Bay HD comments

[poothedrew]

It is good to see you have such an open mind to do research, get the facts & to make such a big leap...good for you!

I am also an IT geek, but don't do as much programming as I do with IT security, networking communications/applications, database admin, etc. However, regarding the Java thing, I honestly don't know enough about it, but I like the fact that it is more of a world standard than a Microsoft standard, not that active-x is bad or even HDi, but I think it is good to go with what the world is standardizing on...I would assume that there are good reasons why Java is the standard.
But regardless of Java being a standard, the BDA and others are working hard to make sure a great variety of stable/debugged BD Java templates are available for studios to use which will make it much easier to produce & program the extras with much less effort, and I think the BDA has many resources available (programmers) to assist studios in this regard, to make the BD Java less programmer intensive & easier/more stable overall.

My programmer buddies that write in Java a great deal, also use that common joke you mention above, however we've discussed the whole BD Java thing on numerous occasions, and they think the BD Java won't be that programmer intensive, and feel confident that the templates & resources made available to studios will be very stable & help studios to produce extras with relative ease. They are up on the whole Blu-ray/BD Java stuff more than I am. At first they questioned it with similar concerns as you mentioned above, but as they researched it more, they don't think it will be an issue. I have to agree that it won't be an issue, when you really dig into it bit more and look at what's available.

Thx!
I hope that is true and my only real concern in this area was for smaller film makers that may not have the resources to take advantage of the interactivity creating a have and have not culture.

[sphdle1]

Well, I agree on the fact that Java is a real pain to code for. I always prefer C++ to Java or C#/.NET. Virtual Machines based languages like those ones in theory should simplify things a lot by having automatic memory management and garbage collectors avoiding the burden of using pointers for everything but the truth is that they end up closing the programmer in a so controlled environment that doing anything beyond what the virtual machine pre-compiled classes allow to do becomes almost if not completely impossible to achieve.
However being BD-J a subset of Java allows to re-use a lot of available code to do many more things than with other more limited language. BD players like the PS3 surely won't have issues running micro Java applets at very high speeds in any condition. Cheap BD players and badly coded firmwares might have troubles with early disc releases.

I think you just explained the technical reasons behind my assumption that there must be good reasons why Java is the standard. Sounds like BD-J has great flexibility to do more things, and with the BDA helping to provide more templates & programmer assistence, studios should be able to do whatever they like for the extras, with lots of support & thus relative ease.

[sphdle1]

Thx!
I hope that is true and my only real concern in this area was for smaller film makers that may not have the resources to take advantage of the interactivity creating a have and have not culture.

The BDA are pretty good with this. I don't think they'll leave any studio out in the cold in this regard. They have many resources available & too much invested in Blu-ray as a technology to let that happen. From what I hear from BD insiders, they are a great group to work with, especially for all the studios!

Michael Bay seems eager to work with them!

[darinp2]

When you convert from progressive to interlaced you cause a corruption of the optical flow axis for every object on the screen, the signal bandwidth gets almost halved (and that's the reason why interlacing was invented in the first place, to save on costs!) with all its tradeoffs and issues.

Are you going to address why you are using the argument that there is less bandwidth when the conversation was 1080p24 sources and 1080i60 signal type?

Do you still think there isn't enough bandwidth in a 1080i60 signal for 1080p24 sources?

--Darin

[Tesseract]

[FONT='Calibri','sans-serif']dobyblue: I've seen you in the thick of all the HDDVD Bluray scuffles from here to Amazon and all the hidef message boards between. This is the first time I've seen you imply that you work in the industry and you did many pages back in this thread. I can't remember the exact post but you said something to the effect of "We released something encoded some way before other studio Y."[/font]

[FONT='Calibri','sans-serif']Who exactly do you work for?[/font]

[wingzero]

Are you going to address why you are using the argument that there is less bandwidth when the conversation was 1080p24 sources and 1080i60 signal type?

Do you still think there isn't enough bandwidth in a 1080i60 signal for 1080p24 sources?

--Darin

You fail to understand the meaning of bandwidth in regard to video coding, scan rate/frame rate conversion and deinterlacing techniques.

How much bandwidth a 1920x1080 interlaced 60Hz signals requires has nothing to do with the issue of extracting the right information to reconstruct the correct image. Interlacing corrupts the optical flow axis, you end up with signal information that has a very high added noise compared to the original source before interlacing is applied at any stage. The Philips document I linked you to before clearly explains that, if you can understand the terminology and how complex mathematical formula are used along with their meaning and real world applications. But that is just an example, there are many developers books explaining what interlacing is really about along with all the issues it creates.
If you had ever studied any developer book with a 64-Dimensions or higher analysis of an interlaced video signal compared to a progressive one and if you had the needed background to understand all that at least in part, you wouldn't keep repeating incorrect statements here and considering incorrect articles on websites and forums as the truth on the subject.

It's always better to have a progressive source than an interlaced one. In the literature there is quite some debating about that and even during the development of HDTV format there were some groups of developers of various multinationals that didn't agree on that, but when a standard has to be set there are compromises to be done to achieve the final goal to create a standard. So during the definition of HDTV the compromise was to increase the spatial resolution over the initially proposed resolutions trying to address the concerns of some stating that a lower resolution source even if progressive didn't look as good as an higher resolution one but interlaced. So yes, you can find some debating like that in the literature, however nowadays there is a certain level of agreement in the field that progressive is the way to go anyway and in any case. That's the reason why the next target for HDTV is the mass adoption of 1080p 60Hz.

[darinp2]

You fail to understand the meaning of bandwidth in regard to video coding, scan rate/frame rate conversion and deinterlacing techniques.

You claimed that signal bandwidth went down, so why not address whether 1080p24 or 1080i60 take more bandwidth, given that this is the case we are discussing? Very simply, does a 1080i60 have enough bandwidth for the information in a 1080p24 source? Why did you make the less bandwidth claim, if you don't think it is relevant?

The Philips document I linked you to before clearly explains that, if you can understand the terminology and how complex mathematical formula are used along with their meaning and real world applications.

I looked at the article at:

http://www.ics.ele.tue.nl/~dehaan/pdf/21_ProRisk.pdf

and it looks like it is all about interlaced signals where the fields are from different points in time (not what we are discussing here). In that case, true interlaced sources. In this case the source isn't an interlaced source (fields from different points in time), the source is progressive and it is just packaged within an interlaced signal. As I said before:

Motion compensation is for interlaced originals where the odd and even fields correspond to different points in time, not progressive originals which happen to be sent one field at a time, where each odd field has an even field for the same point in time and vis versa.

A 1080p24 source sent with a 1080i60 doesn't have odd fields with no corresponding even field from the same point in time or vis versa, like the interlaced stuff discussed there. In this case the goal is reconstruction, not interpolation, object detection, motion compensation, etc.

If line filtering had been added when going to 1080i60 then you would be right that the original could not be reconstructed, but people made sure long ago that the sources for these things on HD DVD and Blu-ray were not going to be linearly filtered (at least for Toshiba decks and things encoded with VC-1 by certain studios).

Once again, you brought up the bandwidth issue. Do you believe there is enough bandwidth in a 1080i60 signal to send the information for a 1080p24 source, or don't you?

How do you think the signal that a projector gets from a VP50 scaler outputting 1080p24 after combining fields from a 1080i60 signal (1080p24 source) differs from the signal the projector would get if the scaler wasn't there and a player that could output the 1080p24 original as 1080p24 was used? What do you believe the fields that the VP50 scaler gets in the 1080i60 signal represent?

Twice now you've avoided answering:

Do you know what 1080p24sf is and how the timing relates to 1080i48?

Would you argue that if a 1080p24 source is sent as 1080p24sf the progressive frames can't be reconstructed.

--Darin

[tosvus]

Seriously guys:

a) Do you think this forum is the place to discuss every little detail about interlacing, optical flow axis and other stuff? Let's all realise the main reason why Blu-Ray CAN be better is due to ability to have higher bandwidth, increasing quality by using less compression (if done specifically to take advantage of this).

b) For those slinging mud against Mr. Bay: Everyone (including him) is entitled to his own opinion, and your responses are rude. Secondly, you are visiting his site. If you were visiting someone, would you insult them to their face? Lastly, I have a hunch that Mr. Bay is more successful than all the rest of us at this forum put together, so some respect in that record is in order as well....

[wingzero]

You claimed that signal bandwidth went down, so why not address whether 1080p24 or 1080i60 take more bandwidth, given that this is the case we are discussing? Very simply, does a 1080i60 have enough bandwidth for the information in a 1080p24 source? Why did you make the less bandwidth claim, if you don't think it is relevant?

You are mixing things up in a desperate hope to prove you right. Interlacing almost halves the needed bandwidth to transport the signal, this a fact and not my claim or personal opinion. Comparing apples to oranges doesn't prove you right. You can't tell that a 120Hz encoded interlaced signal requires more bandwidth than a 30Hz encoded progressive one...(Well, actually yes, it does but it's not carrying more spatio-temporal information as it should and that's caused by interlacing along with its side effects--up to the point that an higher interlaced frame rate can look way worse to the human eye than a lower progessive frame rate) which is the same thing you are trying to do with a comparison between 1080i 60Hz and 1080p 24Hz. And anyway a 120Hz interlaced video signal doesn't need more bandwidth than a 60Hz progressive one at the same spatial resolution.
1080i 60Hz requires almost the same bandwidth as a 1080p 30Hz. It's quite obvious that a 30Hz signal needs more bandwidth than a 24Hz one if both are progressive, since the higher frame rate, there is more spatio-temporal information in the signal.

I looked at the article at:

http://www.ics.ele.tue.nl/~dehaan/pdf/21_ProRisk.pdf

and it looks like it is all about interlaced signals where the fields are from different points in time (not what we are discussing here). In that case, true interlaced sources. In this case the source isn't an interlaced source (fields from different points in time), the source is progressive and it is just packaged within an interlaced signal.

No! Your way of thinking is wrong. When you use interlacing and due to how interlacing works you have to consider the time axis. You can't assemble and discard frames and fields thinking that it's not affecting the signal. If you had studied the needed theory and you had the needed background you wouldn't be thinking in such a simplistic way. But since you don't have a clue about how a video signal really works and its relationships on the time axis and in the frequency domain, it's obvious that you can't understand the basics.
Whenever you use interlacing you have to take the time axis into account along with how the optical flow axis gets distorted.

As I said before:
A 1080p24 source sent with a 1080i60 doesn't have odd fields with no corresponding even field from the same point in time or vis versa, like the interlaced stuff discussed there. In this case the goal is reconstruction, not interpolation, object detection, motion compensation, etc.

You are wrong. Being from the same point in time means nothing. Do you really think that if that was that simple then multinationals would have spent and keep spending billion of dollars in R&D to develop always better algorithms to achieve better results ? If it was all said and done correctly with the ancient outdated telecine pulldown techniques as you keep repeating over and over here, do you really think that developers would have wasted millions of human work researching for better ways to solve the issues caused by interlacing ?

If line filtering had been added when going to 1080i60 then you would be right that the original could not be reconstructed, but people made sure long ago that the sources for these things on HD DVD and Blu-ray were not going to be linearly filtered (at least for Toshiba decks and things encoded with VC-1 by certain studios).

What ? You really don't have a clue of how much filtering is going on at the codec level, trust me. You need to use various filtering types both for encoding any source into digital format with any codec and for properly decoding it. Filtering occours both vertically and horizontally (spatial) as well as on the time axis and in the frequency domain. Nothing is as simple as you think, and designing filters correctly it's an hard task which needs experience, time and resources.

Once again, you brought up the bandwidth issue. Do you believe there is enough bandwidth in a 1080i60 signal to send the information for a 1080p24 source, or don't you?

I tried to explain how things work many times here now but you fail to even trying to understand that. You just want to have the final word and proved to be right. Sorry but that's not going to happen, I can't tell you that you are right when you are wrong and you are stating incorrect and wrong things as facts.

How do you think the signal that a projector gets from a VP50 scaler outputting 1080p24 after combining fields from a 1080i60 signal (1080p24 source) differs from the signal the projector would get if the scaler wasn't there and a player that could output the 1080p24 original as 1080p24 was used? What do you believe the fields that the VP50 scaler gets in the 1080i60 signal represent?

A simple basic scaler would work just in the spatial domain (and so it was many years ago), however modern scalers perform algorithms and filtering in the spatial, temporal and frequency domain and have complex motion estimation/motion compensation algorithms running on one or more DSPs along with all the other code.


This is the VP50 scaler you are referring to, right ?

------
http://www.ivojo.co.uk/dvdo-iscan-vp50.htm
http://www.anchorbaytech.com/products/systems/vp50.php

VRS Precision DeinterlacingT technology
480i/576i/1080i 10-bit Motion, Edge & Source-Adaptive Deinterlacer

• Five-field motion adaptive deinterlacing
• Edge adaptive processing for smooth diagonal edges
• Three frame video processing delay (Max)
• Game Modes with very low latency

• Sub-1 frame delay with edge adaptive processing
• 2 frame delay with edge and motion adaptive processing

------


Do you see that ? Did you read the specifications ? "Five-field motion adaptive deinterlacing" it means that it's using either field interpolation techniques on both spatial and temporal domains at least which is the basis for some motion compensation techniques other than the most used and fastest block matching one or it might be using some sort of hybrid telecine pulldown with added motion compensation algorithm. They might be using more advanced algorithms but the product it's not a professional one targeted to the broadcasting sector, it's a prosumer product so I don't think that they managed to implement very complex motion compensation algorithms but I might be wrong.

Twice now you've avoided answering:
Would you argue that if a 1080p24 source is sent as 1080p24sf the progressive frames can't be reconstructed.

--Darin

I know that websites and forums are claiming that a 24Hz interlaced signal reconstructed with telecine pulldown techniques and then sent at a doubled scan rate would be like the original 48Hz progressive signal BUT that is wrong. It doesn't work that way.
Just doubling the field rate or frame rate or a combination of both without any motion compensation would not be able to achieve optimal results. Those were techniques in use when motion compensation didn't exist or was too expensive to implement due to hardware costs. That is not the case anymore. Even when a 24Hz progressive signal gets multipled up to create its own multiples such as 48Hz,72Hz,96Hz,120Hz and such, motion compensation and more advanced techniques need to be used to achieve optimal results. Plain old outdated telecine pulldown technology is not enough, if used all alone and not even in an hybrid form with motion compensation of some kind being added it does even causes more issues to the video signal than it solves. It's a flawed by design technology and that's the reason why the industry spent billions over the last 25+ years developing better techniques.

[Tobi]

Correct me if I'm wrong, but VC-1 doesn't support 4:4:4 color space, right?

I wasn't addressing either codec in that respect, all I was impying is that both HD DVD and Blu-Ray use VC-1 and AVC/MPEG 4, so the issue of color space is not a pro or a con for either format, as a matter of fact the codec used for the Transformers HD DVD encode was AVC. I'm assuming that you mentioned this in regards of Bay liking Blu-Ray due to the color space issue of the codecs used, if I misunderstood you, disregard.

[wingzero]

I wasn't addressing either codec in that respect, all I was impying is that both HD DVD and Blu-Ray use VC-1 and AVC/MPEG 4, so the issue of color space is not a pro or a con for either format, as a matter of fact the codec used for the Transformers HD DVD encode was AVC. I'm assuming that you mentioned this in regards of Bay liking Blu-Ray due to the color space issue of the codecs used, if I misunderstood you, disregard.

However, given an higher available space on Blu-Ray it's possible to encode with the same codec at an higher bitrate which ensures less compression of colour space samples too.

[darinp2]

You are wrong. Being from the same point in time means nothing. Do you really think that if that was that simple then multinationals would have spent and keep spending billion of dollars in R&D to develop always better algorithms to achieve better results ?

They are spending that money for interlaced sources, not progressive sources which happen to be sent with interlaced signals.

What ? You really don't have a clue of how much filtering is going on at the codec level, trust me. You need to use various filtering types both for encoding any source into digital format with any codec and for properly decoding it.

There is filtering going on when compressing, but with VC-1 and 1080p24 sources they aren't doing the vertical filtering for interlaced because they are storing progressively on the discs and then flagging it as interlaced (even though it isn't). I know somebody who works on that stuff and has verified that they are not vertically filtering for interaced either when doing compression (at the ones they work with) or inside the Toshiba players when outputting 1080i. Again, there is filtering for the compression going on all around, but not for interlaced when the source is 1080p24.

This is the VP50 scaler you are referring to, right ?

------
http://www.ivojo.co.uk/dvdo-iscan-vp50.htm
http://www.anchorbaytech.com/products/systems/vp50.php

VRS Precision DeinterlacingT technology
480i/576i/1080i 10-bit Motion, Edge & Source-Adaptive Deinterlacer

• Five-field motion adaptive deinterlacing
• Edge adaptive processing for smooth diagonal edges
• Three frame video processing delay (Max)
• Game Modes with very low latency

• Sub-1 frame delay with edge adaptive processing
• 2 frame delay with edge and motion adaptive processing

------


Do you see that ? Did you read the specifications ? "Five-field motion adaptive deinterlacing" it means that it's using either field interpolation techniques on both spatial and temporal domains at least which is the basis for some motion compensation techniques other than the most used and fastest block matching one or it might be using some sort of hybrid telecine pulldown with added motion compensation algorithm. They might be using more advanced algorithms but the product it's not a professional one targeted to the broadcasting sector, it's a prosumer product so I don't think that they managed to implement very complex motion compensation algorithms but I might be wrong.

Yep, that is the scaler and it does do motion adaptive deinterlacing when set to that mode. I already told you about locking it into a mode for 1080p24 sources sent as 1080i60, but you didn't seem to comprehend and made some smart alec comment. The VP50 has a mode where it will lock onto the 3:2 cadence and not do any motion adaptive stuff, because what should be being sent is progressive source within an interlaced signal. Try sending it a real interlaced source after locking it for the 1080p24 source in a 1080i60 and of course things go wrong. It doesn't need some super advanced AI to lock onto the 3:2 cadence when it is told that is what it should expect. And it is possible to lock on the wrong fields in the cadence, but once it locks on correctly, as long as the input signal continues with that cadence there is no reason for it to ever lose that lock.

I know one person who has helped on that scaler. The motion estimation is not for the situation we are talking about here. It is for the much more complicated situation of true interlaced sources, where it makes sense to spend a lot of money trying to get better.

I know that websites and forums are claiming that a 24Hz interlaced signal reconstructed with telecine pulldown techniques and then sent at a doubled scan rate would be like the original 48Hz progressive signal BUT that is wrong.

Maybe you have a typo here, because it isn't the original 48Hz progressive signal. The original isn't 48Hz progressive, it is 24Hz progressive. And the interlaced signal can be 48Hz or 60Hz, with the 48Hz in fields being the same frame rate as 24Hz in frames.

Just doubling the field rate or frame rate or a combination of both without any motion compensation would not be able to achieve optimal results.

It isn't just doubling the field rate or frame rate, it is sending one field at a time at the same frames per second rate (meaning send fields at twice the rate whole frames were sent). At least for 1080i48. 1080i60 sends duplicate data since the bandwidth is higher than 1080p24.

Do you disagree with this wikipedia entry about 1080PsF:

http://en.wikipedia.org/wiki/1080PsF

"To allow us to see progressively scanned frames, each frame is segmented into two fields. Each field always relates to the same frame, and the two can be re-interlaced and de-interlaced without any loss of quality."

Do you know what 1080p24sf (or 1080p/24sf) is? I believe that Sony has been one of the biggest supporters of that, using it for professional stuff and I know their Sony Qualia projector will take 1080p24sf.

To be clear, is your position that if a 1080p24 source was sent with a 1080p24sf signal, the Sony Qualia could not reconstruct the original frames?

Those were techniques in use when motion compensation didn't exist or was too expensive to implement due to hardware costs. That is not the case anymore. Even when a 24Hz progressive signal gets multipled up to create its own multiples such as 48Hz,72Hz,96Hz,120Hz and such, motion compensation and more advanced techniques need to be used to achieve optimal results.

If by optimum you mean making fake frames, which has its advantages, but currently has some disadvantages for film also. Another way is to still display the original 1080p24, even if it was sent at 48Hz, 72Hz, etc., with duplicate data. Creating fake frames gives a different look, which some like and some don't like.

It's a flawed by design technology and that's the reason why the industry spent billions over the last 25+ years developing better techniques.

There is money being spent for fake frames (new frames created with some algorithm) to smooth out 24Hz material and also money for doing the difficult job of deinterlacing basically video sources (true interlaced sources). Those require very complicated algorithms kind of like calculating Pii does. But if the question is what 9 times 9 is, complicated algorithms aren't necessary. Much like reinterlacing progressive sources that have been interlaced without adding filtering.

If you believe that the 1080p24 sources on these discs are getting filtered when they are sent with 1080i60 signals, who do you think is doing that filtering? Please don't say the codec as its job was done long ago (so we can ignore codec filtering for things stored as 1080p24). Do you think every player adds vertical filtering if set to 1080i60 mode for things stored on the disc as 1080p24?

--Darin

[wingzero]

They are spending that money for interlaced sources, not progressive sources which happen to be sent with interlaced signals.

Wrong. There are no two way of interlacing. Either you have an interlaced CMOS/CCD sensor in a camera or the sensor is progressive and you apply interlacing during processing, you end up with similar results and you can't get back to the original progressive signal/source like nothing happened.
How many times have I to repeat the same thing over and over to you ?
Do you understand that your assumptions are just wrong or not ? I'd say you don't so far...

There is filtering going on when compressing, but with VC-1 and 1080p24 sources they aren't doing the vertical filtering for interlaced because they are storing progressively on the discs and then flagging it as interlaced (even though it isn't).

What ? What the heck are you talking about ? What is the need to flag a progressive source as interlaced for you ? You don't need to use any flag to turn a progressive video into interlaced, it can be done in realtime at the decoding stage for any progressive source.

I know somebody who works on that stuff and has verified that they are not vertically filtering for interaced either when doing compression (at the ones they work with) or inside the Toshiba players when outputting 1080i. Again, there is filtering for the compression going on all around, but not for interlaced when the source is 1080p24.

Then that friend of yours if he/she works in the field should get fired because his/her knowledge on the subject it's worse then yours to tell such a thing.
Interlacing occours at the decoding stage when you turn a progressive video into interlaced if needed and filtering is used to lower aliasing effects among other things.

Yep, that is the scaler and it does doing motion adaptive deinterlacing when set to that mode. I already told you about locking it into a mode for 1080p24 sources sent as 1080i60, but you didn't seem to comprehend and made some smart alec comment.

If that scaler has a legacy telecine pulldown mode then that means nothing. If motion compensation can be turned off thru menus by the end user then that is provided to work with legacy hardware, ancient projectors and to please those like you who think that telecine pulldown is better than motion compensation.

The VP50 has a mode where it will lock onto the 3:2 cadence and not do any motion adaptive stuff, because what should be being sent is progressive source within an interlaced signal.

You don't know what you are talking about. There is no "progressive source with an interlaced signal" thing. Stop mixing things up trying to create confusion to prove you right. You are wrong.

Try sending it a real interlaced source after locking it for the 1080p24 source in a 1080i60 and of course things go wrong. It doesn't need some super advanced AI to lock onto the 3:2 cadence when it is told that is what it should expect. And it is possible to lock on the wrong fields in the cadence, but once it locks on correctly, as long as the input signal continues with that cadence there is no reason for it to ever lose that lock.

Yeah, sure.. and someone spends £1,200+ on a product that works by chance.. sometimes it works and sometimes it doesn't... smart move to use ancient telecine pulldown and avoid motion compensation... So they wasted their time developing a 5-field motion adaptive algorithm with spatio-temporal interpolation and filtering... because you think that a legacy mode of operation provided in the product is better than all the added algorithms that make the product what it is good for.

I know one person who has helped on that scaler. The motion estimation is not for the situation we are talking about here. It is for the much more complicated situation of true interlaced sources, where it makes sense to spend a lot of money trying to get better.

Ok, now you will start telling me that you know everyone in the industry, maybe you know the CEO of every multinational group working the field and such, right ? Well, whatever, even if that was true either they told you false things for whatever reason or they never told you anything about how things work and get developed.
There is no "true interlaced sources". *Interlacing IS interlacing*. Period.
There are no multiple interlacing versions.

Maybe you have a typo here, because it isn't the original 48Hz progressive signal. The original isn't 48Hz progressive, it is 24Hz progressive. And the interlaced signal can be 48Hz or 60Hz, with the 48Hz in fields being the same frame rate as 24Hz in frames.
It isn't just doubling the field rate or frame rate, it is sending one field at a time at the same frames per second rate (meaning send fields at twice the rate whole frames were sent). At least for 1080i48. 1080i60 sends duplicate data since the bandwidth is higher than 1080p24.

You have seriously confused ideas. I won't repeat myself once again. Re-read what I wrote, it seems a real waste of time trying to explain things to you.

Do you disagree with this wikipedia entry about 1080PsF:

http://en.wikipedia.org/wiki/1080PsF

"To allow us to see progressively scanned frames, each frame is segmented into two fields. Each field always relates to the same frame, and the two can be re-interlaced and de-interlaced without any loss of quality."

Do you know what 1080p24sf (or 1080p/24sf) is? I believe that Sony has been one of the biggest supporters of that, using it for professional stuff and I know their Sony Qualia projector will take 1080p24sf.

1) Wikipedia is full of fake, incorrect and false info about anything. Never trust Wikipedia as a reliable source on anything. Everyone can write on Wikipedia whatever they want and no professionals in any field is checking nor double-checking it.
2) Segmenting a frame into fields is interlacing when the split image is sent even(odd) fields first and odd(even) after those. Deinterlacing with no loss of quality is not possible. Whoever wrote that stuff on Wikipedia is wrong and wasn't able to explain double scan-rate technology at all.

To be clear, is your position that if a 1080p24 source was sent with a 1080p24sf signal, the Sony Qualia could not reconstruct the original frames?

What are you, a lawyer ? If the signal gets interlaced you can't reconstruct the original progressive with whatever implementation used. Being done by Sony means nothing. You confuse marketing with technology. Marketing people always mix things up when promoting any technology even due to their own ignorance about anything (with rare exceptions).

If by optimum you mean making fake frames, which has its advantages, but currently has some disadvantages for film also. Another way is to still display the original 1080p24, even if it was sent at 48Hz, 72Hz, etc., with duplicate data. Creating fake frames gives a different look, which some like and some don't like.

Your precious old outdated telecine pulldown techniques creates more fake frames than what motion compensation techniques allow to achieve. You really don't get it, right ?

There is money being spent for fake frames (new frames created with some algorithm) to smooth out 24Hz material and also money for doing the difficult job of deinterlacing basically video sources (true interlaced sources). Those require very complicated algorithms kind of like calculating Pii does. But if the question is what 9 times 9 is, complicated algorithms aren't necessary. Much like reinterlacing progressive sources that have been interlaced without adding filtering.

You really have a lot of confusion in your head about this stuff. You don't know what you are talking about. You can't imagine how many silly things you wrote in the paragraph above alone, and not counting the rest of wrong things you said. How many times do I have to explain the same thing to you over and over ?

If you believe that the 1080p24 sources on these discs are getting filtered when they are sent with 1080i60 signals, who do you think is doing that filtering? Please don't say the codec as its job was done long ago (so we can ignore codec filtering for things stored as 1080p24). Do you think every player adds vertical filtering if set to 1080i60 mode for things stored on the disc as 1080p24?

--Darin

Filtering occours at the decoding stage of the supported codec format. It's the decoder that has filters added to properly reconstruct the signal. There are filters at different stages and there is not a unique way of performing these kind of tasks. Developers and manufacturers can use their own techniques as long as they comply with the supported format correctly.


24Psf info on a Sony website:

---
http://bssc.sel.sony.com/Professiona...nitorsfaq.html

The BVM-F24U removes frame flicker using a newly designed frame store system that will allow multiple vertical frequency rates to be displayed. While not changing any parameter in the picture other than multiplying the vertical rate, 24P and 24Psf signals are displayed at 2 and 3 times the vertical frequency, thus removing any frame flicker.

---


Segmentation of frames into fields just perform a double,triple or higher scan rate trying to increase horizontal/vertical frequency and reduce flickering on the human eye it's simply a way of doing that simple thing. It's nothing new, although the marketing guys are now promoting it as the new achievement in the industry in the digital world. Scan rate doubling has nothing to do with interlacing and it was used in the theatres with projectors many years ago already. 24Hz progressive films were turned into 48Hz to increase the overall perceived temporal resolution and reduce eye strain.

[darinp2]

What ? What the heck are you talking about ? What is the need to flag a progressive source as interlaced for you ? You don't need to use any flag to turn a progressive video into interlaced, it can be done in realtime at the decoding stage for any progressive source.

The HD DVD format requires interlaced flags even when the source is stored as 1080p24. Did you not know that?

Interlacing occours at the decoding stage when you turn a progressive video into interlaced if needed and filtering is used to lower aliasing effects among other things.

Filtering can be used when converting the 1080p24 for sending in a 1080i60 signal, but it isn't required. For instance, Toshiba could have decided to vertically filter when outputting 1080i60 for 1080p24 sources in order to try to avoid problems with displays which display interlaced, but long ago they were talked to and it was made sure that they wouldn't filter like that.

Filtering occours at the decoding stage of the supported codec format.

The source we are discussing is 1080p24. Let's say VC-1. It is decoded and still 1080p24. Now it can be output as 1080p24 or it can be interlaced and output as 1080i60. Again, who do you think is adding new filtering when it is output as 1080i60 that isn't there when output as 1080p24? Any filtering that went on before the decision to output as 1080p24 or 1080i60 is irrelevant, since it is the same in both cases (unless you want to argue that 1080p24 output has the same problem you are claiming here).

Developers and manufacturers can use their own techniques as long as they comply with the supported format correctly.

Exactly. And Toshiba chose not to add filtering for 1080i60 output that isn't there for 1080p24 output (at least for the HD-A1 and I believe that has continued for all their players, although the XA2 may differ).

If you are claiming that they are adding filtering in all their players, why are you claiming that? If you are claiming that the filtering is there before the stage that decides to output the 1080p24 source as 1080i60 instead of 1080p24, then that filtering applies to 1080p24 output as well as 1080i60 output.

Please answer this question that you ignored:

If I sent you a picture by sending the odd lines today and the even lines tomorrow, could you reconstruct the original?

If I filtered it so you didn't get all the information of course you couldn't (that would be true if I filtered the original image and sent you the filtered version too, even if I sent it as one frame), but if I sent you all the info except in 2 fields you could reconstruct the original, right? Is filtering after the decision point of whether to output 1080p24 or 1080i60 now the reason you claim that the original 1080p24 cannot be reconstructed? If so, why are you claiming the Toshiba players add filtering there?

--Darin

[wingzero]

Please answer this question that you ignored:
If I filtered it so you didn't get all the information of course you couldn't (that would be true if I filtered the original image and sent you the filtered version too, even if I sent it as one frame), but if I sent you all the info except in 2 fields you could reconstruct the original, right? Is filtering after the decision point of whether to output 1080p24 or 1080i60 now the reason you claim that the original 1080p24 cannot be reconstructed? If so, why are you claiming the Toshiba players add filtering there?

--Darin

If you discard even just 1 field you won't be able to reconstruct the original signal as it was, you need interpolation in order to approximate it as much as possible and you must take the time axis into account, interpolating on the spatial domain only is not enough.

If you think that Toshiba is not using filters you are absolutely wrong. Just like the other stuff you kept repeating.

It's a real waste of time trying to explain things to you. You don't want to listen, you keep repeating the same stuff over and over like a little kid, and what's worse is that you want to look smart and cool and act like you were a professional working in the field. Well, if anyone was paying you for doing this kind of stuff then they surely must have a lot of money to waste and expect nothing in return from their employees.

[darinp2]

If you discard even just 1 field you won't be able to reconstruct the original signal as it was ...

Obviously. So no field that isn't just a repeat in the cadence is discarded when this is done correctly. Despite your claims earlier about bandwidth, there is plenty of bandwidth in 1080i60 signals to send 1080p24 sources. Not that hard to understand, so I'm not even sure why you mentioned this. I hope you won't stoop to claiming that a field that is discarded because it is just a repeat and known to be a repeat would keep the original signal from being reconstructed.

If you think that Toshiba is not using filters you are absolutely wrong.

I think you are wrong, so why don't you tell us why you believe Toshiba is doing filtering for 1080i60 that they aren't doing for 1080p24. As I said, somebody talked to them long ago about this issue to make sure they wouldn't add filtering like that and screw things up when doing 1080p24 from 1080i60 signals (with 1080p24 source).

It's a real waste of time trying to explain things to you. You don't want to listen, you keep repeating the same stuff over and over like a little kid, and what's worse is that you want to look smart and cool and act like you were a professional working in the field. Well, if anyone was paying you for doing this kind of stuff then they surely must have a lot of money to waste and expect nothing in return from their employees.

Let's see. First you use the less bandwidth argument. You were wrong. Now you say that if any field is lost you can't reconstruct things, as if wasn't obvious that doing it correctly means not losing fields, just like doing 1080p24 with 1080p24 signal types means not losing frames. Now you tell us that Toshiba is filtering. Please, tell us what filtering you claim they are doing and why you claim they are doing it.

Let's go back to a basic here though. If no filtering was added for the 1080p24 to 1080i60 stage in the player, would you still argue that the original 1080p24 could not be reconstructed?

--Darin

[Dino]

I have read this very long thread in the hopes that Mr. Bay would elaborate as to why he feels that Blu Ray suits his movies better than HD DVD. I certainly respect his views, but was curious to know what technical reason
led him to this stance. Unfortunately, I just saw people posting their guesses on why, such as "more color space", "higher bandwidth", "higher bit rate", "more capacity".

While I hope Mr. Bay will revisit this thread and offer more background to his
preference for Blu Ray, I would like to address the potential technical advantages that Blu Ray supporters tout in regards to an alleged difference in picture quality over HD DVD.

Color Space--to my knowledge, ALL Blu Ray and HD DVD discs have been authored with 4:2:0 color space, regardless of codec or bit rate. Same as
all DVDs.

Higher Bit Rate--yes, Blu Ray has the potential for about 10Mbps more of raw video stream than HD DVD. This is important when encoding with MPEG2.
However, VC-1 and H.264(AVC) modern codecs are optimized for low bit
rates. They were designed to use HALF the bit rate of MPEG2 and yield
equal or better image quality. So a Blu Ray title encoded in MPEG2 at
the maximum video transfer rate of 40Mbps would be considered ample
for that codec. AVC and VC-1 only require 20Mbps for equal quality,
much lower than HD DVD's maximum video transfer rate of 30Mbps.

So, we know that color space and transfer rate using low bit rate optimized
codecs does not favor Blu Ray over HD DVD in image quality. Indeed, I haven't heard of one Blu Ray title that looks better than the same title on
HD DVD. Some would say that is because the title was encoded at a lower
bit rate for HD DVD's limits and used for the Blu Ray title, but I maintain that throwing more bit rate at a modern codec like VC-1 or AVC would not make a difference visually. If there are already no visible compression artifacts at 20 or 30Mbps, encoding a disc at 40Mbps won't yield any discernable quality advantage.

I continue to be curious about what it is that Mr. Bay sees in regards to an image quality difference between Blu Ray and HD DVD given the same title,
same transfer and same codec.

Dino

[darinp2]

Back to one of your claims:

2) Segmenting a frame into fields is interlacing when the split image is sent even(odd) fields first and odd(even) after those. Deinterlacing with no loss of quality is not possible.

I ask again:

If I sent you a picture by sending the odd lines today and the even lines tomorrow, could you reconstruct the original?

Please tell me that you would be smart enough to be able to reconstruct the original if I did that. And that if sent you the odd lines today, the same odd lines tomorrow, and then the even lines the day after tomorrow (telling you that this is what I would be doing), that you could discard one of the odd fields and still reconstruct the original. If you think you couldn't figure out how to reconstruct the original under those conditions, please explain why you think you couldn't.

--Darin

[The Drizzle]

I don't know who is feeding you these lies but these statements are completely false.

No matter what size the HD DVD disk gets 51gb or 30gb the bandwidth will not change. This is the biggest flaw with the HD DVD spec. It is limited to the 36mbs bit rate for the rest of its life or risk all current players being useless.

If you think that Blu-ray is not pushing its superior 48mbs peek bit rate you are mistaken. Several titles reach this limit.

Check out this link for average bit rates for Blu-ray movies. Some have average bit-rates over 40mbs
http://forum.blu-ray.com/showthread.php?t=3338

Yep, and VC-1 and AVC don't get a lot better after 25mbps or so. ALl of that extra bandwidth really doesn't get BD anything noticeable. The fact remains that at least at this moment the HD disc of Transformers is better off on HD DVD. When BD gets it's act together with features, it may be a different story.

[The Drizzle]

If I sent you a picture by sending the odd lines today and the even lines tomorrow, could you reconstruct the original?
--Darin

I believe that every HD player presently interlaces the images before either sending it to the display or reinterlacing it. The lower priced HD DVD players don't have the deinterlacer that all the other HD players have. And, i and p are really the same thing. All that is different is the way the information is displayed and how fast it refreshes. Interlaced video refreshes twice as fast, obviously. Oh and your question is ridiculous. In 99% of cases the interlacing is seemless no matter where it occurs. The video mixers will most likely prevent pure video from passing over the buss, though it would be cool if it could be done and we could bypass the features when we cant, something like they presently do with the audio in 1.3 players (except the PS3 which lacks the silicon).

[darinp2]

I believe that every HD player presently interlaces the images before either sending it to the display or reinterlacing it. The lower priced HD DVD players don't have the deinterlacer that all the other HD players have. And, i and p are really the same thing. All that is different is the way the information is displayed and how fast it refreshes. Interlaced video refreshes twice as fast, obviously. Oh and your question is ridiculous. In 99% of cases the interlacing is seemless no matter where it occurs. The video mixers will most likely prevent pure video from passing over the buss, though it would be cool if it could be done and we could bypass the features when we cant, something like they presently do with the audio in 1.3 players (except the PS3 which lacks the silicon).

How is my question ridiculous? wingzero has been claiming that the original 1080p24 can not be reconstructed if the signal type is 1080i60 (even if the source is 1080p24) and although he didn't seem to even know what 1080p24sf is, he searched around and found something, but still doesn't seem to understand why it was created, and he claims that things cannot be interlaced and then deinterlaced without loss. That is ridiculous. I am asking a simple question because he doesn't even seem to get that. If he understands that he could reconstruct that picture then he should understand that progressive things can be sent with interlaced signals and then the progressive frames can be reconstructed, as long as the information is there (not filtered out). And as I have said, Toshiba was talked to a long time ago about not filtering it out just because the output is going to be 1080i60.

The case we are discussing is 1080p24 sources sent with 1080i60 signals and then displayed on a progressive display, not an interlaced display.

And as far as what the Toshiba players do, the HD35 at the very least sends the 1080p24 out as 1080p24 correctly (so actually 1080p23.976, but I didn't want to go into that detail). The XA2 was doing it incorrectly because it was sending out 1080p24 exactly. I've heard that has been corrected to 1080p23.976. That player may interlace it first even for 1080p24 output, where with the HD35 it never goes through an interlaced stage for 1080p24 output, from what I have been told by somebody who has tested some of these. I think the HD20 and HD30 also do 1080p24 correctly, but I'm less sure about those.

--Darin

[The Drizzle]

How is my question ridiculous? wingzero has been claiming that the original 1080p24 can not be reconstructed if the signal type is 1080i60 (even if the source is 1080p24) and although he didn't seem to even know what 1080p24sf is, he searched around and found something, but still doesn't seem to understand why it was created, and he claims that things cannot be interlaced and then deinterlaced without loss. That is ridiculous. I am asking a simple question because he doesn't even seem to get that. If he understands that he could reconstruct that picture then he should understand that progressive things can be sent with interlaced signals and then the progressive frames can be reconstructed, as long as the information is there (not filtered out). And as I have said, Toshiba was talked to a long time ago about not filtering it out just because the output is going to be 1080i60.

The case we are discussing is 1080p24 sources sent with 1080i60 signals and then displayed on a progressive display, not an interlaced display.

And as far as what the Toshiba players do, the HD35 at the very least sends the 1080p24 out as 1080p24 correctly (so actually 1080p23.976, but I didn't want to go into that detail). The XA2 was doing it incorrectly because it was sending out 1080p24 exactly. I've heard that has been corrected to 1080p23.976. That player may interlace it first even for 1080p24 output, where with the HD35 it never goes through an interlaced stage for 1080p24 output, from what I have been told by somebody who has tested some of these. I think the HD20 and HD30 also do 1080p24 correctly, but I'm less sure about those.

--Darin

So, am I to think that you were just picking on the guy because you think you know more than he does? Kind of childish don't you think? You sound like you know what BD and HD DVD players do and the fact that they are very similar and have damn near the same parts. You also seem to understand that the present way video is dealt with in BD/HD players isn't perfect, but that very few people are going to be able to tell the difference. They are simply different ways of doing the exact same thing. I wish they had ironed it out and saved the consumer all of this crap, though. Peece.

[darinp2]

So, am I to think that you were just picking on the guy because you think you know more than he does? Kind of childish don't you think?

No, I am trying to keep misinformation from being spread. He's been calling me childish over and over even though he has been posting incorrect information, and now you go the "childish" direction too.

Honestly, I think it would be nice if wingzero learned something here. Especially since he still doesn't seem to comprehend the main reason for 1080p24sf. But with his attitude, I'm doubting he will learn anything and will keep stating that anybody who thinks a progressive source can be interlaced and then deinterlaced without loss is wrong. He seems to have learned what motion adaptive deinterlacing is, but has misapplied it, thinking that it is needed for something where it is not.

--Darin

[The Drizzle]

No, I am trying to keep misinformation from being spread. He's been calling me childish over and over even though he has been posting incorrect information, and now you go the "childish" direction too.

Honestly, I think it would be nice if wingzero learned something here. Especially since he still doesn't seem to comprehend the main reason for 1080p24sf. But with his attitude, I'm doubting he will learn anything and will keep stating that anybody who thinks a progressive source can be interlaced and then deinterlaced without loss is wrong. He seems to have learned what motion adaptive deinterlacing is, but has misapplied it, thinking that it is needed for something where it is not.

--Darin

I can understand that, but as I understand it all players are currently being forced to interlace at some point. I expect that whatever format(s) we are finally dealing with, the present players and discs will seem inferior to the players and discs we get in the next couple of years. Look at DVDs development. The first players and discs were horrible.

[darinp2]

I can understand that, but as I understand it all players are currently being forced to interlace at some point. I expect that whatever format(s) we are finally dealing with, the present players and discs will seem inferior to the players and discs we get in the next couple of years. Look at DVDs development. The first players and discs were horrible.

I think the discs will improve (hopefully masters improve for one), but the only Blu-ray player I've heard of as far as interlacing the 1080p24 source for 1080p24 output was that old Samsung (the BD-P1000). Not sure about the 1200, but I believe the 1400, the PS3, and other Blu-ray players do the 1080p24 without going through an interlaced stage inside the player. And on the HD DVD side I have been told that the HD35 does the 1080p24 without interlacing inside also, from somebody who is familiar with those things (and lets Toshiba know when they want something changed).

--Darin