Ceton support infinite loop - macroblocking

[jziggity]

I am appealing to the audience at large here, in hopes that Motz or Erkotz, or anyone else can offer some additional suggestions. Awhile back I upgraded my InfiniTV4 PCI card to the latest beta firmware to address the single line macroblocking issues. I am running Win7 on a brand new HTPC, plenty of RAM, SSD, etc. etc. The update resolved that problem, but caused another. Now I am seeing full to partial screen macroblocking errors, on a regular basis. It is no longer confined to just a couple of lines worth. No problem, I'll just submit a support ticket to Ceton and they'll get this resolved, or so I thought...

On the same trouble ticket (#MGF-725-43604) they have had me repeat several processes multiple times now. It appears from my end that the ticket gets passed around to various different people (I have no problem with that) and each new person doesn't read what was already attempted previously (now that's annoying). I have been asked to describe my wiring topography (coax and home network) three times now. I have been asked twice to submit a diagnostics capture file after manual tuning each of the 4 tuners to the problem channel. After the first time, they told me the signal levels were fine, but there were still excessive demodulation errors. Now they are telling me it's probably my 8 way splitter/amplifier that has gone bad. I highly doubt that, given that they already told me the signal levels were fine!! Arrgghhh!!

So, with the trouble ticket support system rapidly falling apart, what are my other options? I have run the CetonSignals program several times, and the levels look good. For S/N, I'm above 35 except for a couple of frequencies that show 34.8 and 34.9. The lowest I have seen it go is 34.6. In the signal column, the highest number I have is 9.0 at frequency 687000 and the lowest I have is 1.6 at frequencies 111000 and 117000. I am certainly not an expert at these numbers, but doesn't that seem like an excessive swing in signal levels between my highest and lowest points?

I just want the InfiniTV card to work as advertised. Ceton support already threw in the towel when it came to my issues with waking up from sleep mode. They just told me to leave my machine on 24/7. Yes I know that will technically resolve that issue, but that's not what I was looking for there either. It seems to me that when it comes to this macroblocking issue I'm stuck in an infinite support loop. Any advice at this point would be appreciated. Thanks!

[newfiend]

There is another thread on here started by Barnabas1969 for basically the same issue. After doing multiple tests I believe he determined it wan't the Ceton card causing the issue but his video card. I would PM him and ask him what he finally did that fixed the issue for him. I know he ended up switching to a SilconDust HDHRP tuner (I believe) but still had macroblocking issues. After changing out his video card I think that finally solved the issue for him. I would PM him just to be sure though. I have two Ceton PCIe cards in one machine. I rarely get macroblocking issues. I used to get them more frequently but once i had the Comcast guy out and got my signal levels as high as possible the issues all but vanished. I haven't had a macroblocking issue for some time now. I am running the latest Beta Diagnostic Utility/Drivers and Latest BETA Firmware as well with no issues.
HTH,
newfiend~

[crawfish]

This "macroblocking" can be due to momentary glitches in the signal that do not show up in the signal meter software that comes with the tuners. One potential cause for this appears to be LTE. Maybe a year ago, it was happening with my HD Homerun Prime, and it also happened on the cable company DVR in the other room, and the problem had the same characteristics as described by the people with Cetons. The cable company came out and of course found nothing wrong at my house. The guy escalated the problem though, and a couple of days later, I got a phone call from a tech in the field saying that the problem should be fixed. It was, and it has been fine ever since.

The conclusion I drew from my experience is that unless you can verify the problem doesn't happen using an independent tuner, preferably a box from the cable company, it's just guessing to blame it on the PC.

[WhatHappend]

@jziggity,

Have you ran the latest version of the cetonsignals script? http://experts.windows.com/frms/windows ... spx#554438

I see that you have signals that are less than -15dBm. That is very low. Do they contain channels you watch? (largest 3 channels in your results)
I got that from your post here: http://www.thegreenbutton.tv/forums/vie ... 092#p39092

The updated script will correctly flag those signal levels as red along with their crappy SNR values.

Keep in mind the script is only looking at upper/lower&mid frequencies. You should manually check channels via the Ceton web page for channels you are having issues with that don't show up in the list of channel in the results.

Can you post your script results here along with the some manual values for Signal/SNR on the channels you're having macro blocking on?

UPDATE: I uploaded a new script AllCetionSignals.HTA that will capture your results for all of your channels.

[jziggity]

Okay, thank you to everyone who has weighed in. The latest response I got from my Ceton support ticket was:

1. Update your video card drivers to the most recent version. (Done, has been for awhile now.)
2. Change the recording folder to a different hard drive. (Can't, only have 1 SSD and 1 traditional hard drive installed, and I'm not setting recordings to use the SDD.)
3. Temporarily uninstall your antivirus program to see if the problem persists. (Haven't tried this, but I'm already using MS Security Essentials, which I thought was supposed to be the one option that played the nicest with WMC.)

WhatHappend, those bottom of the barrel numbers from my other post are not active channels for me. When I run the CetonSignals script, those signals actually show up as black boxes on my screen that I can't even see. I did confirm through my guide on WMC that those channels don't even exist for me. I tuned to channel 1004 (local NBC channel, one of the problem ones) several times manually through the InfiniTV card web interface, and I am consistently receiving signal values of 6.3 to 6.4 and the S/N value remains steady at 36.6. See below for a screen capture of the results from the latest CetonSignals script. Where is the AllCetonSignals.HTA?

My other thought at this point is that I could try using a decent coax splitter to make a connection split directly off of the incoming cable line, instead of going through the splitter/amplifier, to eliminate that as a potential problem (but I don't see why it would be). I was also thinking of trying to split the cable at the wall jack by the PC, and have one split go to the tuning adapter and the other split go to the Ceton card. Right now I'm using the internal coax pass-through of the TA, so I suppose I could eliminate another possible culprit that way. I have to go out to get a decent splitter though - used to have one from TW but of course now I can't find it. I will try both of those alternatives and post back here with results.

At one point in the trouble ticket process, Ceton suggested contacting TWC to have someone come out to examine signal levels and see what they could do, but now they have backed away from that suggestion. Back when I had TWC's equipment, I would rarely see just momentary pixelation issues, like only a couple of random squares here and there on the screen. So, I'm guessing there ARE occasional glitches in the signal TWC is providing, but it appears the Ceton card is magnifying that 10 times over. For me, the fact that the macroblocking errors changed from the single line to partial to full-screen errors after the Ceton firmware upgrade still points to the card as being the major cause. No other hardware or cabling changes have been made since that firmware upgrade.

I will touch base with Barnabas too and see what he thinks.

Crawfish, can you tell me what LTE is? Thanks!

[crawfish]

Crawfish, can you tell me what LTE is? Thanks!

It's a wireless technology used on an increasing basis in phones. I've seen it mentioned a bunch in the Silicondust forums as a cause of "macroblocking" and "pixelation", particularly for channels in the range 700-800 MHz. If you come up empty after exhausting all the potential PC causes, I'd recommend renting a cable company DVR for a month. If you can record the problem on it, call your cable company and demand they fix it. If the problem is not with your computer setup, you can replace everything, and it's not going to fix it, and I've seen that happen. Search my messages here over the last year, and you'll see what I'm talking about, and I'll leave it at that.

[jziggity]

I'd recommend renting a cable company DVR for a month. If you can record the problem on it, call your cable company and demand they fix it. If the problem is not with your computer setup, you can replace everything, and it's not going to fix it, and I've seen that happen. Search my messages here over the last year, and you'll see what I'm talking about, and I'll leave it at that.

I hear you, but I can't go down the road of picking up a DVR box from TWC again, not after the last disaster that ended up in me filing an FCC complaint against them. I am confident that a test such as this would show that everything is fine when using TWC equipment, as I have been using their stuff for several years without issue (relating to picture quality, that is).

I still have some splitter experimentation to try, so here's to crossing my fingers on that...

[crawfish]

I hear you, but I can't go down the road of picking up a DVR box from TWC again, not after the last disaster that ended up in me filing an FCC complaint against them. I am confident that a test such as this would show that everything is fine when using TWC equipment, as I have been using their stuff for several years without issue (relating to picture quality, that is).

I still have some splitter experimentation to try, so here's to crossing my fingers on that...

What, they blackball you or something? (Just kidding. )

Definitely replace splitters, break and retighten connections, etc. Do all the easy reasonable stuff first. If you have a TV with Clear QAM tuner, and it can still get your locals, you might try watching it for extended periods as a test. I wouldn't assume that just because the TWC box worked fine for several years, it would work fine now, and that would be something I'd keep in the back of my mind if all else fails.

[jziggity]

Well, they didn't want to play nice in the sandbox when I turned in my TWC equipment and went strictly with the cablecard. The whole ordeal is detailed nicely in the cable company section of the forums here, if you're looking for some light bed time reading. (http://www.thegreenbutton.tv/forums/vie ... =13&t=4008)

I don't have any splitters in the mix now, just an 8 way splitter/amplifier that all coax runs originate from. I was considering adding a splitter to the mix to move the run to the Ceton card either around the amplifier or around the TA, instead of through. We'll see what happens - it's an easy adjustment.

Oddly enough, the problem seems to ebb and flow at various times. Last night I recorded the 5PM news on the local NBC channel and watched it after work. I only saw 2 glitches in the 30 minute program, which I would consider a great improvement over previous observations. And the first one didn't even happen until 15 minutes into the recording. If this were the case all the time, I would be happy with that and call it a day.

I do have several TV's with clear QAM tuners, one of which I should be able to move around without much effort. If I watch the clear QAM TV next to the WMC Ceton card fed TV, and see a glitch on the WMC version but not the clear QAM, is it safe to say the problem lies with the card and/or computer...?

[crawfish]

Oddly enough, the problem seems to ebb and flow at various times. Last night I recorded the 5PM news on the local NBC channel and watched it after work. I only saw 2 glitches in the 30 minute program, which I would consider a great improvement over previous observations. And the first one didn't even happen until 15 minutes into the recording. If this were the case all the time, I would be happy with that and call it a day.

I think you can and should expect much, much better performance than that.

I do have several TV's with clear QAM tuners, one of which I should be able to move around without much effort. If I watch the clear QAM TV next to the WMC Ceton card fed TV, and see a glitch on the WMC version but not the clear QAM, is it safe to say the problem lies with the card and/or computer...?

It would definitely support that position. However, it doesn't mean the fault is not in the signal or broadcast. One example would be the 29/59 bug, which PCs have trouble with, but cable boxes do not. Another would be the glitch in the H264 channels in my area that puts Media Center into a strange state in which tuning most channels results in a black screen, solved either by stopping ehrecvr and restarting or by tuning a very small selection of local channels, such as my local NBC affiliate. It was not the HD Homerun Prime at fault, because using two PCs, once one Media Center got into a bad state, it stayed there until fixed as previously described. Making it use a different tuner didn't help, and a tuner used to put one Media Center into a bad state could be used in the other Media Center without a problem. SiliconDust support chalked this up to a bug in Media Center, possibly a bad reaction to something in the signal/broadcast that Microsoft never fixed, that cable boxes have no trouble with, sort of making it Yet Another 29/59 Bug. I didn't care about the H264 channels and sometime later stopped subscribing to the service level that included them, so I didn't pursue this any further, but I would have called the cable company before Microsoft, as Microsoft would want to charge me just to talk about the problem, and I have zero expectation they would try to fix it.

Hopefully, you'll see the glitch in Clear QAM using the TV tuner, proving the other thing. (Or given your conflict with TWC, maybe that wouldn't be a good thing. ) Anyway, good luck with it all.

[WhatHappend]

WhatHappend, those bottom of the barrel numbers from my other post are not active channels for me. When I run the CetonSignals script, those signals actually show up as black boxes on my screen that I can't even see. I did confirm through my guide on WMC that those channels don't even exist for me. I tuned to channel 1004 (local NBC channel, one of the problem ones) several times manually through the InfiniTV card web interface, and I am consistently receiving signal values of 6.3 to 6.4 and the S/N value remains steady at 36.6. See below for a screen capture of the results from the latest CetonSignals script.

The signal in your results look pretty good (except the last two). It is best to have no yellow in the SNR column. I have noticed the SNR values in the Yellow may macro block occasionally (10min or so).

Where is the AllCetonSignals.HTA?

http://experts.windows.com/frms/windows ... spx#555792


Things I would try would be running the main cable feed right to the tuner (no amp or splitter) and do a 1 hour test and see if your image quality improves. I would also run the AllCetonSignals script to get a reference idea of the best SNR quality you have to work with. FYI - an amplifier can't improve SNR over what is provided at its input, it only improves signal level to minimize further SNR loss in spitter/couplers/wall plates/cabling.

I would use a XBox360 as a test to view your recorded TV and Live TV feeds and see if that minimizes the blocking.

Do you use MOCA or have a device that may have it enabled? If so, you may need a LPF filter at the tuner input.

[jziggity]

Hi WH,

Thanks for the feedback. I picked up a splitter at RadioShack today, but before going that route I will try the direct cable feed option. That will involve using a female-female coupler to make the physical connection, but that would be it. It will also knock out my internet and phone, as the modem would no longer have an active coax feed running to it, but for an hour or so I think we'll survive.

I will definitely run the AllCetonSignals script too and see what that shows, thanks for the link on that. If I understand you correctly, there's not much I can do on my end with getting those yellow S/N values higher? If I do have to get TWC out to work on the S/N issue, and I show them the results of the AllCetonSignals script, do you think they will understand what the problem is and what needs to be done to correct it?

I do not have any MOCA devices. I have 2 Xbox's that handle all of the main TV watching at 2 different locations in the house (living room and basement). They are connected via cat6 cables to a gigabit switch, which then ties into an older wireless router. While watching football this past weekend, I was able to observe macroblocking on both the PC and an XBox fed TV, at the same point in time.

I will conduct the above testing within the next day or so and post back here with results. Thanks again!

[WhatHappend]

If I understand you correctly, there's not much I can do on my end with getting those yellow S/N values higher? If I do have to get TWC out to work on the S/N issue, and I show them the results of the AllCetonSignals script, do you think they will understand what the problem is and what needs to be done to correct it?

The SNR Yellows you get after attaching to the direct feed, those are the ones you won't be able to fix with an AMP. Your Yellows you captured so far are really fine (just barely less than 35dB with good signal strength Read here: http://cetoncorp.com/support/index.php? ... -explained).

Your Provider might charge you a service visit fee if your complain about 34.6dB SNR value.

[jziggity]

Okay, ran some more testing, and the results are not good. These are the lowest values I've seen yet. Below is the result of the AllCetonSignals with the feed running through my amplified 8-way splitter, and also the results with just a direct cable connection (using the female/female coupler) to the InfiniTV PCI card. The black number in the chart is supposed to be 31.3 (per the Ceton web interface for the card). I've never had a number that low before today. I'm not sure what changed in the last day or so. Naturally, channel 1004 (NBC) is one of the bigger problem ones, as is 1006. Not surprisingly, when I watched channel 1004 for a couple of minutes, there was quite a bit of pixilation and audio drop outs. So now what do I do? Is it time to bring out the cable guys?

Ah crap, it's getting worse!! I reconnected the cable run through the amp and the S/N for frequency 57000 is now down to 28.3, but seems to be bouncing all over the place. What the $%@# is going on?

[WhatHappend]

What kind of AMP do you have? Your amp is adding lots of noise. The SNR should stay the same (within 0.1dB) before and after the AMP. Your AMP seems to be adding excessive noise at the lower frequencies. Also, your 8 way amp is not keeping the level the same as the pre-input level was (do you have additional splits after/before the amp?)

Is channel 1004 Clear QAM? Have you viewed it on a clear QAM tuner and had the same issue (with direct connection to cable feed)?

I would call the cable company out for a SNR that low. Or if your clear QAM with direct connection shows the same issues.

Note - With only 27 frequencies in your cablecard map, does that mean you have SDV in your area? Do you have tuning adapters? Were they connected when you did the direct feed test (that may reduce your direct feed signal some)?

[jziggity]

TWC definitely changed something in the last 24 hours. If I try to tune 4-1 or 6-1 directly on my TV (coax feed directly into back of the TV) I get a TWC provided screen that says "Please reprogram your TV in order to continue receiving this channel." Those channels are the cablecard equivalent of 1004 and 1006. After I run the TV's channel scan again, then they work (still on 4-1 and 6-1). I have a feeling TWC reassigned a couple of channels as previously I never saw a frequency in the list that low in the CetonSignals program. I think they reassigned these 2 channels to frequency that is much, much weaker. I noticed the same problem on 2 TV's at work last night, on the other side of town. TWC provides cable at work too, and it's just a straight cable feed into the back of 2 TV's. There I also had to rescan them to get the same channels to work again (4-1 and 6-1).

I'm not sure if channels 1004 (NBC HD) and 1006 (FOX HX) are clear QAM? To get them on my TV's by tuning directly, I have to tune it as 4-1 and 6-1, so I'm guessing they are not then?

Reference the amp, it is an Electronline EDA-FT08300 that I bought here: http://www.cabletvamps.com/Products/EDA-FT08300.htm It wasn't cheap, and It has been working great thus far.

Before I rescanned my TV's, when I tried tuning 4-1, 6-1, and a few other HD sub channels, the TV kept reporting "Poor Signal Quality." After the scan, no problems, they are working fine again.

Yes, we do have quite a bit of SDV in our area, and I'm using a TA between the wall jack and the coax input for the Ceton card. I haven't changed anything there at all - same setup for the last couple of months. I think I'm going to get on the horn with TWC and have them start looking in to this. Since the exact same issue was replicated at work, I can't hardly believe that I'm the only one with this problem. I'll let you know how it goes!

UPDATE: I just got off the phone with TWC. They did confirm some changes to the channel lineup very recently, which was the cause of the "Please reprogram your TV" message. A lot of people were calling in about that. Beyond confirming that, regular customer support was not able to assist me further, so I now have a tech appointment tomorrow morning at 8AM. I'm hoping during tomorrow's appointment is where the CetonSignals program will really start paying off. It should be clear for the tech to see what is going on. Oddly enough, given the new, much lower S/N values on these 2 channels (which I watch a lot, of course), I'm not seeing nearly as much pixilation and audio drop-outs as I would have expected. We'll see what happens tomorrow!

[WhatHappend]

The channels are ClearQAM since you receive them on the TV without a cablebox or cablecard.

That AMP has a forward gain of 3dB per port, so if you just bypassed the amp for the test, the bypassed connection should have been 3dB less than the connection through the AMP. Something seem wrong to me or I didn't understand your setup for the two signal captures.

Cablecards auto remaps the channels for you when the cable company changes frequency assignments. As you figured out, without a cablecard you have to re-scan for the TV to find the re-mapped channels. That is one of the advantages of having a cablecard tuner, you don't miss recordings when the cable company adjusts their channel line-up.

[crawfish]

Isn't it recommended to use a splitter rather than daisy-chaining the TA and tuner? If you have a good splitter, e.g. a Regal, and a good, short RG-6 cable, you might give it a try... It is conceivable it could help with the macroblocking. The SNR problem is new and different and reproducible across town, right?

[jziggity]

Well, I don't have a way to test S/N at work, but the evidence of the same channel remapping was there. The work TV's also showed the "Please reprogram your TV" message, and then started working again right after the rescan. I do have a splitter that I picked up yesterday to move the cable feed to the Ceton card around the TA, instead of through it. At this point though, with this new S/N development, I'm going to hold off on doing that until after the tech visit tomorrow to see what he (or she) says. (Random question, has anyone ever seen a female cable field service tech? I haven't.)

WH, the test was conducted just as you previously suggested. I used a female-female coupler to connect the incoming coax in my basement to the feed that runs directly to the wall jack that feeds the Ceton card. That was the "direct connection" test results. The "through amp" results were obtained running through the splitter/amp as I normally would. The results that I posted jive with what you're saying about 3db forward gain. Are you saying though that the 3db should only affect the Signal column and not the SNR column?

I really feel that TWC jacked up the signal levels by moving 1004 and 1006 to such a low frequency that is clearly not providing a clean enough signal. We'll see what the tech coming out tomorrow says - I really hope he knows his stuff when it comes to all this. It was humorous actually when I called in, I could just hear the blank expression on the face of the customer service rep I talked. I tried to explain to her what my setup was like, what the signal levels meant, how I was measuring them, channel and frequency assignments, etc. I could tell that was way above her head, because she told me as much. I don't expect regular CSR's to have an intimate working knowledge of the cable system though, so no fault on her part.

[jziggity]

Well, the cable tech visit was largely a waste of everyone's time. He brought in a meter but didn't bother hooking it up. When he saw my then-current SNR level for the lowest frequency (57000), a value of 32.2, he said that was still well within their specs. He said they don't even bother looking at it unless it is below 29. He said if it goes below that (which is was for a bit yesterday) that the engineers would have to look at that. So my natural follow-up question is, how do I get in touch with the engineers if that happens again? In classic TWC fashion, he looks at me, shakes his head, and says, "You can't." Of course not, why would I be able to...?

He did look at the wiring, and immediately complained that my splitter/amp was not provided by them. He also complained that all non-TWC provided coax runs were loose. I do have commercially purchased coax runs that have boots on the ends which are admittedly loose, but just the boot part, not the threaded screw-on part that actually makes the connection - the threaded parts are on tight. I know the entire cable system can be very sensitive, but could it really be the cables that I used (and have already run through walls) are the major culprit? Could it really be that simple of an explanation? Here is an example of the coax cables that I purchased: http://www.cablewholesale.com/specs/rg6 ... -01115.htm

I don't agree with the tech's response of "we don't look at anything unless it's below 29." That seems to fly in the face of everything posted here and online. Man how I wished I had gotten a tech who was actually interested in spending some troubleshooting time with the customer. This guy just wanted to get the heck out of Dodge.

Of course as I was showing the tech the real-time SNR values by refreshing the tuned channel on the Ceton card webpage, they had climbed even higher to 34.5. With the cable tech visit being a bust, I will continue to monitor the issue. I have seen a lot less of the macroblocking to be honest. I watched a 30 minute program on FX last night (channel 1205) and there wasn't a single blip. That's good news, I suppose. But given the readings for that channel, that's what I would have expected.

I still have other things to try, like using the splitter to go around the TA, or the amplifier, to see if that makes any difference. For right now though, I'm going to let everything sit as it is for a day or two, just to get a baseline feeling of what I'm dealing with again since TWC moved channels around.