Not all EPG data shown in Guide
|
2021-01-18, 07:37 PM
(2021-01-18, 06:31 PM)mvallevand Wrote: No that's the same only 39MB I'd expect bigger and there is no epg data in it. That can't be right -- I just checked the link in incognito mode, and it shows the file's size as 143.59 MB https://www.dropbox.com/s/2e4b8w50oiyvdt...re.ts?dl=0
2021-01-18, 08:07 PM
Sorry I checked the first file twice. It looks like there is 12hours of data for the five channels on that frequency but I guess it doesn't get updated every hour so you'd have to figure out when it rolls.
Code: dump_epg_event: id:5 - 58.1: CW58 2021-01-18 10:00-10:30 Family FeudMartin (2021-01-18, 08:07 PM)mvallevand Wrote: Sorry I checked the first file twice. It looks like there is 12hours of data for the five channels on that frequency but I guess it doesn't get updated every hour so you'd have to figure out when it rolls. Ok, so I captured that at 11:46am, and, in just those 60 seconds of transport stream, it had 12 hours of guide data? So, why is NextPVR only receiving three hour chunks of guide data that are somewhat in the vicinity of whenever I click the Update EPG button? If the stations didn't add the EPG data to the stream every hour, as you're hypothesizing, then I'm imagining that I would sometimes get *no* guide data, and other times get 12 hours of guide data. But, that's not what's happening. Instead, across all channels, no matter when I click the Update EPG button, I always get a three hour chunk of guide data that is somewhere in the vicinity of the current time. This doesn't seem to match what was in that 60 second sample. Edit: Like right now about 2:40pm, I cleared the EPG data, then clicked Update EPG, and I ended up with guide data that starts at 12pm and ends at 3pm. Besides it not being useful to get guide data for all the shows that already aired for over the past two hours, it gives me no data on the shows that will start in 20 minutes! Anyway, it seems like a bug in NextPVR, as it can't just be by chance that you were able to get 12 hours of data out of that 60 second sample.
2021-01-18, 08:59 PM
Correction if you captured that at 11:46 you only had about 10 hours of data. Typically more than 60 seconds will loop the EPG, this is TV sets not PVR use.
Hard to say about the specific issue since you didn't post any logs of an EPG scan. It is known that ATSC EPG is low priority for sub especially when a good source of EPG is available The logic is different for each o/s platform and tuner type and getting DVB working (which this thread is about) would be much higher as it often gives reasonable data. However I expect in time sub will be able to use the file you provided to see what might be the issue Martin
2021-01-18, 09:15 PM
(2021-01-18, 08:59 PM)mvallevand Wrote: Correction if you captured that at 11:46 you only had about 10 hours of data. Typically more than 60 seconds will loop the EPG, this is TV sets not PVR use. Thanks, for sure, no argument at all, I'm willing to provide logs and make a separate thread specific to ATSC EPG, too, if that would be preferable. I can work on that now. There's just clearly something inconsistent going on here, in spite of the logic potentially being different for each OS platform (Windows 10 in my case) and tuner type (the HDHomeRun CONNECT, HDHR4-2US, which was a quite popular previous generation unit). Basically, the output doesn't match the input. 
2021-01-18, 09:28 PM
As my capture example showed just for the HDHR there are two methods needed on Windows to get the data. Once captured the EIT guide data is send in many formats in the stream each one neededing to be decoded Then the data for multiple channels needs to be extracted and formatted combining the title, time and description. From there the data has to be store in the database. It is easy for me to see how it might not be perfect.
NextPVR is closed source with one developer who doesn't have access to ATSC except with files like yours. He has to prioritize his time and it's no doubt a labour of love. I'm speculating here but users who find $2 a month for guide data too expensive probably aren't that eager to donate for NextPVR either. Martin
2021-01-18, 09:39 PM
(2021-01-18, 09:28 PM)mvallevand Wrote: As my capture example showed just for the HDHR there are two methods needed on Windows to get the data. Once captured the EIT guide data is send in many formats in the stream each one neededing to be decoded Then the data for multiple channels needs to be extracted and formatted combining the title, time and description. From there the data has to be store in the database. It is easy for me to see how it might not be perfect. Agreed on the first part. No denying it's complicated. On your second point, all I can say is that if there's a fair chance he can get this HDHomeRun ATSC EPG guide scenario working more effectively, I'm willing to donate directly to NextPVR's future development. I've already liked what I've seen so far. Full disclosure, I've donated to other devs before, like krustyreturns. Does emveepee have a donation link on GitHub? He deserves something, too. 
(2021-01-19, 03:25 PM)sub Wrote: Yeah post the logs and I'll see if anything stands out. These are verbose logs from where I've emptied the EPG, and then updated the EPG (they don't overlap with that previously posted capture, but still exhibit the same issue). This is all with having the ATSC devices selected, not the HDHR devices, just to be clear: https://www.dropbox.com/s/sn9meldd7n3ffi...9.zip?dl=0 (2020-12-24, 03:59 PM)sub Wrote: Ah, sorry - I didn't think to check it on an 'HDHR' style device. That fix was for BDA style device, ie USB/PCIe device connected to the machine, or HDHomeRun devices using the BDA drivers ('DVB-T' / 'DVB-C' device types on the Devices screen). The 'HDHR' style EPG is done a slightly different way. I'll check that too and fix. Any update on this? Do you want me to test this on my Windows and docker instances? Is this fix scheduled for next release, if so any ETA? Thanks appreciate it. |
|
|
