DVD Player W/Out The Layer Change? (1 Viewer)

[John Kotches]

I played with a Toshiba 5700 earlier this year (for about 2 weeks). I was impressed with it sonically for the price, but the deinterlacing was textbook Genesis (read bad).

Regards,

 

[Dave E H]

Hello!

I am thinking about getting a Pioneer DV-47A. The girlfriend just hates the layer change on our SD6200 - anyone know if this is any better on the Pioneer?

If not, it may make me think twice about making the jump into SACD/DVD-A. at this point in time.

 

[Jeff D]

John, you don't like the RP-91 deinterlacing? That uses the Genesis, right? Interesting, I didn't think the RP-91 and Tosh players both use the Genesis?

 

[StaceyS]

The Tosh 4700 is not using a Genesis. It is integrated into the Zoran Vaddis V.

Previous Pioneer players have always had the slow layer changes. I Have not seen the 47a thought.

Apex A600, Denon 2800, Meridian 800, Meridian 596, WinDVD (PC), PowerDVD (PC), and XBOX are the fastest that I have seen.

 

[John Kotches]

Didn't know that Stacey. Thanks for the tip -- that's part of why I don't review the players for video
Regards,

 

[Wayne Bundrick]

The layer change should not depend on seek time. So that seamless branching will work, DVD players are supposed to have enough buffer memory to withstand a "full stroke" seek of the laser during wide open throttle 10 Mbps video.

The problem, which I think someone already mentioned in this thread, seems to be that DVD players flush the buffer before changing layers.

 

[Dave E H]

Any one know of a DVD (and it's location) that's particularly problematic? I'm going to go try out a DV-47A soon.

 

[Ian Montgomerie]

The layer change should not depend on seek time. So that seamless branching will work, DVD players are supposed to have enough buffer memory to withstand a "full stroke" seek of the laser during wide open throttle 10 Mbps video.

The reason non-seamless transitions (and thus most layer changes) depend on seek time is due to the traditional way of implementing DVD player control code. You have playback ranges A and B, and between them you have to reset the audio and video decoders due to the non-seamless transition. The easiest way to do this is play range A, wait for it to finish, then reset the decoders, then play range B. Because range B doesn't start reading until range A finishes, there is a seek while no new video is being displayed. The seek could be avoided, it would just mean a different and harder way of doing the software.

It wouldn't produce a zero-time layer change, though. Still need to reset some underlying hardware. But the 1/10 of a second or less that would take, would be less noticable.

 

[Jeff]

Even better, someone has to come up with a DVD player with better error correcting. Working at a video store, we get so many customers bringing back DVD's now because they are scratched or dirty. What I really hate is when they exchange it for a VHS copy.

Jeff

 

[Han]

You know what, at this point, I'd rather people exchange DVDs for VHS tapes than say a widescreen DVD for the Pan & Scan version. But anyway, that's neither here nor relevant to this thread...
Dave E H, "Saving Private Ryan" has a pretty jarring layer change. One, I've watched recently though I doubt many people have it, is "Ed TV" which also has a noticeable change. "Armageddon" the Buena Vista (non-Criterion) version has a pretty awful layer switch too. As far as Pioneer players, I heard some older ones had trouble with "Independence Day" and "T2:UE" for the seamless branching, but I'm not sure about the newer players. I'd hope something like that would have been caught by now.

 

[Dave E H]

Thanks! I've got SPR DTS - I'll boot that up and see if I can find the layer switch. Anyone know where it is?

 

[Kwang Suh]

Magnolia has a terrible layer change as well.

 

[Dave E H]

I just tried ou the DV-47A - with Rush Hour 2 (layer change during Red Dragon Scene) it didn't seem any faster than my SD6200.

Audio performance was pretty spectacular on that player though - video was tough to tell since the display is different than mine at home.

Might be in the cards though - for $750 delivered from some places, that's a screaming deal.

 

[Ian Montgomerie]

Even better, someone has to come up with a DVD player with better error correcting. Working at a video store, we get so many customers bringing back DVD's now because they are scratched or dirty. What I really hate is when they exchange it for a VHS copy.

DVD player error correction is already pretty good in most cases. In fact the real tradeoff is knowing when NOT to use it. The various methods to fix errors (try to read it again, apply the error correcting code, try to cover up errors in the decoder, etc.) all take time. More time than it would normally take to play the data if it had no errors. Sometimes much more. If a disc is badly scratched or really dirty (hairs or strands of fabric can be a nasty case), there will often be long sequences of playback, say 30 seconds, which are chock full of errors.

What happens if you try and apply maximum error correction to this is that playback slows to a crawl, and it will still look really bad because most of the errors are uncorrectable anyway. So once a certain level of errors are encountered, the player has no choice but to either slow playback to a crawl, or to start skipping ahead to try and get past the bad data as quickly as possible.

The core problem is physical, not electronic. VHS is highly resistant to dirt because it uses a mostly-enclosed case, and the surface area of the tape is huge compared to the size of a given piece of dirt. DVDs are physically easier to beat up if you throw them around "naked". And it's not like people have a very high incentive to take care of rental DVDs. On the other hand, if you don't abuse them, DVDs will last much longer. They experience essentially no wear from the act of being played, unlike VHS.

Even so, I must say that I've rented a fair number of DVDs, and have only had serious errors with a few - all fixable by wiping with a damp cloth.

 

[Paul Dube]

There are probably a couple of other DVD players that can go through the layer switch without a hiccup, one of them being the Proceed PMDT.

I searched for such a player for a long time. I owned some Toshiba, a Sony 7700, then a PMDT. I could not find a player that would be able to pass the layer switch without a pause without having to pay several megabucks.

What I am wondering is why no publication talks about it. I mean I am not asking for a whole review on the subject. I would just like one sentence about the layer switch speed when they review a DVD player.

Paul

 

[Han]

Paul,

I think the reason it isn't mentioned much, is because it's just about a forgone conclusion that whatever player reviewed will pause during a layer switch.

As it's been mentioned in many past threads, the Denon players, Apex players, and a Meridian player zip through layer changes quite fast. Almost DVD-ROM drive on a computer speed, well, pretty much because those players are DVD-ROM drive based. I think Stacey mentioned a new Samsung is relatively fast on layer changes as well, but has other problems.

 

[Mark Tranchant]

My Philips DVD-762 is pretty quick. Less than a second, I'd say, and the audio doesn't break with my Marantz SR-4200.

Continuing the seek time buffer discussions, I was under the impression that there was no seeking during a layer change, it's just a re-focus with no head movement. One layer reads from the inside outwards, and then back again on the second layer. Yes?

If this is the case, the layer change break is purely a function of data buffering and is easy (but not free) to engineer out. With a maximum bit rate of 9Mb/s and up to 2 seconds for refocussing, you'd need a layer change buffer of 18Mb, or 2.25MB.

 

[John Kotches]

Mark,

A few things...

1) Not all titles are authored RSDL, some (perhaps many) are simply Dual Layer. This means seek to track 0 + refocus.

2) Most titles are designed for non-seamless layer breaks as per DVD-V specs.

3) Allow for some time to rebuffer data so that you are working from buffer instead of raw off the spindle.

4) Mechanical time to stop/restart the spindle in reverse.

The other item to consider is where the buffering occurs. If it occurs in front of the MPEG decoder, then you actually have to account for maximum raw data rate off the spindle, which exceeds 20Mb/second with all the ECC involved. Perhaps Stacey or Ian can chime in on whether the buffer is implemented off the raw spindle data, or after all the ECC is stripped. If raw spindle data, you're talking about something more along the lines of 16MB.

Both Ian and Stacey know far more about this topic than me.

Regards,

 

[StaceyS]

Mark,

Don't forget you must have enough buffer to zip through durring the highest FF speed a player supports.

 

[Eugene Hsieh]

I'd pay an extra US$20 for a seamless layer change as long as the rest of the player were good.

My RP91's layer change is slow compared to my Apex, but the Apex quality is poor.

I've never seen a layer change take longer than 2 seconds in any player, including first generation players (assuming they don't screw up on the layer change). Thus an additional 4 MB output buffer should be sufficient. I don't see what the big deal is. Sub-$100 CD burners these days have 2-8 MB buffers. Even if the data rate were 20 Mbps, that's only 40 Mbps over 2 seconds, or 5 MB of memory.

Why is it such a big deal to ask that a $500 DVD player to have 4-8 MB too? Considering that this is the market of well-educated buyer, a seamless layer change would be a significant marketing advantage.

I hate progressive scan combing, but I hate layer changes even more.

P.S. You guys can add Contact to the list. The layer change has no dialogue, but nonetheless the change is in a pretty obvious spot.