Announcing the general release of ReplayTV software 5.1 for RTV5000s and 5500s - Page 2  

Ok, your response helps me hone my question: do networks insert similar markers at "scene" changes?

One interesting omission from the description of this feature is the "CYA" language that SB adopted to describe the CA reliability ("we don't guarantee it will skip properly all the time"). (Maybe they'll need one down the road, perhaps. We shall see....!) Even if this feature works 90% of the time, I'll be thrilled with it.

Richard,

Good news and thanks for the update. Now there's just one thing I'm looking for so I can upgrade my ReplayTV 2020 (now that the reliabilty is back) - HDTV support.

With the release of the 5500s and Alcatraz, I hope that means that resources are available to start working on HDTV support. If that's the case can you please confirm?

Also, is DNNA/ReplayTV going to be at CES this coming January? I didn't see the name on the exhibitor list, but I wasn't sure if you might be in the Denon or Marantz booths.

Thanks,

-phil

Hint: The 5.5K series doesn't have something that the 5K series DOES have.

Great news. Hopefully this will fix the RTV/wireless bridging issues (ethernet port not working with wireless bridges). After many hours on tech support this was the best solution they had to the issue :(. Fingers crossed.

No one has mentioned the improvements on channel changing times? Richard said "several seconds" of improvement, can anyone comment on this?

Hey Lyndon... on the DNNA website you guys have the "football" point listed twice in th bullet list.

"Some people only care about the fourth quarter of the game? - With Show|Nav you can jump right to it. "

I hate contradicting heathriel (no mileage in it ;) ), but I don't think that's correct. As far as I can tell, SA will do the following:

1. If you are in the middle of a show, SA will skip to the beginning of the next commercial block.

2. If you are within a commercial block, SA will skip to the beginning of the next part of the show.

Basically, it seems like a scene is defined as wherever the show breaks to a commercial or a commercial breaks back to the show.

Successive presses will for example jump first to a commercial block, then to the beginning of the next scene in the show (i.e. where the show returns from the commercial break), then to the next commecial block etc.

So, if you want to skip the silly action part of the superbowl and 'show brittney's nav' (as RR keeps hinting), you just hit the arrow a few times and it will skip all that pesky content and jump over to successive blocks of commercial.

I don't think SA is using any fancy scene detection algorithms like the dvd authoring programs have.

Am I incorrect, here?

Yeah, this is why I'm pressing the issue because I can't believe SA can be this "smart." (or at least even remotely reliably smart).

I think your take is correct, especially since Richard went out of his way to make the point that 55xx owners will appreciate the feature which is essentially "manual" CA. (Of course, I hope you and I are wrong!)

One other benefit of SA is that for those occasions when CA misses a set of commercials, SA should be able to get you through them without having to QS all the way through. That in and of itself is great.

That is correct. SN is using the same markers as is used by Commercial Advance. If you are watching a two hour HBO movie, the only thing SN will get you is back to the beginning of the show once you start watching it.

Not exactly, it will skip to the next product placement screen... Oh, damn! that's on the list for Release 7.0 :(

That presumably means that SA should be no more or less accurate than CA itself, so that any claim that it is comes simply from the placebo effect, right?

I'm really happy to hear about the roll-out of Alcatraz. The new features are a little underwhelming to us 5xxx series owners, but it certainly makes the 55xx series much more attractive; a smart and safe way to *replace* ACA.

Now if a future release would include these two features...

1. A list of shows scheduled to be recorded (I really thought this might happen with this release, but maybe it's harder to implement than I imagine)

2. Serial control compatibility with most or all cable/sat boxes. (the IR blaster codes are a nice addition, but serial control would REALLY be nice)

It's good to know your looking into the audio hiss issue, but I'm not very optimistic it can be fixed through software; but serial control of a cable box would take that out of the equation for me.;)

Can we at least call it by DNNA's name? :) ShowNav (SN), not SA....

I *believe* (no science in this....) that there are no 'markers' per se anywhere in this. Not in ACA, not in SN. I believe (see earlier post) that ACA works on fade to black with a fade out of black some mutiple of 30 seconds away. ACA will fail if there are bugs on the screen, a 'too quick' fade. There will be a 'false positive' ACA if a 'screen wipe/fade to black' is in a show, and commercials start 120 seconds (for example) later. I've seen this on a L&O recently.

As for SN, I think it's the same: a 'scene' is a chunk of show with fade to black at the endpoints. So, of course, a commercial break will be a scene end/start. But I have seen in some shows (no hard examples, I'm at work.. :) ) where doing the SN back and forth will go between 'chunks' of a show.

Anyone want to confirm this with some experimentation?

Good point. I had forgotten about this behavior because I haven't seen it recently.

That has been my experience thus far. SN and CA are only going to be as good as the quality and strength of the signal coming into the box and the availability of the markers transmitted.

I believe signal strength and quality have a lot to do with the accuracy of CA (and now SN). I have digital cable with a signal amplifier and I would estimate that CA has worked about 90% of the time. My dad has analog cable and says CA only works about 50% of the time. I used IVS to transfer some shows I had recorded to his unit and sure enough, he got a 90%+ hit rate with CA.

I'd expect the same performance with SN.

Lee, the .EVT file contains the "markers" (for want of a better term). And it isn't just a fade to black of the video, the audio stream also gets examined by the algorithm, that's why a low audio level, when detected throughout the show, causes gross malfunction of ACA. We've grossly oversimplified the description of ACA as far as detection and performance. Obviously it's a far more complicated mechanism than it appears to be at first sight.

| CONTENT(a) | COMMERCIAL(a) | CONTENT(b) | COMMERCIAL(b) | CONTENT(c)

If the speculations above regarding SN are correct, there is a
risk that hitting SN while in CONTENT(a) could unwittingly bring you to the beginning of CONTENT(c) when you think you should be at the beginning of CONTENT(b). It seems as if paying attention to the HUD will be critical until we become convinced of SN's reliability. If you see a 6 minute "skip," the SN probably missed the mark(er).

In any event, it still should be fun to play with.

Bayside, thanks for the additions... I'd forgotten about the audio component. And, as was said, signal strength sure counts. When I moved from digital cable to DirecTV, things that were good got better.

As for the markers, I was trying to stop the reference to 'markers in the original from Hollywood', whatever. I agree that the 'recording' (encoding?) machine seems to create markers that are in the recorded MPEG.

ISTR some discussion somewhere about this being done only on the playback side, but I find that hard to believe....

Brian, you definitely need to look at the HUD to see if you are where you expect to be until you get used to it. It's perfectly possible that SN could lose a transition and be too far down the road. I plan on taking some recent recorded shows that I know where the 'scenes' are and doing the HUD/SN back and forth to see where I move to!

I haven't even seen 5.1 yet and I'm already dreaming about possibilities. For instance perhaps LeeThomson's friend will create a new rtvconvert that you can insert custom CA/SN markers into the .evt file. Then the SN will have nothing to do with commercial breaks and you can be in control of scene selection in your DVarchive show database. Can't wait for 5.1 and can't wait for all you great software developers to have at it.

I think people are missing the best part of ShowNav. It means that Replay has reason to continue to improve its Commercial Detection algorithm.

When I heard that the 5.5ks didn't have ACA, I feared that commercial detection woudl never be improved. Now DNNA has reasons to work on it.

The reason i'm asking is because i'm still debating whether or not to trade in my lone 4504 to talk to my other 5040s. The 4504 captioning works quite well, while there have been acknowledged CC problems on the 5xxx units.

Lyndon, Richard -- could someone please comment??

Am I the only one w/ networked RTVs that wishes control OF any of them FROM any of them should be transparent? It doesn't seem technically complicated to be able to perform all the functions on another (networked) replay that I can do from the one receiving the remote's ir signal. It's aggravating to have to go to another room (or basement in my case) to see exactly why I can't record some show 'on a networked replay', etc etc.

Also would like to be able to watch a show being recorded on another replay remotely, while it is in progress (no I don't need live, just a scheduled recording, like I can do on a 'local' RTV). No mention of improving this stuff in this release. While stability is definitely job 1, I hope the next release will make it irrelevant WHICH Replay on the network I'm pointing the remote at, functionality should be the same - this isn't that hard to do, at least conceptually (hardware constraint's another story).

That said, the skipping feature ("SN", we seem to have tagged it) is welcome. Though I've gotten used to "number, skip fwd" too often I don't get the skip button recognized in time so it does a channel change and everyone growls.

Oh, is the speeding up of the reaction time gonna help? I have to be very careful to aim the RTV remote, it does not seem to have a very "broad" beam and between that and the delay, even after a year's experience I have trouble getting the remote signals recognized consistently. No software will help the beamwidth but at least I shouldn't have to wait 2 or 3 beats to see if the remote command was accepted.

Finally YAY for the improved IR blaster, hope it fixes the occasional glitch I have trying to control the Philips DirecTV box (why oh why did they stop putting serial port control in the tuner boxes?).

PS to the HDTV'ers: think about program content encoded at the source and transmitted digitally all the way to the hard drive in some box in your house: you essentially have an exact copy of the master, and could dupe to your heart's content. Think Hollywood will sit and watch perfect copies of their product be distributed for (between, say) $0 and $65 / month? Their secondary market would disappear (DVDs, syndication, etc) overnight. Sometimes that's the only place they make money on a product.

HDTV PVRs will have some copy protection, or they will be sued into SonicBluedom. Watch and see, and look at http://www.avsforum.com/avs-vb/showt...71#post1620871 for the whole spiel.

Can anyone say if the remote control buffering problem has been fixed (or at least improved)? Many times my 5040 will see 2 button presses, when I only pressed it once. Even more annoying is when I do press it 2-3 (or 4-6) times because I can't be sure any of them were detected, only to have all of them be processed about 15 seconds later.

It seems that my 3060 and ShowStoppers behave much more intuitively, by only buffering 1-2 remote clicks at any time.

There he goes teasing us again.......

Maybe I missed it... is the Screen saver bug fixed?

I don't have mine turned on since it has been known to cause problems, but I'd love to turn it on and start using it. That'd be waaaay cool.

IS IT FIXED??

They said every known lockup or hang.

Bayside, you ever make the test video? I'm still curious about the results. I'm sure we are pretty close in understanding how and what the detection looks for, but I'd love to see your test results, I'm most interested in the fade to dark greys with a IRE level of 10-30... =)