Industry Insiders Q&A Thread: only Questions to insiders please - Page 149  

What I'm wondering though, is if the strange behaviour around grain is the result of too high a Qp around the background areas where these film grain are (quite a few titles come to mind) or because of the film grain filtering. If it is filtering, then it doesn't work, or to be specific, it doesn't work when you eyeball a 50" plasma from 12 inches away. The film grain does not look natural anymore. I do agree that from more than 6ft away, this is very difficult to see.

After thinking about this, I'm now more inclined to guess that it is the filtering that is causing this strange unnatural grain. Consistency in film grain within a picture is not a bad thing, but what I find objectionable is that the grain looks unnatural, as if it is splotchy. (from 12 inches away)

This is all conjecture as I have no idea what the original YUV looks like.

I assume you mean that this perceptual lossless will be with a sample that the subject is intimately familiar with right? Because if not, then noone can know which is the original. In a sense, the perceptual lossless side of things for most homo sapiens doesn't have to be near optimal for either video or audio --- us consumers don't even have the original, so how can we tell whether a compressed presentation is truly perceptual lossless or not?

At best we can say that it is perceptually artifactless, and this is the extent of the consumers' experience unless they happen to have another version that is closer to the original.

My point is that all subjective tests for perceptual lossless as a bar to present audio and video makes sense from economics and efficiency, but from known tools today, it's bunk from presenting the true content itself because the bar itself is movable, and not based on any science, only by subjective tests of ears and eyes based on the compressionists judgment.

Because this bar can move, who's to say that there is any point in changing from 16bit to 24 bit sampling? Who can really tell between 96kHz and 48kHz? For all we know 16bit 48kHz PCM is perceptually lossless from the original 24bit 96kHz. So why bother doing this? Isn't it because the economic and efficiency measure in all of this is suspect and because we recognize that though we crossed the point of diminishing return in accuracy, there is STILL the need to be as close to the original as possible?

As for video, I have even bigger objection to terms like transparency to master or perceptually lossless, not just because of the reasons above, but simply that too many who work in this field and have the golden eyes can take their favorite stress samples and can still pick out the original from any 28Mbps PEAK VC-1 encode of their favorite test samples.

It might be transparent to me, but it is not transparent to them, so unless we agree that the limits in peaks and bandwidths are for economic reasons targetting 90% of the population is enough, then the term is misleading, because the general public now has the perception that the movies presented in the advanced codecs are like PKZIP or RAR when in reality they have fairly large Qp values and are nowhere near lossless.

Lossless is the optimal, economics drives us to strive for bars with qualifiers in front, but since these bars are still subjective, I would say they can't be used because they mislead.

Come to think of it, even the term "artifactless" is false, because while we are all familiar with MPEG2 artifacts (because it is very easy to pick out), AVC and VC-1 hides artifacts very well, and it will take a while for the industry working with the encoding parameters before we can pick them out easily.

After reading the subsequent posts... I think I'm tending to agree with Gary... "DP Approved".

For cases like PiP, where the avg rate drops to ~0 for extended periods, it's still the worst cae "aligngment of peaks" that would constrain something like being able to add a completely new track, right?

IOW: if at any point during the film, if all peaks align to allow only 300Kbps of headroom, then addaing another 448Kbps audio track just ain't gonna happen (unless that peak-shifting feature I've heard rumored actually exists...)

However, if the avg. headroom is ~3Mbps with statmux turned on, then that xtra bitrate can be used to tune the picture, add IME where previously there was room for none, etc..., correct?

amir/ben,

could the hd dvd players' interactive implementation permit/enable buying the new video content available on xbox marketplace in a streamed manner? i presume network bandwidth could be an issue if it is possible?

Paidgeek, I´m reading at davisdvd that Sony has now announced Spiderman 2: Extended Edition, can you comment on it now? And a question for all insiders: has any title shipped with branching or are they thinking about it? I think it´s a very important capability rarely used, and Blu-ray has the advantage with the bandwith over HD-dvd, so they should push it (as Hd-dvd should use more PiP, even though I don´t see the point of watching Tom Cruise´s face in a corner of a movie...)

Regarding PIP (which i feel "interactivity" is totally overrated and pointless in this format ware)...

QUESTION - Have you studio guys ever considered the usability of "interactivity" and to make "watching" the "extras" more useful by swapping PIP and having all content viewable in a simple, concise manner?

For example in King Kong it's tedius if you just watched a 3hr movie to find the "extras". who is going to sit 3 more hours waiting for an "icon" to popup. Ugh. it would have been more useful to watch all the "extras" running full screen with the movie part being discussed showing in the PIP window.

Scanning at 2K resolution has been the most popular and feasible.

4K 4096x3112 is slowly becoming viable as storage costs drop and processing and transport speeds escalate.

Bill Pope, ASC awhile back did a screening for director Sam Raimi, the editors and the producers (and some other people with normal homo sapien eyes) which included a series of 2K and 4K resolution blinded tests, and all preferred the 4K input/output. The perceived improvement in resolution is that obvious.

Beginning with 6K, that perception starts to drop off for some people.

I'm of the opinion that when a consumer format is "transparent to the master", it is the sign that it is time to improve the master. :)

pm, do you mean a screening of the actual 2k and 4k scans or encodes based off those scans?

I have a question for paidgeek, in regards to the region encoding, why does Sony appear to be putting that on almost every new release, I had heard that catalogue titles wouldn't have it, but Black Hawk Down was region encoded?
I'm in Australia and love my blu-ray and although Sony Australia has been very helpful in providing me with upcoming release dates, I do get a little frustrated when stuff like Marie Antionette is announced and I have to wait for an Australian release that probably won't have the special features of the US version.

Region encoding isn't for Piracy it's to control the consumer market, A pirate in China can probably quite easily get his hands on a region A player if he wants.
So please tell me if Sony will be region encoding all their upcoming titles?

However, the optimal 2K scan shouldn't have those issues.

Clearly, a 4K is a better source than a 2K that isn't optimally derived from the 4K.

I couldn't agree more. The first viewing example (2 picture heights or closer on a big screen, instead of 4 picture heights or more, as in the 46" LCD example) is my definition of what the objective of High Definition should be.



Make a very transparent picture at an extreme bit rate, please.

I concur, except to say, preserving the much finer grain of the original negative, not the added grain of any other film elements made from it, the worst of which is the projection print.

Yes! They can use the sharpness control or DNR to reduce it, while if the original grain or fine detail isn't put on the disc, it can not be brought back.


We finally get a medium that can reproduce film, and people don't want it!
Well I want it, and if anything, if we should be doing something to a film to make it look its best is to use the original negative and try to push it a little bit (bring the MTF up a little because surely the O-negative can withstand it, and compensate for the deficiencies of cross color in the dyes of the day), if we get a little more apparent grain, well its part of film, I prefer a sharp, resolved grain look with bite than a fuzzy gloss, or an artificially sterilized looking one. SD gives us enough fuzz. Any tool used to minimize grain should be used sparingly without sacrificing any of the detail, or the look.


I want to add to the welcomes for paidgeek, and thank Penton-Man for pulling him here. :)

If you can pick out which one is better, it's definately not perceptually lossless by any definition.

In order to convince a certain studio to do an all 4K DI on a certain major motion picture, a blind test was done showing three scenes………
a 2K scan with a 2K film out,
a 4K scan with a 2K film out,
and a 4K scan with a 4K film out.

Everybody choose the 4K scan and recognized that it was superior to the other samples.
It wasn’t even a close call.

No argument there against 4K (which I assume will be the next epoch of home theater). I'm only speaking of the value of 2K v. 4K when going to 1080p, and that it won't neccessarily make a difference (but there are cases where it would).

thanks for the further clarification but could you provide some more :D

i don't know what is meant by 2k/4k film out? TIA

paidgeek,

Thanks for your excellent post about TFE.

I have been very puzzled. I had predicted that SPHE would single-handedly win the format war for Sony by shipping a huge number of really stunning titles. I expected Lawrence, Spiderman, Navarone, Da Vinci, etc., all to have shipped by now.

Most studios are holding up their treasures but it is clearly in Sony corporate's interest to get their best titles out immediately.

Can you shed some light on this?

Thanks!

I think we are getting feedback based on released titles like BHD, Click, RV others with considerable grain or noise. We will have to handle on a case by case basis.

Difficult interlaced sources are working better with AVC or VC1. Concerts are a typical example. Look for those to be encoded from us with AVC at very high bitrates.

SPE will likely take a more conservative approach with interactivity than some of the other studios. We have been experimenting with Java code and are getting the results we want on most, but not all players. I think it will pay to wait a bit longer to see how the player firmware updates progress. What interactive feature is your highest priority?

easy call: picture-in-picture video commentary

There is no official word yet on any Spiderman releases (I just checked). Once it is on the schedule and in production, I will try to provide some details.

My information is that catalog titles will be coded "all region". This will likely vary by studio, but it is current SPE policy. New titles that have not been released theatrically in all markets will likely be coded to specific regions.

If it is not a production or technical question, I can't comment, sorry.

That is the first rational explanation and policy I have ever heard for region encoding. By placing "catalog" titles as all region, SPHE may set a trend.

If other studios do the same, then that would make me more likely to support the format. Thank you.

One question though, there is no stealthy way to encode region coding on a disc or is that is possible? Either it has it on release, or it doesn't have it at all?

Paidgeek, welcome to the forum! It was long due for someone from Sony to be here. Thanks a lot for taking time to answer our questions.

Can you explain the overall MPEG2 encoding process using the real-time encoder? For example: How many passes? How much time (in ballpark)? Is there something similar to segment re-encoding done for the difficult scenes? Etc.

BD is more sophisticated than DVD in how it approaches region coding. On DVD, the disc is identified for a particular region or set of regions and it is up to the player to stop playback if it does not find a match. With BD, the software decides what to do if the player does not have the expected region code value. It is possible in theory for a studio to allow the disc to playback even if a region mis-match exists using whatever criteria they want (for instance getting some information from a network connection). I don't know of any discs using an elaborate program like this, but it is possible.

Thanks to you and the forum for the hospitality and the chance to answer your questions.

The VM700 and VM770 are both Sony real-time, two pass encoders (The VM770 has a few additional GUI features and can encode all code values 0-254, aka super black and super white).

These encoders are similar in user operation to the original Sony SD encoders used for DVD to this day.

The basic operation is this:

A project is initialized

The title is loaded into whatever tape machine is required, usually HD-CAM SR

A first pass is run in real time, during this pass the encoder collects information about the entire program, 3/2 cadence if applicable, entropy (complexity), scene change locations and other parameters

The data collected by the first pass is software processed by the encoder in preparation for the second pass.

A second pass is run with the specifications that have been pre-determined for the project, principally the target file size.

The 2nd pass encode can be split screened in several ways against the master or viewed full screen according to operator preference.

The operator typically takes note of any scene that catches their attention for possible customization

The operator will adjust parameters, mainly bitrate for scenes that require it. The segment that contains the customized areas will be re-encoded.

Completed files are available by network for the software engineers to process.

For a typical two hour title, the whole process from beginning to end should be finished in about two days.


That's it in a nutshell..

That’s a heck of a of first post ! :)

Welcome to the AVS forum.
If memory serves, Noita is Finnish ?

I’m signing off for awhile.