shows recorded from KTLA are clean as always and nothing appears wrong. It is just the disc space usage, or rather how much disk space opens up once I delete a KTLA show.

    Thoughts? Oh, and sorry if this is a repeat post.
     
  2. Stuart Sweet

    Stuart Sweet The Shadow Knows!

    37,060
    287
    Jun 18, 2006
    Could be a less efficient encoding strategy.
     
  3. gphvid

    gphvid Godfather

    498
    0
    Jun 19, 2007
    Thousand...
    It reminds me of how much space was used by MPEG2 but I'm sure they have long since moved to 4. Just curious why they are the only ones this way. Would this be a DirecTV encoding to the sat that would do this?
     
  4. inkahauts

    inkahauts Well-Known Member

    22,661
    1,103
    Nov 13, 2006
    Are you recording OTA? That's much larger. If not...

    It may not be what you are thinking. You'd really have to play around to check, but if another show is using say 1.49 % and a ktla one is using 1.5%, then you will see 1 vs 2 most likely, if your receiver is empty.

    If it has other programs on it you have to wonder what your dvr is actually at as well.

    If your unit says 10% but is really 10.49% and you loose 1.98%, then it is now 8.51 but will now show 9% making you think it only was using 1%.

    If your unit says 10% but is really 9.51% and you loose 2.02%, then it is now 7.49 but will now show 7% making you think it was using 3%.

    We don't really know how they round, but obviously they must, and in some cases its just a matter of having to. Record 10 seconds on an empty drive and it will show only 99% free, even though we know that didn't actually take up a full 1% of space..

    Add in the few mins some shows run short or long, the fact that some shows may have 2 mins worth of non listed padding, and others have none (suddenly those two things right there can add up to 4 or 5 mins in programming length), 720 vs 1080, amount of compression based on hi movement, low movement, etc, and there is a lot of room for these differences that can make it look like ktla shows are taking up a lot more space, but the really aren't.

    Also, some channels send DirecTV an MPEG4 program already, rather than making them do the conversion to mpeg4 on the fly. On the fly isn't as efficient and will generally take up more space.

    There are so many minute differences that can make something appear much larger, and sometimes it is, sometimes it isn't. But your specific example, I doubt its really as different as it appears.
     
  5. HoTat2

    HoTat2 Hall Of Fame

    7,312
    185
    Nov 16, 2005
    Los...
    Good points;

    And I'd say the effect of integer rounding is likely the cause of what the OP is seeing.

    Interestingly though if DIRECTV for whatever reasons wanted to use less compression on KTLA's signal thereby requiring more recording space, they actually have plenty room to do it.

    According to the recent SB transponder map KTLA has an entire transponder (# 21 from D10 or 12 at 103W) all to itself with no other locals multiplexed with it.

    And for that matter it also has an entire spotbeam, "A4BB," all to itself. One of the few spotbeams in the grid assigned only a single transponder. Most others have 2-3 tps. maximum per spotbeam.
     
  6. gphvid

    gphvid Godfather

    498
    0
    Jun 19, 2007
    Thousand...
    I am going along with the idea that KTLA is running less compression or that they are on the sat with less compression. The percentage free reading on my DVR gains about 1% from most regular 1 hour HD shows but always gains 3% off KTLA shows no matter when or what the show is. This is on deletion.
    And then seeing that the transponder map has KTLA on its own seems maybe to indicate less compression which I assume maybe increase in bandwidth therefore the need of its own tp vs others that are multiplexed? I wouldn't surprise me if so. But it does seem odd to me that KTLA would get such a position over other channels. Seems like a favored channel.

    Interesting case in point is that I started last Sunday with 85% free and after recording many shows and their premieres which included 3 KTLA shows, the DVR reported 63% free. We here watched those shows and several others including stuff off HBO and the channel formerly known as SciFi, and the biggest jump in regained space was when each of the KTLA shows were deleted vs the other shows. 3% for each, vs the others with 1% each. Every time. So I am back to 81% free at the moment (can't watch everything...)

    Just thought it was interesting...
     
  7. skyboysea

    skyboysea Icon

    540
    0
    Nov 1, 2002
    The same thing happens in Seattle where KING 5 consistently uses 3%/h while other locals use <2%/h. My guess was that the encoders are not the same for all channels.
     
  8. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    :confused: :confused:
    While TPs have bandwidth, spot beams don't, which is to say it all comes down the the TP frequencies being used.
     
  9. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    "A few things":
    The DVRs with 300 GB drives will use twice as much "percentage" as DVRs with 500 GB drives.
    You can use DirecTV2PC to check the bit-rates of recordings, by monitoring your network while streaming.
    Local HD channel feeds [to DirecTV] won't be any higher than their OTA FCC limit of 19 Mb/s [MPEG-2].
    1080i "may have" a higher bit-rate than 720p.
    By the nature of MPEG-4, it strips out the repetitive bits between frames of MPEG-2, so it only "sends" the bits that change between frames.

    Now with all of this said, I have had recordings take up twice the space that they "should have".
    I recorded Ken Burn's The National Parks, off my local PBS SAT channel.
    My PBS is 720p, and fed to DirecTV via OTA.
    I notice my free space almost completely disappear. :eek2:
    On my 300 GB DVR each hour took 4%.
    Checking the bit-rate with DirecTV2PC showed 21+ Mb/s of MPEG-4 :confused:
    Airing of this series during prime time, caused video breakups and many posts about this on the forum from all over the country. Airings late night recorded perfectly [here].

    This series has been the only recording that have ever shown this here. YMMV
     
  10. HoTat2

    HoTat2 Hall Of Fame

    7,312
    185
    Nov 16, 2005
    Los...
    Oh, I didn't mean to infer there was any increase in available bandwidth by the SB having only one transponder, just that it was interesting.

    Almost all the 49 spotbeams carried by the B-band birds D10-11-12 (D12 actually has only 44 enabled) have two or three transponders assigned to them.
     
  11. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    What you're seeing is the ability to re-use frequencies with spot beams, that can't be done with the CONUS beams.
     
  12. evan_s

    evan_s Hall Of Fame

    2,136
    0
    Mar 3, 2008
    I think that there are probably a number of factors here.

    1) As already mentioned the space remaining gauge is a pretty rough indicator at this point. You can improve this by deleting larger blocks at once. Delete 5 hours of shows at once and instead of getting 15% you might only get 12%. The more tests you do or the larger the blocks the more accurate a value you can get.

    2) The source- How many sub-channels and of what type(s) does that station have compared to others in the area? An OTA channel with only a single hd channel should be a higher bit rate source than one with several SD or even other HD sub-channels. This difference could end up impacting the re-encoding process resulting in a higher bit rate mpeg4 stream.

    3) The reencoding step. From what I've seen posted here DirecTV (and really all providers) use dynamic shaping to get the maximum quality out of a block (tp or QAM channel or what ever) for the channels that are in that block. Since that channel apparently has no other channels on that spot beam it's quite possible that the result is a much higher average bit-rate as it never has to worry about sharing with other channels.

    I do agree that the best way to confirm that it is actually a higher bit-rate would be to use DirecTV2PC to stream it and use tools on computer end to monitor how much information was sent. This should give us pretty accurate readings on the size of recordings. Unfortunately, there really isn't anything you could do about this if that channel is actually at a higher bit rate altho it isn't entirely a bad thing as it should result in higher quality.
     
  13. HoTat2

    HoTat2 Hall Of Fame

    7,312
    185
    Nov 16, 2005
    Los...
    Sure I'm aware of that;

    My point was that very few of the spotbeams have only one transponder assigned to them as does A4BB which KTLA is using.
     
  14. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    I'll bet more will be added.
     
  15. HoTat2

    HoTat2 Hall Of Fame

    7,312
    185
    Nov 16, 2005
    Los...
    I don't think the transponder to spotbeam configurations can be altered, at least not without special FCC approval, as that technical information is on the satellite's formal LOA application's Schedule S officially filed with the FCC.

    However, the number additional of channels from this or a near adjacent local market sharing or multiplexed onto that transponder most likely will be increased in the future. I mean I can't believe that KTLA gets to maintain the luxury of having an entire 36 MHz wide SB transponder all to itself this way.
     
  16. inkahauts

    inkahauts Well-Known Member

    22,661
    1,103
    Nov 13, 2006
    LA has 11 channels right now that are digital of Directv. And they are arranged very strange, over what looks like 5 tpns. I think they are just prepared to add all the local digital stations eventually, because they will need to in the next couple years I believe, and we have more than most markets do, and I am not even counting sub channels. I can easily see them showing 25 LIL digital stations in LA via sat.
     
  17. TomCat

    TomCat Broadcast Engineer

    4,153
    100
    Aug 31, 2002
    IIRC, there was some reason that the national network channels, which I think come primarily from LA and NY, were going to be remaining as MPEG-2 while virtually every other spot-beam market would be MPEG-4. I don't remember the logic behind this strategy; it would seem that they could distribute the nationals as MPEG-4 without issue. It may have been the large universe of MDU customers that were getting the nationals that played into it, as changing them over to the MPEG-4 delivery infrastructure was problematic.

    KTLA, while not a national channel, is backhauled along with the other network channels as a local channel for LA, so I guess they treat it as MPEG-2 as well.

    So KTLA, even if off the sat, may be anywhere from 12-18 mbps, although most stations are typically 13 to 16. But that is still about twice the bit rate of an MPEG-4 channel on DTV, roughly. Alphas is from Syfy, and cable channels are typically in the 7-8 mbps range as MPEG-4.

    That MPEG-4 takes less HDD space for virtually the same quality is why I only use sat locals (unless it rains really hard) even though I have a perfectly good OTA system connected.
     
  18. inkahauts

    inkahauts Well-Known Member

    22,661
    1,103
    Nov 13, 2006
    All la locals, including the distant feeds are now mpeg4. Ktla has been mpeg4 always, was never a mpeg2 local as I recall. Only the big 4 where. One of the reason I got a hr when it first came out.

    They actually have had the local la channels in mpeg4 since they launched mpeg4 locals off the space way sats. It was only the dns feeds that where kept in mpeg2, and that was for the reason you mentioned, all the tivos and mdu installs that had to be converted first before they switched the dns feeds to mpeg4. I think that was finished about two years ago now.
     
  19. TomCat

    TomCat Broadcast Engineer

    4,153
    100
    Aug 31, 2002
    Well, counting the subs is irrelevant, as they are muxed into the primary program stream, and there is only one program stream per broadcast source. Plus, DTV is not required to carry the subs, although they may opt to. Those they don't carry are probably muxed out before backhaul.

    The September Satellite Transponder Guide lists 26 SD channels and 9 HD channels for LA at 101 (DirecTV 4s, 8, and 9s). 5 spotbeam xponders for 9 or even 11 HD channels is not that strange, as a typical xponder only has enough bandwidth for about 2 MPEG-2 HD channels at typical ATSC bit rates, and I assume they are still MPEG-2 due to them being available as nationals as well. Phoenix only has 2 spot beam xponders for 6 HD channels, but those are MPEG-4, and with statmux, 3 or even 4 MPEG-4 HD channels will fit into a xponder.
     
  20. TomCat

    TomCat Broadcast Engineer

    4,153
    100
    Aug 31, 2002
    That may be your understanding, but I am skeptical that it is reality (no offense meant, please and thank you). If there is a perfectly good MPEG-2 infrastructure in place for DNS, why not use that same infrastructure for LA and NY locals rather than go to the expense of creating a second one at a cost of millions? The fact that they have so many xponders for so few HD channels seems to indicate that they are still MPEG-2. There may be a practical point to migrate over to MPEG-4, but what would drive that, and at what cost and for how much trouble? If it were my company, I would be tempted to just keep the MPEG-2 infrastructure going, as moving the rest of DTV to MPEG-4 has solved the bandwidth issue, which is what drove MPEG-4 in the first place.

    I am not declaring you to be incorrect, I just think we need a verifiable source for that information as to me it seems unlikely. But then I don't know the whole story either.
     

Share This Page