My TivoHD Pixelation Story + 9.1 with Verizon FiOS in NoVA

Discussion in 'TiVo Series3 HDTV DVRs' started by edrock200, Oct 9, 2007.

  1. Oct 9, 2007 #1 of 28
    edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...

    Advertisements

    I know there are a ton of threads on this, so sorry for yet another, but I wanted to start a thread to chart my progress.

    My setup:
    Verizon FiOS in Northern VA
    2 single stream cable cards in each Tivo HD
    4 Tivo HD's
    1 Motorola FiOS DVR

    So I began to notice the Tivo pixelation issues very early on on practically all cable-HD channels. This issue did not seem to affect OTA channels that were in HD. The pixelation issue occurred on both encrypted and unencrypted channels delivered via FiOS.

    So, first I began basic troubleshooting. All 4 Tivo's were experiencing the pixelation issues. One of the Tivo's was in the same room as the motorola DVR. The motorola DVR never pixelates. So, I swapped coax cables between the Tivo and the moto DVR. Moto DVR does still does not pixelate, Tivo still pixelates.

    So then I check signal levels on the Tivo and see it's bouncing around, wildly, from 0 to 80 to 90 to 100. Generally stays between 90 and 100 but drops to the 60's. I believe it was durring these drops that the pixellation occurred. The bizarre part is a forced reboot seemed to fix the problem for at least several hours.

    I then decided to give Tivo a shot at addressing the situation. So I contacted Tivo today and explained to the first CSR the issue. She immediately knew of the problem, which was a good sign. So claimed that 9.1 would fix the problem and told me how to get on the priority list and gave me a service credit. So far so good. Signed up for the priority list then performed a forced update this evening. Tivo HD #1 now on 9.1

    Tune to a FiOS HD channel and immediately see pixelation. It was too late to call Tivo so I read some threads on the subject and saw someone suggest checking RS Corrected and RS Uncorrected under the diag screen.

    I go in and sure enough both numbers are slowly going up, and signal is all over the place. Heres the odd part. After about an hour of the unit being updated the signal is now steady between 90 and 100 and rs corrected and uncorrected remain at zero, even after being tuned to the same channel for about an hour.

    Does any of this make sense to anyone? Is there anything that would have required time after the update to correct "signal-locking?"

    Anyway, I'll keep monitoring the situation and if it comes back, I'll contact Tivo again and update this thread.

    I don't get why the signal would swing 10 points though, is that normal?

    I do encourage anyone who is having this issue to open a ticket with Tivo. The more tickets they have to track the more information they can collect on the issue and it can possibly get the problem higher on the priority list.
    -Ed
     
  2. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    Just a follow up, my TivoHD *appears* to be working properly now with the 9.1 update, but its still too early to tell. Haven't seen pixelation on recordings anymore. Personally, I think in my personal situation, the signal was "too hot" and the software may be able to attenuate the signal somewhat.
     
  3. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    Just an update, as many have already figured out, I spoke to soon. Watched an HD show last night and indeed the pixelation is still there. it's much better, and IMO the shows are now "watchable" but its still there.
     
  4. AbMagFab

    AbMagFab What happened, TiVo?

    3,638
    0
    Feb 5, 2001
    I've found that with my FIOS (in MD), it's the SNR that's far more imporant. I've attenuated my signal down by -11db to get the SNR to ~31/32. My pixelation issues are basically gone.

    My signal strength is in the 60's, but apparently that's not a problem.

    I've had it stable now for a couple days.

    Some of my symptoms, for what it's worth (RS Corrected/Uncorrected is the quantitative measure of the problems):

    - Most of my issues were on a fixed set of channels, mostly in the same frequency band, according to the diag screen. Some specific channels were 169 (BBCA), 177 (FOX Reality), and 851 (HBOHD) (the latter is in a different frequency).

    - Strangely, the problems seemed to correlate to time of day. So from about 10am to about 2pm, there were almost no problems. Then around 3pm to around 11pm, problems galore. Then it seemed to calm down again. It was consistent enough that I noted this, but it was weird, and I have no explanation (except that it strangely correlates to kids home from school).

    - When the problems occured, they would occur on both tuners. Sometimes one tuner would get the problem quicker than the other, but it would show up on both.

    - Once the problem started (i.e. RS Uncorrected skyrockets), it generally wouldn't stop until I changed to a different frequency and then back.

    - The signal strength would be all over the place

    - For other frequencies, the signal strength would be stable, and there would be no macroblocking, and the RS numbers would be 0 or near 0.

    Once I attenuated the signal way way down, the signal is now worse, but it's relatively stable, fluctuating only about +/- 5, and the SNR is stable at +/- 1.

    In any case, it's worth a shot. PM me if you want more detail on how to attenuate the signal. Again, since we're in the same area and seem to have the same problem, it's worth a shot!

    (You can attenuate with in-line attenuators, about $2 each, or if you have some cheap splitters handy, the 4x ones are -7.4db, and the 2x ones are -3db. Try one or more until you get your SNR down to 31/32, and see what happens. Don't worry about signal strenth unless it gets too low.)

    This worked for me, I'm not sure it's really a solution or not, but it's worth giving it a shot since we're in the same area with similar problems.
     
  5. SCSIRAID

    SCSIRAID Active Member

    2,339
    2
    Feb 2, 2003
    Vero Beach, FL

    Advertisements

    31/32 db SNR is still too low... should be a couple db more for reliable operation.
     
  6. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    Thanks for the info guys. Really apreciate it. AbMagFab I will try what you mentioned. I remember seeing a variable 20db attenuator at radioshack. It might be worthing picking one up just to figure out what the sweetspot is.

    With the SNR, what is better, higher #'s or lower #'s?
     
  7. bizzy

    bizzy New Member

    707
    0
    Jan 20, 2004
    San...
    Make sure the attenuator you pick up is rated as flat through at least 1GHz!
     
  8. gt7610c

    gt7610c New Member

    32
    0
    Oct 5, 2002
    NoVA
    FWIW and I'm probably jinxing myself, but I have 8.1.7c with signal strength ~68 on one CC and ~81 on the other with SNR 32-34 and I haven't seen any pixelization at all (knock on wood). Not looking forward to 9.1 if it is going to cause pixelization :(

    I should add I'm in NoVA with FiOS with a THD and a VZ DVR (no pixelization there either, no shock).
     
  9. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    hmmm...so I just daisy chained two 2-way splitters (at 3db each this means -6db right?) and got the signal to the same exact level as abmagfab, signal around 60, SNR ~31/32 (down from 34/35.) Again, not sure if higher or lower is better here. Will try it out for the weekend.

    Should also add with only one splitter signal dropped to ~85/90 and SNR stayed at 34/35.
     
  10. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    I'm gonna turn in for the evening but just an FYI. It's been about 25 mins and so far no RS uncorrected, lots of RS corrected, no signal "blips" though I've been flipping between the two tuners. I'll update this thread over the weekend. Thanks again for the info.
     
  11. SCSIRAID

    SCSIRAID Active Member

    2,339
    2
    Feb 2, 2003
    Vero Beach, FL
    Higher SNR is better. 34-35 is excellent 36 is perfect.
     
  12. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    Well I can't explain it then. I will go ahead and leave the splitters on and see what happens since as of now I still have not experienced the blocking issues. If it is an issue of the signal being too hot I may just try one splitter to get the signal to 90ish and the SNR to 35. Anyway, I'll report back my findings.
     
  13. AbMagFab

    AbMagFab What happened, TiVo?

    3,638
    0
    Feb 5, 2001
    There is no such thing as a better SNR. There is theory, then there is what actually works.

    One of the more annoying things on this board (and in life) is when people get all "academic", and start spouting what is theoretically good or bad.

    In this case, the THD + FIOS appears to work best with a SNR at around 31/32. It's been working for me for over 3 days now, when it wouldn't be stable for more than about 30 minutes before.

    If it works for edrock200 also, then we have two "reality" data points. I tend to ignore the theory in the face of reality.

    edrock - I also get some RS Corrected, and even a few RS Uncorrected, but so far I haven't *seen* any macroblocking. So the RS Uncorrected is either somewhere I'm not looking, perhaps during commercials, or it's at the beginning/end that I'm not watching. In any case, my recordings and LiveTV have been rock solid at 31/32.

    For what it's worth, when I bumped it up to 33/34, it was worse - the macroblocking was visible. Not as bad as when it was a straight signal, but worse than 31/32.

    For FIOS at least, and for me, 31/32 SNR on the THD seems to be nearly perfect. My S3's (for what it's worth) don't seem to have any issue with a 35-38 SNR, but I will probably attentuate them down a little after a week of stability on the THD, just to see if it makes anything worse.

    (Can you post a list to the radio shack variable attenuator? I'm using the $2 in-line ones, mixing and matching -3db, -6db, -10db ones.)
     
  14. SCSIRAID

    SCSIRAID Active Member

    2,339
    2
    Feb 2, 2003
    Vero Beach, FL
    http://broadband.motorola.com/ips/pdf/QAM.pdf

    See page 4. SNR directly effects BER. Different coding schemes have different requirements. Did you consider that the effect you are seeing may not be due to SNR?

    QAM is rather sensitive to BER. For example... ATSC is designed around an 18db minimum SNR.
     
  15. AbMagFab

    AbMagFab What happened, TiVo?

    3,638
    0
    Feb 5, 2001
    Again, there is a book, and there is reality. The book will have to bow-down in the face of reality.

    This is the ongoing battle between the book-smart folks and the actually-smart folks. Guess which one I side with?
     
  16. SCSIRAID

    SCSIRAID Active Member

    2,339
    2
    Feb 2, 2003
    Vero Beach, FL
    You arent considering that the effect you may be fighting ISNT directly related to SNR. Success is wonderful and what always counts is results. However, if you really believe lower SNR is better then youre mistaken. BER is a multidimensional issue and more than likely whatever you are doing to ease your situation is trading off higher BER due to SNR for lower BER due to the actual problem and in your case, the BER due to the actual problem appears to be much more significant so the trade off is successful. In total though, your BER is going to be higher than it should be if everything was 'right'. Smart people designed QAM256 and characterized its channel needs.

    Calling me not-smart doesnt help your case.
     
  17. AbMagFab

    AbMagFab What happened, TiVo?

    3,638
    0
    Feb 5, 2001
    Don't know what to tell you, but if:

    1) SNR > 32 = severe macroblocking
    2) SNR <=32 = no macroblocking (and no other negative side effects)

    Then, lower SNR is better, for me. Put all the words around it you want, but it boils down to lower SNR is better in this case.

    And I'd be willing to bet a diet coke that it will be better for a good chunk of other FIOS/TivoHD people. Going on and on that, theoretically lower SNR might not be better based on the theoretical design of QAM is sort of meaningless, isn't it? Unless of course someone is looking for an education in the theory of QAM.

    (and I wasn't calling you not smart, I was just characterizing various approaches)
     
  18. SCSIRAID

    SCSIRAID Active Member

    2,339
    2
    Feb 2, 2003
    Vero Beach, FL
    I agree that in this case the lower SNR is better... but that doesnt make it right. Success is the final arbitor.

    I'd lose a diet coke with that bet as I believe you are right. And no... its not meaningless. It speaks volumes. It says that the actual issue likely isnt SNR related. There is something else much more significant going on that is swamping the effect of lower SNR. The few RS Uncorrected's you are getting are probably due to the 31db SNR... However... that appears to be small potatoes compared to massive pixelation when the SNR is up where its supposed to be.
     
  19. AbMagFab

    AbMagFab What happened, TiVo?

    3,638
    0
    Feb 5, 2001
    Right, so unless there is some other attribute I can change by itself, the lower SNR is better.

    It's like rebooting a Tivo to fix something (there are a couple problems that are only fixed this way). Sure it fixes it, but there was clearly something else that was the actual problem. But without the ability to tweak that one little thing, rebooting is good, and it works.

    Lower SNR is good, it works, and there are no apparent other (bad) side effects.
     
  20. edrock200

    edrock200 New Member

    2,569
    1
    Feb 7, 2002
    Northern...
    AMF: I owe you a beer. 24 hours later and no drop outs on any recordings. Not counting my chickens yet but this is 100x better than before.

    SCSIRAID: FYI when my signal was 100 and SNR was 35/36 my RS uncorrected and uncorrected were extremely high. Now uncorrected is under 100 and corrected is only a few thousand. Keep in mind its been on the same channel for hours, so these numbers aren't only after a few minutes.

    AMF: In response to your question about the variable attenuator, it looks like they don't sell it anymore, and for good reason as its only 0-500mhz so it wouldn't work for us anyway:
    http://support.radioshack.com/support_tutorials/video/faq-1500578.htm

    Looks like I'll just ebay a 10 pack of 6db attenuators. Before I order them though can you confirm that two daisy chained two way splitters = roughly 6db loss? Or is there another calculation when you daisy chain them?

    Thanks again for your help. I was really losing hope after 9.1 didn't fix the problem.
     

Share This Page

spam firewall

Advertisements