1.5.28 Skip settings for new recordings

[skinny]

I've been hanging out for the next release of NPVR, I thought this was going to help things...

- added a workaround to give more accurate skipping when playing .ts recordings (for comskip users etc). New recordings on NTFS drives are those that will benefit from these improvements.

I'm having trouble with the skip functions in new recordings made with V1.5.28 they do not adhere to the settings of SkipSecondsForward and SkipSecondsBack in Config.xml
I was playing around this morning and to get it close to what I want, I had to use something like this

- to skip forward approx 60 seconds I needed to set the SkipSecondsForward to 40
- to skip approx back 20 seconds I needed to set the SkipSecondsBack to 40

Old recordings and avi files adhere to the settings in the config file

Is it just me, or is this a bug?
Cheers

[johnsonx42]

I'm getting reports of the same sort of thing from my wife; I've been gone for a day-and-a-half. I've come home to a wife who has absolutely had it with NPVR 1.5.28. She reports that skipping based on comskip is more off than ever, and manual skipping either doesn't work at all or goes the wrong way. I've not had chance to play with it yet myself, but will expirement and make a full report in the next day or so.

[skinny]

I'm getting reports of the same sort of thing from my wife; I've been gone for a day-and-a-half. I've come home to a wife who has absolutely had it with NPVR 1.5.28. She reports that skipping based on comskip is more off than ever, and manual skipping either doesn't work at all or goes the wrong way. I've not had chance to play with it yet myself, but will expirement and make a full report in the next day or so.

Yeah, when I had the SkipSecondsBack set at 10, it would actually go forwards at times

[XJW]

I am having trouble with skipping in TS files. I can select skip forwards and it will actually skip backwards. The behaviour can appear to be quite random.

[johnsonx42]

I had just a bit of time to play with this last night, and found that for some new recordings skipping with comskip is dead-accurate, every bit as good as it was with GB-PVR and DVR-MS. For others though the skip points are not even close. Likewise, for the recordings where the auto-skips were spot-on the manual skipping was perfect too, while the ones where the auto-skips were off, the manual skips were screwy. As far as I know, the channels in question both broadcast 720p so it's not as simple as 1080i vs 720p.

[sub]

You could see if this patch helps.

It uses the duration from the Timing.Info stream, giving more accurate reporting of duration and current playback position, which should give more accurate skipping.

I should mention these new skipping features are only relevant to completed recordings, not in-progress ones.

[skinny]

Thanks for the patch Sub,
unfortunately no apparent improvement Here's 2 examples

TV1 News
Skipping forward 60 sec - actual skip time forwards 50, 61, 66, 54, 41, 30, 30
Skipping backward 20 sec - actual skip time back 44, 34, 24, 18, 17, 16, 26

TV3 NCIS
Skipping forward 60 sec - actual skip time forwards 74, 82, 92, 100, 101, 106
Skipping backward 20 sec - actual skip time back +31, +28, +20, +19 (going fwd not back)

Cheers

[sub]

Do you see the timing information if you run:

notepad "c:\path\to\recording.ts:timing.info"

[skinny]

Do you see the timing information if you run:
notepad "c:\path\to\recording.ts:timing.info"

Yep it's there, below is a snippet from the One News recording, but when I try
notepad "R:\NCIS\NCIS_20100808_20302130.ts":timing.info" it says it's too large for notepad

Code:

0,0
920,670784
1887,1341568
2839,2012352
3791,2683136
4742,3353920
5709,4024704
6661,4695488
7613,5366272
8471,5953208
9422,6623992
10389,7294776
11341,7965560
12293,8636344
13151,9223280
14102,9894064
15007,10564848
15974,11235632
---------snip----------
3947808,3787078768
3948759,3787749552
3949711,3788420336

[sub]

You might need to try to make a short recording that exhibits the problem, and get it to me, so that we can both test with the same file.