Jump to content


Welcome to DBSTalk


Sign In 

Create Account
Welcome to DBSTalk, like most online communities you must register to view or post in our community, but don't worry this is a simple free process that requires minimal information for you to signup. Be a part of DBSTalk by signing in or creating an account.
  • Start new topics and reply to others
  • Subscribe to topics and forums to get email updates
  • Get your own profile page and make new friends
  • Send personal messages to other members.
 
Guest Message by DevFuse

Photo
- - - - -

Recordings start late/end early


  • This topic is locked This topic is locked
26 replies to this topic

#21 OFFLINE   tsantanni

tsantanni

    Cool Member

  • Registered
  • 22 posts
Joined: Nov 08, 2007

Posted 27 November 2007 - 06:55 PM

I have two HR-20 700's and one exhibits this issue and the other does not. The thing I notice is the time will switch to say 8PM, but if I have the guide up when it goes from 7:59 to 8:00, the guide delays about 20-30 seconds later before the 7 time slot dissapears and then the recording starts. The clock appears to show the right time so its something else that's causing the sync problem.
Samsung HL-S6188W 1080P 61" DLP
HR20-700 & Cavalry 750GB eSata HD

Samsung LN-S4692D 46" LCD
HR20-700 & Cavalry 750GB eSata HD

Philips 37PFL5322D 37" LCD
H20-100

...Ads Help To Support This SIte...

#22 OFFLINE   RoberMR

RoberMR

    AllStar

  • Registered
  • 70 posts
Joined: Oct 23, 2007

Posted 27 November 2007 - 08:35 PM

I think Tiger nailed it. It is most likely a thread synchronization/prioritization/optimization issue. It is also possible that the HR20 is wimpy in the CPU department. Maybe it just has trouble keeping up with everything that has to be done all at one time. But my guess is that it's most likely poor programming techniques used in the design and development of the code. When I used to be involved in designing and developing software (many years ago), we spent a good bit of time on optimizing the execution of the code that was written before we shipped it. Nowadays, it seems like if a piece of code "works", it's stamped as finished and is shipped without any consideration for reliability, performance and defensive abilities to handle unexpected events. I am not singling out DirecTV with this statement, it is an industry disease that has taken over the IT world.

Just as a curiosity thing, has anyone run into this symptom with the R15? It would be interesting to find out if the R15 also has the same issue.

#23 OFFLINE   Dr. Booda

Dr. Booda

    Godfather

  • Registered
  • 385 posts
Joined: Jun 18, 2007

Posted 27 November 2007 - 09:32 PM

I just did an apples to apples test between a SD DirecTivo I have and the HR20. And definitely, the HR20 starts recording about 20-30 seconds later than the DirecTivo. Here are the details:

I set up the same SD program from the guide in both machines starting at 9:30 pm with no padding or anything. I let them record for about 5 minutes and then went to play them on my split screen TV set side by side. The Tivo based machine started recording right on time while the HR20 took at least 20 seconds to 30 seconds more before it started recording.

Same guide
Same satellite
Same multiswitch feed
Same channel (SD)
Same program
Same time

For some reason, the HR20 takes longer to start recording.

It is not the networks, not different timing between the SD and the HD version of the same channel. It is the timing in the HR20.

I hope this helps the DirecTV engineers.


Thanks for performing a functional test that should have been done a year ago at D*. I have observed this issue for 12 months now, and it indicates bad engineering and poor oversight. If the hardware is overburdened, can we please have an upgrade in future HR generations?

#24 OFFLINE   TigerDriver

TigerDriver

    Electronics Engineer

  • Registered
  • 570 posts
Joined: Jul 27, 2007

Posted 28 November 2007 - 12:12 AM

It is also possible that the HR20 is wimpy in the CPU department.


Possible, but not likely. These boxes are full of DSPs, disc controllers, etc, all of which do DMA. Anyhow, 20-30 seconds is an eternity in CPU time.

I smell a deadly embrace between two threads, each waiting for the other to complete.

IMO, this issue ought to be triaged to the top of their bug list (if they have one)--right along with the bug that causes Dolby Digital to get mangled for 20 seconds after FF/RW/etc.
Joe

HR20-100, HR24NC-100, HR21-700
Samsung HLS5687WX 56" DLP
Onkyo TX-SR705 A/V receiver
Samsung Blue-Ray/DVD player
4-channel NTSC modulator

#25 OFFLINE   RoberMR

RoberMR

    AllStar

  • Registered
  • 70 posts
Joined: Oct 23, 2007

Posted 28 November 2007 - 01:05 AM

Tiger, I haven't heard the phrase "deadly embrace" in a long time. We used it often when our sofware got all tangled up in its own underwear.

Thanks for goving me a chuckle before I sign off for the night.

#26 OFFLINE   Milominderbinder2

Milominderbinder2

    Cutting Edge: ECHELON '08

  • Registered
  • 4,107 posts
Joined: Oct 08, 2006

Posted 28 November 2007 - 11:47 AM

I still have this problem. Every single recording starts late.

Part of the problem with my HR20's is with their inabillity to keep the correct time. I have two HR20's and have tested them split screen, scheduled to record the same show. They never start at the same time. The clocks are never the same. Neither ever matches my atomic clocks or cell phones (which are right).

I think that there is another part of the problem.

Say they program the recordings to start exactly at the right time. Let's even assume that their internal clock is exactly right.

When the triger comes, the HR20's still need to take about a year to change channels and then start recording. Watch your HR20 when it begins recording a show and you will see this is exactly what happens.

Do that by hand and see how long it takes. 5-12 seconds?

So they will always be late by 5-12 seconds if they begin the processs of changing channels and starting a recording "on time".

So at least part of the solution has to be to start recordings a little early whenever possible. Maybe automatically pad a minute to the beginning if possible?

Still this does nothing to explain why they end early so consistently.

- Craig

#27 OFFLINE   rkreitl

rkreitl

    Duplicate User (Account Closed)

  • Banned User
  • 96 posts
Joined: Aug 24, 2007

Posted 03 December 2007 - 01:10 PM

Maybe if there wasn't code to spin those magical blue lights on the front panel the unit could actually start recording on time. :lol:




spam firewall