View Full Version : Is the clock drift fixed?
Orient Express
11-10-08, 08:52 PM
Programs seem to be recorded for me with correct start times now. Is that the consensus? Did the last SW revision fix this?
Doug Brott
11-10-08, 08:57 PM
It's good that you noticed this ..
Hard to tell here. Our unit has, uncharacteristically, locked up several times in the past few days and we've had to do RBR's to get it going again. So clock would have been re-set from that as well. Hope others have been more fortunate and can give you better perspective.
tenholde
11-10-08, 10:35 PM
One thing that I think was included in the latest release was automatic buffering. All recordings start and stop 1 minute (I think) earlier/later than the schedule indicates unless that would create a conflict with other recordins.
tenholde
One thing that I think was included in the latest release was automatic buffering. All recordings start and stop 1 minute (I think) earlier/later than the schedule indicates unless that would create a conflict with other recordins.
tenholdeHmmm. At first I thought that was just a nice benign new feature, but now it seems more like a band-aid response to the clock drift issue.
Doug Brott
11-11-08, 04:45 AM
Hmmm. At first I thought that was just a nice benign new feature, but now it seems more like a band-aid response to the clock drift issue.
No, softpad is more of a band-aid response to networks that don't follow their own schedule. I've heard of some stations starting as much as 2 minutes after the scheduled time.
Best way I've found to test your DVR's recording start accuracy is to ignore the INFO bar clock and put "time.gov" up on your web browser a minute or two before a recording is scheduled to start.
Do it for situations where the recording light is off, meaning there are no other recordings happening before the one you're checking. When the RECORD LIGHT* turns on, check the browser time of day. Now, when you have issues with the viewed recording start, you can tell if it's the network or DirecTV.
When I've tested my units as above, I find that my HR20-700's are starting about 4-5 seconds late, on average, and my HR21's about 13-14 seconds late, on average. When a show can be "soft padded", this is generally not an issue. I have several shows scheduled to record during prime time that can't be soft-padded, however, because of previous recordings on different channels tying up the tuner, so I still sometimes miss a few seconds of the starts of those shows.
/steve
*NOTE: If a recording is eligible for soft padding, the RECORDING light will still come on at the scheduled time, even though it may actually have started 30 seconds earlier. If the light comes on early or late, the start of the softpad will be correspondingly early or late. Manual recordings are never soft padded.
Doug Brott
11-11-08, 05:48 AM
Thanks Steve, I'd be interested in seeing what results folks see ..
StephenT
11-11-08, 06:36 AM
Yes, there is soft padding now, but if I'm not mistaken the soft pad at the beginning of a recording can only be seen if you rewind after you start watching. I've noticed that without having to do that my recordings are starting on time where in the past they weren't.
It seems to me soft padding was added AND the clock drift was fixed.
Soft padding is pretty cool by the way. Shows don't only start early they sometimes run late. I've found more than once where I would have missed the ending of a show if I didn't have it.
The soft padding helps. There is still clock drift. If I record the same program on all four DVRs no two of them start at exactly the same time. There is generally a differential of about 5 to 10 seconds between the earliest starter and the latest.
trainman
11-11-08, 10:02 AM
My check on the DVR clock is "Late Night with Conan O'Brien," which I have a Series Link for, and which I happen to know NBC starts at exactly 12:37:00. Before version 0290, the recording would consistently start 1-5 seconds late; now it's consistently starting 1-5 seconds early (always includes a little bit of the very end of Leno's closing credits).
My check on the DVR clock is "Late Night with Conan O'Brien," which I have a Series Link for, and which I happen to know NBC starts at exactly 12:37:00. Before version 0290, the recording would consistently start 1-5 seconds late; now it's consistently starting 1-5 seconds early (always includes a little bit of the very end of Leno's closing credits).Which model HR? TIA. /steve
frederic1943
11-11-08, 11:32 AM
I've noticed that the shows are recording from the begining now even when recording on two channels. On Mondays I record "The Big Bang Theory" & "How I Met Your Mother" on CBS and "Terminator" on FOX. "The Big Bang Theory" includes the first 30 seconds or so of the first part of "How I Met Your Mother" at the end. It used to be that "How I Met Your Mother" would start recording where the over-run on "The Big Bang Theory" ended. Now "How I Met Your Mother" starts before the begining and includes the overlap from "The Big Bang Theory".
I've noticed that the shows are recording from the begining now even when recording on two channels. On Mondays I record "The Big Bang Theory" & "How I Met Your Mother" on CBS and "Terminator" on FOX. "The Big Bang Theory" includes the first 30 seconds or so of the first part of "How I Met Your Mother" at the end. It used to be that "How I Met Your Mother" would start recording where the over-run on "The Big Bang Theory" ended. Now "How I Met Your Mother" starts before the begining and includes the overlap from "The Big Bang Theory".This means the HR scheduled both CBS shows on the same tuner, as it should have, so you'll always get 30 seconds of add'l pre-padding on the second show and a full 90 seconds add'l post pad on the first show. This "back-to-back on same channel" padding should be the same, regardless if you're recording or watching another channel LIVE on the other tuner.
I say should, because I recently experienced a rare case where the HR2x scheduler recorded back-to-back shows on the same channel on two different tuners, but that normally shouldn't happen. I knew it took place because I was expecting the "end of first show/beginning of second show" overlaps, and they weren't there.
/steve
Best way I've found to test your DVR's recording start accuracy is to ignore the INFO bar clock and put "time.gov" up on your web browser a minute or two before a recording is scheduled to start.I've got a clock that syncs every midnight with WWV in Ft. Collins, CO and in the past my HR2Xs have started as much as 45 seconds after the correct time. I don't care if the problem is fixed via software or the soft padding is doing it, I am finally seeing all the programs. ;)
I've got a clock that syncs every midnight with WWV in Ft. Collins, CO and in the past my HR2Xs have started as much as 45 seconds after the correct time. I don't care if the problem is fixed via software or the soft padding is doing it, I am finally seeing all the programs. ;)SOFTPAD is just a band-aid that will only work when it has a free tuner available on either end of the recording.
If you're recording CBS and NBC from 8-9, and ABC and FOX from 9-10, none of the 8PM shows will be end-padded and none of the 9PM shows will be front-padded. That's why the recording start times still need to be accurate. /steve
The time drift problem is always "fixed" after a reboot, including one that follows a software update. There were two software updates recently, so the time drift problem would not have been as noticeable of late. I like the recording starting on time (give or take a few seconds) at the moment, but will that continue a few weeks from now?
In the case of my HR21, the drift is gradual and takes at least a couple of weeks after the latest reboot to be noticeable. Until that much time has passed, I can't say one way or the other whether it's been fixed for real.
Doug Brott
11-11-08, 01:54 PM
The time drift problem is always "fixed" after a reboot, including one that follows a software update. There were two software updates recently, so the time drift problem would not have been as noticeable of late. I like the recording starting on time (give or take a few seconds) at the moment, but will that continue a few weeks from now?
Check back in a few weeks .. from what I understand, you should be reporting positive results.
trainman
11-12-08, 10:41 AM
Which model HR? TIA. /steve
HR21-700, as noted in my signature.
Monday night's Conan recording actually started exactly on time.
HR21-700, as noted in my signatureD'oh! :) My HR21's now start on time as well, but just for padded shows. When there's no opportunity prepadding, there's been no change... my HR21's still start up to 14-15 seconds behind my HR20-700's. I hope Doug is hinting that a start-time fix for each model may finally be in the works, because it doesn't seem like "one size fits all" works when it comes to this issue.
/steve
vBulletin® v3.7.6, Copyright ©2000-2012, Jelsoft Enterprises Ltd.