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

Tivo HD Pixelation Troubleshooting

Discussion in 'TiVo Series3 HDTV DVRs' started by Chimpware, Aug 3, 2007.

  1. Aug 5, 2007 #21 of 1597
    GoldenTiger

    GoldenTiger Tivo Pony =)!

    90
    0
    Apr 10, 2002

    Wow, what a disaster this launch is turning into... I will be returning two Tivo HD's within the return window if this isn't fixed by then. It is a HUGE issue to me, I have it happening twice a minute sometimes, and always 12+ times in any 10 minute window. It is infuriating while watching, and is not the kind of picture quality I want from this kind of expensive device. Clock's ticking, TiVo... I'm a huge TiVo fan, but I'll be damned if I'm going to pay $300 for the 3-year subscription and $260-300 each of my 2 units for a far inferior unit quality-wise to a rentable cable box other than the software menus.
     
  2. Aug 5, 2007 #22 of 1597
    jfh3

    jfh3 New Member

    4,469
    0
    Apr 15, 2004
    Denver area
    Yes, I've seen this on a THD w/o CCs on both the a and b s/w revisions.
     
  3. Aug 5, 2007 #23 of 1597
    jfh3

    jfh3 New Member

    4,469
    0
    Apr 15, 2004
    Denver area
    Disaster? I think that's a little dramatic.

    If everyone who has this problem calls in and gets a case number, we'll all be better off.

    I have two new THD boxes - I have no doubt that Tivo will fix this problem, so much so that I'm using a LGC to put lifetime on one box and upgrading its' hard drive.

    From the reports I've read, this seems like software, not hardware, so I see no need to panic and jump ship.
     
  4. Aug 5, 2007 #24 of 1597
    Ishma

    Ishma New Member

    38
    0
    Sep 8, 2006
    I agree. Give TiVo a chance to truly fix the issue. I would imagine there are a lot of people working on it now.

    I might recommend holding off a few weeks but panicking now?
     
  5. Aug 5, 2007 #25 of 1597
    Phantom Gremlin

    Phantom Gremlin New Member

    1,555
    0
    Jun 20, 2002
    Tualatin,...
    Unless you have read much more detailed information than has been posted here, your statement is unsupported.

    Please link to the "reports" that make clear this is software.

    IMO it could very well be hardware glitches that are causing this. And yes, at various points in my career I've been paid to write software and to design ASICs and to design systems.
     
  6. Aug 5, 2007 #26 of 1597
    ChrisPA

    ChrisPA New Member

    8
    0
    Aug 2, 2007
    Bucks...
    Yes, I had the pixelation problem both before and after the software upgrade using direct cable connection (no cable cards).

    Since then, cablecards have been installed, with the same pixelation problem.

    Regarding whether this is a "disaster" or not, I do think its extremely important they address this in some way soon. If things haven't been fixed - or there hasn't been an announcement from Tivo that they are sure things will be fixed - it will be a tough call as to whether or not to return the box when 30 days is up. Its a distracting problem that none of the rental units have...
     
  7. Aug 5, 2007 #27 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    Then you have something else wrong. The menus are created internally to the TiVo and are not compressed and I do not believe they are encrypted. They do not pass through the Cable Cards or the Tuner, at all. Which TiVo output are you using?

    You don't know that. He may not be seeing any whatsoever. Typically speaking, I see pixelization at roughly the same rate on my Series III units as the Scientific Atlanta 8300 HD DVR I rented from the CATV company. It's very subjective, and very intermittent, however. I have about 150 HD movies on one of my Series III units, and some of the movies haven't a single pixelization in the entire movie. Zathura only had two that I recall, and they were very minor. Guess Who's Coming to Dinner and Harry Potter and the Goblet of Fire have none at all. Top Gun, on the other hand has so many and such bad ones that it is virtually unwatchable. I'm waiting for it to come on again so I can replace the copy I have. On the other hand, when I had the 8300, it was no better. One episode of NCIS was so broken up I couldn't see what was going on, and it finally locked up the SA 8300 DVR.

    Qualitatively I would say all three devices tend to have a drop-out about every 30 - 45 minutes on average. Sometimes it's much worse, and sometimes it's considerably better.

    If it goes away upon replay, then the signal was fine all the way through the Cable Card. A replay comes from the hard drive, and the hard drive lies completely beyond the CC circuitry. If it is good on the hard drive, then your pixelization is neither from the cable company, the CableCard, nor the tuner. Either the TV or the TiVo has an internal problem, or you are losing bytes in the transmisssion cable. Are you using HDMI or component video? You might try re-seating or replacing the cable between the TiVo and the TV. I've never noticed any streams fail between the DVR and the TV, but then I haven't tried rewind every time I get a pixelization, either.
     
  8. Aug 5, 2007 #28 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    Well, I wouldn't agree with that. One cannot expect broadcast video to ever be 100% perfect, but at some level any persistent issue is going to become unnaceptable in whole or in part. Certainly if all my recordings were as bad as the recording of Top Gun which I mentioned in the previous message, I would be very upset. One or two very brief pixelization events an hour, however, is not enough to consider more than a minor annoyance.

    Define "poor quality". Many of the SD channels on my cable feed are downright horrible. They are snowy, grainy, and frequently have venetian blind or herringbone beats. Indeed, even though many of the SD channels have really great movies on them, since I have moved to HD I usually can't bear to watch them.

    Unfortunately, that is probably a lie. First of all, until the signal leaves the CableCard, there isn't really anything the TiVo can do to it - certainly not in software. The encrypted digital stream is sent without any modification directly from the receiver (which is a shielded monoloithic outboard device - not properly speaking part of the TiVo) directly to the CableCard where it is de-encrypted and then re-encrypted to be sent to the DVR. All this is rigidly controlled by Cable Labs specifications, and a Cable Labs certification warrantees this system meets specifications which will allow the CATV provider to deliver an acceptable signal to the device. If there is a systemic problem (other than a bad unit, of course) which prevents the source video from being reproduced in encrypted form at the receiver interface, and the receiver is certified by Cable Labs, then it is the CATV company's responsibility to fix the issue, not the device manufacturer (in this case TiVo). The CATV company must prove the device in question is not meeting Cable Labs specs. Otherwise, it's their problem, period. That is the law, per FCC regulations and CableLabs specifications, who also certifies the equipment used by the CATV company.

    If the problem is beyond the point where the CableCard encrypts the data, or if the device is not meeting CableLabs specs, then it is the responsibility of the device manufacturer to fix things. If this is the case, then it is also possible a software fix could affect the situation. With the exception of a post just previous to this one, I have seen no firm evidence either the TiVoHD nor the Series II are the source of the problem.
     
  9. Aug 5, 2007 #29 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    Well, first of all, there are several places along the way which could be the cause of pixelization. The CATV plant is the most likely, but at least one poster detailed issues which are clearly not caused by the CATV plant. There are only really two places in the stream where the TiVo might cause issues with it at the software level, and it's unlikely even at those junctures. What is highly likely is there are two or three (or more) different problems being encountered, here. It is also very likely all or at least most are hardware related - the slot 2 issues perhaps being an exception.

    I have a real suggestion. The source codes for the 8.1 software have been posted. I suggest someone point to me (and TiVo) where they think this code is faulty.
     
  10. Aug 5, 2007 #30 of 1597
    Chimpware

    Chimpware New Member

    334
    0
    Jan 30, 2002
    OK, top be honest I don't care where the problem lies, and it should not be mine to solve. The fact of the matter is that I as well as many others who have posted will be returning our Tivo HD units (2 in my case) if this issue is not remedied within the next 2 or 3 weeks.

    Posting that someone should sort through their code to find the issues is assinine as well as many of your other ridiculous comments such as pointing out that analog SD video can be grainy and therefore it makes a pixelated HD pictures from a Tivo HD acceptable.

    I have a better suggestion for you; if you have something constructive in terms of troubleshooting you can add that can help remedy the situation I as well as many others are having please let us know.
     
  11. Aug 5, 2007 #31 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    Well, those are all easy things to do, so I can't say it was foolish to try them, but the chances any of them would affect any given problem in general is very low. The highest probablity of resolving a problem lay in replacing the jumper, but I would guess that has less than a 1% chance of causing a problem.

    For a more effective run at it, call the CATV company and demand they come out with a spectrum analyzer, a distortion analyzer, and a signal flatness analyzer (sometimes called a sweep system). Some CATV systems use a single device which can do all this. Have them verify the signal levels at the back of the receiver across the entire CATVspectrum and look for ingress carriers with the Spectrum analyzer. No video carrier should be more than 10dB higher than any other video carrier across the entire spectrum. No video carrier should be more than 2dB higher than any adjacent carrier. The spectrum should be as free as possible of any ingress carriers, although in some areas ingress is unavoidable. The digital (high) end of the spectrum should be completely free of ingress carriers. The carrier / 3rd order distortion ratio should be no lower than 30dB on any channel. The peak video signal should be no more than 12 dBmV, and the lowest video signal should be no less than 0dBmV. Digital video will probably be run 10dB lower than the analog carriers, but after subtracting the differential, the same specs should apply. Finally, the spectrum flatness should be no more than 4 dB peak-to-valley after compensating for amplifier tilt and cable loss. If all these are good, then the CATV system is probably delivering a good signal to your TiVo.
     
  12. Aug 5, 2007 #32 of 1597
    Chimpware

    Chimpware New Member

    334
    0
    Jan 30, 2002
    Thanks, that was actually helpful, I will give that a try this week. Only concern is that I undertand only about 50% (or less really) of what you wrote so I guess I ahve some studying to do before teh cable guy comes.
     
  13. Aug 5, 2007 #33 of 1597
    Saturn

    Saturn Lord of the Rings

    12,249
    4
    Apr 10, 2001
    TiVo only posts the kernel and tools - essentially only the GPL stuff. The actual TiVo application you see is propriatary, as are the drivers for the hardware. You won't see the source for any of these unless you work for TiVo. :D
     
  14. Aug 5, 2007 #34 of 1597
    btwyx

    btwyx Substantive Member

    11,331
    0
    Jan 16, 2003
    Mountain...
    Even if it is hardware, the hardware guys just get us software guys to work around the problem anyway. The line between hardware and software problem can be a little blurry. The worst case would be a hardware problem with no software workaround.
     
  15. Aug 5, 2007 #35 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    It also isn't TiVo's to solve if it's being caused by substandard CATV signals. Indeed, by law, it is not TiVo's to fix unless they are failing to meet CableLabs certification specifications, nor would it be possible for them to fix it unless that is the case. (Assuming the problem is in reception. One poster here was clearly having a problem in transmission, which is an entirely different matter.) If it were, it should have shown up and been flagged when the boxes were submitted for certification, unless they've had a bad batch of boxes produced in their manufacturing run.

    You're free to do whatever you want, for whatever reason you want, but threatening TiVo they have to fix something which they have no power to fix or you'll return the unit doesn't help.

    No, it isn't. You are claiming there is something TiVo should be able to fix in their code. I'm asking you where you think that code is broken. Otherwise, quit asking for software fixes.

    That's not what I said. I said the local SD programming is all but unbearable to watch. On a scale of 1 to 10 where 10 is the best SD picture quality, some of the channels get a 3, and many only get a 5 or a 6. American Movie Classics, USA Network, Animal Planet, Fox, Sci-Fi, FX, CNN, WGN, and a number of others all look like hammered dog pookie.

    The HD programming I am watching here even with the pixelization is fabulous - with an occasional glitch. If the pixelization on the HD material were happening ten times more often, then the HD content might get close to being unwatchable. It is just that on the Top Gun recording I mentioned, but there the video was pixelating every minute or so, and the picture would be jumbled and the audio gone for several seconds at a time. It would get a 2 at best, but most of the programming on the units gets a 14 even with the few drop-outs. Some get a 15, with no or almost no drop-outs. The best DVD gets a 10.

    You asked, "...what is the point of having an HD device that has poor picture quality, when you could have an SD device with no pixelation, or a cable supplied box that works correctly?"
    To which my response is, "Why would I drop a service whose worst typical offering is a 13 so I can go back to a service whose best possible offering is a 10, but whose typical offering is rarely better than 8, and often worse than 5?"

    To the second half of the question I can only repsond that calling what the SA 8300 does "working correctly" is like saying the Spruce Goose could fly. The rate of pixelization I saw with it was not noticeably better than my Series III TiVos, and the user interface was completely unacceptable.
     
  16. Aug 6, 2007 #36 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    dB is a means of measuring power ratios. it is defined as 10 * Log (P1/P2), where Log is the logarithm base 10 of the power ratio. If one signal is 3.5dB higher than another, then it is twice as powerful. If it is 10dB higher, then it is ten times as powerful. A negative number indicates the upper number is smaller than the lower.

    dBmV is a measure of the power of a signal relative to a 1mV signal across 75 ohms. A +60dBmV signal is 1 volt across 75 ohms.

    An ingress carrier is a signal getting into a transmission line (in this case a CATV coaxial cable) from a source outside the transmission system. There are lots of signals in the air, from radio and television stations to paging system, aircraft communications, HAM radio, electric motors - you name it. If the levels of any carrier getting into a CATV cable are very high, they will cause problems there. For example, if a cable has a crack, then off-air channels will likely cause venetian blind or herringbone beats on the channels at which they reside. An ingress carrier can easily cause intermittent or persistent pixelization. Some receivers are better than others at rejecting in-band ingress carriers.

    Second order distortion is not generally a problem in modern CATV plants because hybrid amplifiers very nearly completely eliminate 2nd order distortion, but composite 2nd order can sometimes be a problem for digital carriers, and if one side of a gain stage is bad, an amplifier can cause large amounts of 2nd order distortion. This is pretty easy to spot in analog signals, and can be readily measured by distortion analyzers. The effects of second order distortion increase by 1dB for every 1dB increase in signal levels.

    Third order distortion is caused by a non-linear response in the output gain stages of an amplifier. No amplifier is perfect, and third order distortion increases as 20 * Log (# of amplifiers) in a cascade. By using fiber nodes, most CATV companies have greatly reduced the number of amplifiers in cascade to the user's house, but there are still often 3 or 4 in cascade before reaching the house - where there may be another amplifier. The effects of third order distortion also increase by 2dB for every 1dB increase in signal level, and more if the signals are not properly balanced. A small change in levels accompanied by a small change in tilt can devastate certain sections of the spectrum. As with ingress carriers, some receivers are better at rejecting third order distortion carriers than others.

    Signal flatness is simple. The lower frequency signals are deliberately run at a lower level than the higher frequencies because the attenuation in the cable plant is lower at those frequencies. Typically at the amp, the high end video carriers may be 8 or 9 dB hotter than the lowest channels. After leaving an amplifier and moving down the CATV plant, the higher frequency signals drop in level faster than the lower frequencies. After some distance, the high end will have dropped to the point where it is the same level as the lower frequencies. The signal is still quite useable well beyond this point, however, so the plant will continue to run until the highest frequency carrier has dropped about 16dB more than the lowest carrier. At the back of your set, the high frequencies may be higher in level or the lower frequencies, may be, but the difference between the highest and lowest must not be greater than 10dB (don't forget to add the offset to digital carriers when comparing to analog carriers). Actually, come to think of it, that spec may be 9dB, not 10. It's been nearly 25 years since I've done any FCC compliance testing, so my memory is a bit foggy on the exact details. No carrier should be more than 2dB above or below its adjacent carriers. Again, distortions in the receiver caused by an overly tilted signal and common mode rejection of adjacent carriers will vary between receivers. One might do fine and another be sent mad.

    Finally, spectrum flatness is also a simple notion. In a perfect scenario, every carrier would be injected at exactly the same level into the plant at the cable headend, and those signals would reach the consumer again with every carrier being exactly the same level. Of course, as I have already mentioned, the levels are actually set up so they leave the amplifiers with a tilt and that tilt is partially or completely obliterated or even reversed by the CATV plant along the way to the subscriber. The tilt and cable equalization can be compensated for in the test equipment, however, and in a perfect world a perfectly flat signal - with each carrier being exactly equal in level - would pop out. This is not the case, however. cable connectors, directional couplers, and amplifiers all have a unique frequency response which will cause some areas of the spectrum to be attenuated just a bit more than others. The overall response is called spectrum flatness, or in the vernacular peak-to-valley. After compensating for signal tilt and cable equalization, the maximum difference between the highest area of the spectrum (usually very distinct from the carrier with the highest level) and the lowest area of the spectrum is 4.0dB.
     
  17. Aug 6, 2007 #37 of 1597
    Chimpware

    Chimpware New Member

    334
    0
    Jan 30, 2002
    Just some quick simple points:

    1. I am not asking for a software fix, that is what the Tech. at Tivo told me they were working on. I am asking Tivo to remedy the situation. Telling everyone that legally Tivo is not responsible is useless. No one is going to take the local cable company to court to get their Tivo working properly.
    2. Many people including me are not happy with a picture that pixelates once every minute or so, and if Tivo cannot either remedy the situation, or provide adequate troubleshooting support to have the situation remedied a lot of Tivo HD unit will be returned, that is just a fact. I am in no way threatening Tivo, I am just stating a fact. At the end of the day the Tivo does not work properly for many people here. I have 2 SA 8300 HD boxes on the same lines as the Tivo HD units and never have had any pixelation issues. Do they work as well as a Tivo, No, but they work. Who the heck would want a Ferrari that stalled every mile, even though it was a great sports car?

    If you are happy with a pixelated picture then you have no problems, no reason to post on any of the thread realted to pixelation then I guess huh?
     
  18. Aug 6, 2007 #38 of 1597
    Chimpware

    Chimpware New Member

    334
    0
    Jan 30, 2002
    Thanks appreciate the assistance. I already have an advanced degree in Chemical Engineering, gues I now need a PhD in Electrical Engineering to get my Tivo working.
     
  19. Aug 6, 2007 #39 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    Your posts - especially the first one - strongly suggest you do not think it is a hardware issue. The only other option is software, and you are asking TiVo to fix the problem, so the only inferreence is you are asking TiVo to implement a software fix. QED.

    No, it isn't. The local CATV company is bound by law to make their CATV plant work with any CableCard certified device.

    I never suggested they should. A phone call or letter to t he FCC can do wonders. For that matter, a call or letter to the General Manager of the CATV company quoting the FCC regulation and promising to call or write the FCC can work magic. In the past the FCC has not hesitated to levy six and seven figure fines to CATV companies and even to shut them down cold for non-compliance with FCC regulations.

    Unless I missed it, I believe this is the first time you have quantified the level of interruption in this thread. It would have been helpful had you been more specific in previous posts. That said, I still find it unlikely the TiVos are the culprit, and even less likely it is a software issue. I cannot guarantee either, of course. I've worked with many, many devices which worked fine on the bench, but failed in the field. In most cases I was eventually able to pin the down the problem to something in the field. Sometimes it was something very subtle, and it was not at all unusual for another device - either similar or of the exact same type - not to show the problem. There are always exceptions, however.

    That's impossible if it is not a problem with the TiVos.

    What do you suggest? They can test returned units, but this will take considerably more than 2 weeks. Other than that, what do you think they should be able to do?

    Of course you are. You are threatening to return the TiVo if the problem is not resolved. It's not a legally actionable threat - you have every right to return the unit under their stated warranty for whatever reason you like. This does not change the fact you are attempting to legally extort action from TiVo by legally threatening to impact their revenue stream.

    To extend your analogy, my point is you may be screaming at Ferrari because it stalls every mile when the reason it stalls is the cheap gas sold by your local filling station. Your beat up old Chevrolet may take the gas in stride, but the Ferrari may be choking on it. Returning the Ferrari will get rid of the problem, but so will buying better gas. If the gas costs the same, which is the better solution?

    So no one who disagrees with your point of view should post?

    I am not happy with pixelization in the video. I wasn't happy with it the first time it started happening. (It was 1984, with the introduction of the Zenith Z-View converters in the local market, and I worked for the CATV company at that time.) I'm not happy with it now. A 90 minute movie with 3 or 4 drop-outs of not more than 2 seconds does not make the video less watchable than a grainy, snowy, beat ridden SD picture, however.
     
  20. Aug 6, 2007 #40 of 1597
    lrhorer

    lrhorer New Member

    6,922
    0
    Aug 31, 2003
    San...
    All of that information should be readily deliverable from memory by every CATV technician. Very few CATV techs have even bachelor's degrees, let alone graduate degrees. Aside from the actual specifications for flatness, etc and the fact that hybrid amps eliminate 2nd order distortion, there is nothing there I did not know when I graduated high school. (Well, that and the single unit test sets. Those did not exist when I graduated high school.)
     

Share This Page