Is There A Layer Change with DVD-Audio? (1 Viewer)

[Chuck C]

Is There A Layer Change with DVD-Audio? Is DVD-A recorded onto DVD-9 discs?

 

[Philip Hamm]

I've never experienced anything like a "layer change" with DVD-A but my experience is limited to just a few titles.

 

[Chris PC]

I know nothing about DVD-A but here is my educated (?) guess. Since a movie is one continuos stream of audio and video, there is often a layer-change that you see and hear because they couldn't fit the layer change in between scenes.

I imagine on a DVD-A disc, there are enough separate tracks that each track resides as one track on one layer. The tracks are split up among the layers, however they fit, but I doubt very much that any tracks are split between layers. Thats just my guess.

 

[John Kotches]

Chris, I think you need to read our article on DVD-A. Your "educated guess" wasn't correct. Here's a pointer showing the directory structures for a few DVD-A discs:
Link Removed
The correct answer to the question is that DVD-A allows for a seamless layer break, where DVD-V does not.
Regardless of whether we are talking DVD-A or DVD-V content is stored in files up to 1GB, with multiple tracks / chapters per file. The tracks/chapters on a DVD-A or DVD-V discs are not discrete files as is the case with Redbook CD.
Regards,

 

[Ian Montgomerie]

The reason you will often see a pause on layer change (which varies between players) is due to a combination of memory size and drive performance of the DVD player. When a layer change occurs, the disc drive has to refocus its layer on a new layer, and for some discs it must move the laser to the opposite side of the disc. During this time, it is not reading data off the disc. All DVD players will have a buffer of already-read data stored in memory, but typically this buffer will run out before the layer change finishes. The result is a pause while the drive catches up.

Most players simply don't include enough buffer memory to guarantee smooth layer changes, because memory is expensive and they can use smaller memory with the only negative effect being a layer change pause ("big deal").

Future players are including more and more buffer memory for features such as improved scan and reverse playback, so eventually the DVD-V layer change pause should disappear.

 

[Lewis Besze]

Actually the disc[dvd-v] is authored,that it requires the player to "unload" the buffer at the layer change.

Having said that there are differences among players but none of them are really seamless.

 

[Chuck C]

thanks for your observations philip and chris, and thanks for answering my original question, john

 

[John Kotches]

Ian,

This is true only to a certain extent, as there will always be a very slight pause at the forced layer break when defined as "non-seamless" per DVD spec.

So far, the best we've tested is ~ 1/4 of a second.

Regards,

 

[Philip Hamm]

Having said that there are differences among players but none of them are really seamless.

Actually, my ProScan PS8680Z was completely seamless 9 times out of 10. A really amazing machine at layer changes. Also, computer based players like many Apex players are seamless as well.

Also, I thought the Columbia "Superbit" branded releases were engineered to be seamless on most players.

 

[Matt DeVillier]

Also, I thought the Columbia "Superbit" branded releases were engineered to be seamless on most players.

they are desgined to be seamless, but they are also slightly out of spec in doing so, which causes problems in some older players

 

[Duke H]

Philip,

Proscan? Is that the same Proscan made by RCA? Their highend, "flagship" line? If so, how is the DVD player doing? I was under the impression that they discontinued the entire ProScan line after their TVs flopped (something like 80% of all units needed repaired or replaced!!!). Just going on vapors here, though. I've never checked it out, beyond what I've heard, and what I've read in CR. Thanks for any info.

 

[Philip Hamm]

Yes. It's a DIVX player. Got it cheap long ago. It's now doing double backup duty in my fiance's bedroom getting very light use.

 

[Ian Montgomerie]

This is true only to a certain extent, as there will always be a very slight pause at the forced layer break when defined as "non-seamless" per DVD spec.

So far, the best we've tested is ~ 1/4 of a second.

Well, you could call "time required to do a non-seamless transition" a third variable along with memory buffering and loader response time. There is no fundamental reason that a seamless transition has to be visible, and as an earlier poster mentioned, PC software players could do it within a field time. A non-seamless transition doesn't mean the player MUST flush all of its buffers, it just means it has to reset enough internal state (e.g. MPEG decoder state) when it needs to actually switch from one stream to the other. Whether this could be done imperceptibly fast depends on chip design.

However, the buffer design required for this is a pain, so there isn't much point to it unless the loader and memory are sufficient for rapid "non-seamless" transitions. The layer change is the only case where people get really annoyed by non-seamless transitions (this kind of transition also happens when you use certain fast forward/backward modes, but those tend to always have pauses because LOTS of loader and decoder muscle is required to make, say, 16X scan look smooth).

 

[John Kotches]

Ian,

I agree there are a variety of issues at hand, not the least of which is the head seek time / speed of transport. There's no question that DVD-ROM drives are faster than traditional transports as implemented in consumer players.

PCs generally have anywhere from 64MB (pretty low end these days) to 1GB or more. So we're not exactly comparing equals, are we?

Regards,

 

[Dan Stone]

Forgive me for asking, because I don't want to start any more dialog about which technology is better, but bottom line...with current technology (either HTPC based or standalone player) is there a player or players which have seamless layer changes?

Thanks for any info,

Dan

 

[John Kotches]

100% of the time? No.

Regards,

 

[Jeremy Hegna]

"...is there a player or players which have seamless layer changes?"
Dan,
The Denon DVD players, the 2800 specifically, has a buffering system...and I believe, John correct me if I'm wrong, the player that John and the rest of the guys at Secrets measured quarter second layer changes was this unit. The new Denon DVD players due out soon will also have this capability. The 2800MKII, the 3800, and the 9000. I have owned the 2800 for a year now and it's really tough to notice the layer change...a quarter second is pretty damn quick...as close to seamless I've seen with the exception of my PC.
I will never own a DVD player without this technology. It's mandatory for me after living with it this long.
Jeremy

 

[Norm Strong]

I own one of the infamous Apex players, and it has no layer change pause. I checked carefully.

 

[John Kotches]

Norm,

Put on WHQL, and run the layer change chapter. It is in the middle of a rolling sequence of credits, and is ideal for measuring layer change.

It's what we use for testing -- and all my numbers are derived from tests with that specific disc.

Regards,

 

[StaceyS]

The Apex A600 has about a 1/4" layer change on WHQL. (I have one of those.) WHQL is a worst case layer change. As Kotches wrote, it is in the middle of scrolling text. It is also not RSDL. (this type of layer change can be done quicker but it does not seem to be used often.)

The Meridian and Denon also measure less than 1/2" a second. The XBOX has the fastest layer change I have seen yet. Using the same PC with loader, one software decoder was faster than another during the layer change. The results were consistent test after test. This was a version or two ago on the software decoders tested.

It is hard to measure the layerchange less than 1/2" second using a stopwatch. I have some ideas on how to improve the accuracy.