I guess this will only work when tuner sharing is enabled. Is this different from HLS ?
Yes this is a new implementation of Tuner Sharing
It is not related to HLS
Thanks will turn it on not worried if I have bad recordings this eve. Will report back.
One thing I notice that if you are watching a channel local and the same channel remote (LTE) it will use a second tuner on that device ? .... is this normal?
The behavior should be the same between both modes, so if its not sharing a tuner when it should be that sounds like a bug.
I see what the problem was one of my clients did not connect to the DVR so it just used the tuner directly working ok now sorry about the false alarm.
The buf is always 0 ... is that expected when it is running smoothly.
Watching ch744 CNNHD from Edwin's 2nd Fire TV Cube: strength=85% quality=100% symbol=100% rate=3.9Mb/sec buf=0%
Watching ch744 from 72.106.11.37 (Remux Running: 1m35s @ 1.02x (59.90fps)): strength=86% quality=100% symbol=100% rate=3.7Mb/sec buf=0%
Yes, being empty is the preferred state. It will only show above 0 if there is something wrong going on.
All recordings were fine no problem.
One of my older AppleTVs that routinely stopped streaming after 10-15 minutes worked flawlessly last night. Will keep an eye on it, but so far this seems like an improvement.
Yes, I have server side setup for Tuner sharing for all devices. After updating yesterday to the latest DVR pre-release I enabled the new streaming buffer. Just a single evening of use so far, but this particular older appleTV has always given me trouble with streaming stopping after a watching for a while.
I just saw this thread, my issue was more similar to what was referred to here, but applied to all sources, local HDHRs and IP based streams. IP streams do no reconnect automatically when a connection is lost - #14 by eric
Just noticed this during a recording and I have the Experimental New Streaming Buffer unchecked.
Is this the new normal?
It logged this when the recording finished
[SNR] Buffer statistics for "TV/Star Trek Voyager/Star Trek Voyager S01E03 1995-01-30 Time and Again 2022-08-12-1958.mpg": buf=0% drop=0%
Yes, this is the new normal. We are now exposing the buffer usage for all streams.
Without the New Streaming Buffer enabled the drop= statistic will always be 0% (the old buffer is unable to drop) and the buf= percentage will be the same for all clients watching the same channel.
Sorry if off topic but can someone link me to how to sign up for beta access. Thank you
So it appears the DVR record engine is a client?
[SNR] Buffer statistics for "TV/Magnolia Table With Joanna Gaines/Magnolia Table With Joanna Gaines S04E04 Favorite Sides 2022-08-14-0957.mpg": buf=0%-1% drop=0%
I think the buf=0%-1% was caused by Channels DVR re-authing the channel to record (which happens every 8 days for this Magnolia channel).
2022/08/14 09:57:00.000806 [DVR] Starting job 1660496220-12 Magnolia Table With Joanna Gaines on ch=[6108]
2022/08/14 09:57:28.797259 [TVE] action=fill_form u=USERNAME
2022/08/14 09:57:38.911505 [TVE] action=scienceauth done=true
2022/08/14 09:57:38.911552 [TVE] action=authed
2022/08/14 09:57:41.234409 [TVE] stream timestamps: diy: start_at=2022-08-14T09:56:41-07:00 current_at=2022-08-14T09:57:01-07:00 end_at=2022-08-14T09:57:09-07:00
2022/08/14 09:57:41.234598 [TNR] Opened connection to TVE-Comcast_SSO for ch6108 DIY
2022/08/14 09:57:41.267351 [DVR] Recording for job 1660496220-12 from TVE-Comcast_SSO ch6108 into "TV/Magnolia Table With Joanna Gaines/Magnolia Table With Joanna Gaines S04E04 Favorite Sides 2022-08-14-0957.mpg" for 34m59.999087284s
2022/08/14 09:57:41.616220 [IDX] Generating video index for job 1660496220-12
2022/08/14 10:32:00.001037 [SNR] Buffer statistics for "TV/Magnolia Table With Joanna Gaines/Magnolia Table With Joanna Gaines S04E04 Favorite Sides 2022-08-14-0957.mpg": buf=0%-1% drop=0%
Yes, recordings go through tuner sharing
The buffer increases when video data has to be saved in RAM before being received by the client. Maybe in this case the disk was slow to wake up so the memory buffer spiked slightly.
Anything less than buf=100% is normal. When the buffer reaches 100% then drop= will start increasing.