Announcement

Collapse

Fandango at Home Forum Guidelines

Fandango at Home Forum Guidelines

The Fandango at Home Forums are designed to help viewers get the most out of their Fandango at Home experience. Here, Fandango at Home customers may post information, questions, ideas, etc. on the subject of Fandango at Home and Fandango at Home -related issues (home theater, entertainment, etc). Although the primary purpose of these forums is to help Fandango at Home customers with questions and/or problems with their Fandango at Home service, there are also off-topic areas available within the Fandango at Home Forums for users to chat with like-minded people, subject to the limitations below.

Please post all comments in English. When posting a comment in the Fandango at Home Forums, please conduct yourself in a respectful and civil manner. While we respect that you may feel strongly about an issue, please leave room for discussion.

Fandango at Home Forum Guidelines

The Fandango at Home Forums are designed to help viewers get the most out of their Fandango at Home experience. Here, Fandango at Home customers may post information, questions, ideas, etc. on the subject of Fandango at Home and Fandango at Home -related issues (home theater, entertainment, etc). Although the primary purpose of these forums is to help Fandango at Home customers with questions and/or problems with their Fandango at Home service, there are also off-topic areas available within the Fandango at Home Forums for users to chat with like-minded people, subject to the limitations below.

Please post all comments in English. When posting a comment in the Fandango at Home Forums, please conduct yourself in a respectful and civil manner. While we respect that you may feel strongly about an issue, please leave room for discussion.

Fandango at Home reserves the right to refrain from posting and/or to remove user comments, including comments that contain any of the following:

1. Obscenities, defamatory language, discriminatory language, or other language not suitable for a public forum
2. Email addresses, phone numbers, links to websites, physical addresses or other forms of contact information
3. "Spam" content, references to other products, advertisements, or other offers
4. Spiteful or inflammatory comments about other users or their comments
5. Comments that may potentially violate the DMCA or any other applicable laws
6. Comments that discuss ways to manipulate Fandango at Home products/services, including, but not limited to, reverse engineering, video extraction, and file conversion.

Additionally, please keep in mind that although Fandango at Home retains the right to monitor, edit, and/or remove posts within Fandango at Home Forums, it does not necessarily review every comment. Accordingly, specific questions about Fandango at Home products and services should be directed to Fandango at Home customer service representatives.

Terms of Use - User Comments, Feedback, Reviews, Submissions

For all reviews, comments, feedback, postcards, suggestions, ideas, and other submissions disclosed, submitted or offered to Fandango at Home, on or through this Site, by e-mail or telephone, or otherwise disclosed, submitted or offered in connection you use of this Site (collectively, the "Comments") you grant Fandango at Home a royalty-free, irrevocable, transferable right and license to use the Comments however Fandango at Home desires, including, without limitation, to copy, modify, delete in its entirety, adapt, publish, translate, create derivative works from and/or sell and /or distribute such Comments and/or incorporate such Comments into any form, medium or technology throughout the world.
Fandango at Home will be entitled to use, reproduce, disclose, modify, adapt, create derivative works from, publish, display and distribute any Comments you submit for any purpose whatsoever, without restriction and without compensating you in any way. Fandango at Home is and shall be under no obligation (1) to maintain any Comments in confidence; (2) to pay to users any compensation for any Comments; or (3) to respond to any user Comments. You agree that any Comments submitted by you to the Site will not violate the terms in this Terms of Use or any right of any third party, including without limitation, copyright, trademark, privacy or other personal or proprietary right(s), and will not cause injury to any person or entity. You further agree that no Comments submitted by you to this Site will be or contain libelous or otherwise unlawful, threatening, abusive or obscene material, or contain software viruses, political campaigning, commercial solicitation, chain letters, mass mailings or any form of "spam."

You grant Fandango at Home the right to use the name that you submit in connection with any Comments. You agree not to use a false email address, impersonate any person or entity, otherwise mislead as to the origin of any Comments you submit. You are, and shall remain, solely responsible for the content of any Comments you make and you agree to indemnify Fandango at Home for all claims resulting from any Comments you submit. Fandango at Home takes no responsibility and assumes no liability for any Comments submitted by you or any third-party reserves the right to refrain from posting and/or to remove user comments, including comments that contain any of the following:

1. Obscenities, defamatory language, discriminatory language, or other language not suitable for a public forum
2. Email addresses, phone numbers, links to websites, physical addresses or other forms of contact information
3. "Spam" content, references to other products, advertisements, or other offers
4. Spiteful or inflammatory comments about other users or their comments
5. Comments that may potentially violate the DMCA or any other applicable laws
6. Comments that discuss ways to manipulate Fandango at Home products/services, including, but not limited to, reverse engineering, video extraction, and file conversion.

Additionally, please keep in mind that although Fandango at Home retains the right to monitor, edit, and/or remove posts within Fandango at Home Forums, it does not necessarily review every comment. Accordingly, specific questions about Fandango at Home products and services should be directed to Fandango at Home customer service representatives.

Terms of Use - User Comments, Feedback, Reviews, Submissions

For all reviews, comments, feedback, postcards, suggestions, ideas, and other submissions disclosed, submitted or offered to Fandango at Home, on or through this Site, by e-mail or telephone, or otherwise disclosed, submitted or offered in connection you use of this Site (collectively, the "Comments") you grant Fandango at Home a royalty-free, irrevocable, transferable right and license to use the Comments however Fandango at Home desires, including, without limitation, to copy, modify, delete in its entirety, adapt, publish, translate, create derivative works from and/or sell and /or distribute such Comments and/or incorporate such Comments into any form, medium or technology throughout the world.
Fandango at Home will be entitled to use, reproduce, disclose, modify, adapt, create derivative works from, publish, display and distribute any Comments you submit for any purpose whatsoever, without restriction and without compensating you in any way. Fandango at Home is and shall be under no obligation (1) to maintain any Comments in confidence; (2) to pay to users any compensation for any Comments; or (3) to respond to any user Comments. You agree that any Comments submitted by you to the Site will not violate the terms in this Terms of Use or any right of any third party, including without limitation, copyright, trademark, privacy or other personal or proprietary right(s), and will not cause injury to any person or entity. You further agree that no Comments submitted by you to this Site will be or contain libelous or otherwise unlawful, threatening, abusive or obscene material, or contain software viruses, political campaigning, commercial solicitation, chain letters, mass mailings or any form of "spam."

You grant Fandango at Home the right to use the name that you submit in connection with any Comments. You agree not to use a false email address, impersonate any person or entity, otherwise mislead as to the origin of any Comments you submit. You are, and shall remain, solely responsible for the content of any Comments you make and you agree to indemnify Fandango at Home for all claims resulting from any Comments you submit. Fandango at Home takes no responsibility and assumes no liability for any Comments submitted by you or any third-party.
See more
See less

Compression Discussion

Collapse
X
Collapse
 
  • Page of 8
  • Filter
  • Time
  • Show
Clear All
new posts
Previous 1 3 4 5 6 7 8 template Next
    #76
    Re: User-requested enhancements & features?

    Originally posted by redwein View Post
    ]
    Good enough as proof?
    I'm holding judgment until further investigation.

    But there is a factor that I didn't really consider. That's the human eyeball.

    I think it's well documented that the brain can perceive and fill in information that is missing, i.e. a 24 fps movie.

    I don't have a reference for an example, but there are lot's of examples of phrases that are out there that demostrate you can trick your brain into thinking one thing but literally is not true from the actual text.

    So the question is, what would be a better metric, one that actually measures the difference in loss from the source and destination, or what is actually perceived by the observer??

    I'm someone that would like to see the former than the latter.

    Comment

      #77
      Re: User-requested enhancements & features?

      Originally posted by HeadHodge View Post
      I'm holding judgment until further investigation.
      That really doesn't make sense to me given the evidence I presented. Proof by contradiction is a valid way to prove something false.

      Comment

        #78
        Re: User-requested enhancements & features?

        Originally posted by redwein View Post
        That really doesn't make sense to me given the evidence I presented. Proof by contradiction is a valid way to prove something false.
        I reviewed your previous post to see if I missed something. I like to believe that I'm a pretty open minded person.

        But even after reading it again, I can only conclude that you proved that you can convert video into different formats and create different results from the original. So in an abstract way you have proved that it's possible.

        I think a better experiment would be to take two or more frames of video (possibly one with lot's of dark and one with high contrast) and compress then decompress them with H.264 (ideally with the same compression rate that VUDU uses) and compare the results.

        I would be happy to do it, if only I knew how.

        Regards

        Comment

          #79
          Re: User-requested enhancements & features?

          Originally posted by HeadHodge View Post
          I think a better experiment would be to take two or more frames of video (possibly one with lot's of dark and one with high contrast) and compress then decompress them with H.264 (ideally with the same compression rate that VUDU uses) and compare the results.
          That's essentially what I did (though not with H.264). The jpg is the compressed file and the tiffs were uncompressed. I used paint to do the compression/decompression. It wasn't going between 2 different compressed formats. That's why I think it is analogous.

          Also, if you were going to invent the perfect compression algorithm for video, would you really shoot for one with a constant data loss per frame. I wouldn't. I would design one that would lose more in places where it matters the least and lose less where it matters the most. That is almost good enough as a proof by itself since H.264 is pretty much state of the art.

          Comment

            #80
            Re: User-requested enhancements & features?

            Originally posted by redwein View Post
            That's essentially what I did (though not with H.264). The jpg is the compressed file and the tiffs were uncompressed. I used paint to do the compression/decompression. It wasn't going between 2 different compressed formats. That's why I think it is analogous.

            Also, if you were going to invent the perfect compression algorithm for video, would you really shoot for one with a constant data loss per frame. I wouldn't. I would design one that would lose more in places where it matters the least and lose less where it matters the most. That is almost good enough as a proof by itself since H.264 is pretty much state of the art.
            Well if I was going to design the perfect compression algo, it would be one that would take up only one trinary string to represent the total content without loss. Much like a strand of DNA is able to do.

            But beside the point, I was trying to point out (as it's been pointed out to me) that the human biology has an odd way of interpreting raw information. So H.264 may be "state of the art" if you want to call it that. But does it really lose more in places where it matters the least?? That might be a good idea if one has to compromise, but I'm not convinced that is what is happening. Like you said previously, it seems to simply be compressing based on the other bits in the neighborhood and not necessarily where it matters the least.

            I was also suggesting earlier that with VBR compression that some areas are more heavily compressed than others because of the surrounding bits, but that after the content is uncompressed the loss is consistent and not related to the rate of compression. I could be wrong about that, but that's my theory.

            I know what I'm trying to say, but I'm not sure if I'm doing of very good job of it. But I promise to keep trying or until I'm convinced I'm full of it!!

            Comment

              #81
              Re: User-requested enhancements & features?

              If you have a lossy algorithm and it detects and reacts differently to different kinds of image scenarios, it is not possible to have a result that has a constant rate of loss regardless of the image on the frame. If what you are proposing is true, then a frame with all of the same pixel values will be guaranteed to be needlessly corrupted, which is way worse than even jpg compression. That just doesn't make sense.

              Comment

                #82
                Re: User-requested enhancements & features?

                Originally posted by HeadHodge View Post
                Like you said previously, it seems to simply be compressing based on the other bits in the neighborhood and not necessarily where it matters the least.
                Those 2 things can be equivalent. It matters the least when a pixel is close enough to its neighbors to not be noticeable if its true value is lost and its rendered value is computed based on its neighbors.

                Comment

                  #83
                  Re: User-requested enhancements & features?

                  Originally posted by redwein View Post
                  Those 2 things can be equivalent. It matters the least when a pixel is close enough to its neighbors to not be noticeable if its true value is lost and its rendered value is computed based on its neighbors.
                  But as I pointed out in a previous post, the human brain seems to have an affinity to detect changes in dark colors where it turns out that compression is usually the heaviest. So that would explain why dark scenes are so noticeably compressed and subsequently lost.

                  (Don't know about you, but I'm actually having fun!!)

                  Comment

                    #84
                    Re: User-requested enhancements & features?

                    Originally posted by HeadHodge View Post
                    But as I pointed out in a previous post, the human brain seems to have an affinity to detect changes in dark colors where it turns out that compression is usually the heaviest. So that would explain why dark scenes are so noticeably compressed and subsequently lost.

                    (Don't know about you, but I'm actually having fun!!)
                    It is an interesting conversation. I don't buy the argument about dark colors being caused by perception. The same phenonomenon shows up in bright sky scenes. All it takes is to have a large expanse of similarly colored, low detail images.

                    Also, your quest for being able to analyze the image in a non-subjective fashion is what I did. I compressed and decompressed 2 images. One image lost nothing, the other lost a noticeable amount (either by examining the bits or just by looking at the picture). Those characteristics are going to be inherent in any lossy compression algorithm. The lossy algorithms exist because of the desire to compress video more than can be done with lossless algorithms. So you are always dealing with tradeoffs and compromises in that case.

                    Given that, anything but the most idiotic of algorithms (just eat every n'th pixel) will optimize for getting the smallest possible file size with the best possible image. That can not be accomplished with a constant rate of loss as it will force you to lose data where it isn't necessary or desirable.

                    Comment

                      #85
                      Re: User-requested enhancements & features?

                      Originally posted by redwein View Post
                      Given that, anything but the most idiotic of algorithms (just eat every n'th pixel) will optimize for getting the smallest possible file size with the best possible image. That can not be accomplished with a constant rate of loss as it will force you to lose data where it isn't necessary or desirable.
                      I'm not sure if your confusing the rate of compression with the rate of loss. So I'm not sure if I can agree with you or not.

                      I think if you took the content you proposed to NA9D earlier where each frame is white, you should be able to compress the whole content losslessly into a file size of about 16 bytes.

                      So I don't believe that smaller file sizes indicates that it must mean that it represents a greater loss. It should just mean that the compression is more effecient than others.

                      Given a 90 minute movie that is all white, I should be able to replay the whole movie with only a file that is only about 16 bytes in size.

                      Comment

                        #86
                        Re: User-requested enhancements & features?

                        Originally posted by HeadHodge View Post
                        Given a 90 minute movie that is all white, I should be able to replay the whole movie with only a file that is only about 16 bytes in size.
                        That's my point. If you compress and decompress a movie that is totally white frames, you will have very good compression and zero loss when you decompress it. If you compress and decompress a movie that actually has typical video content, you will get a larger file size but will end up having data loss when you decompress it. Within a normal movie, there will be scenes that compress very well and lose little or nothing and those that compress less well and lose data, and every combination of compression efficiency and data loss possible. Given that, the loss per frame will not be constant. It will depend on the content of that frame. My jpg/tiff experiment demonstrated just that.

                        Comment

                          #87
                          Re: User-requested enhancements & features?

                          Originally posted by redwein View Post
                          That's my point. If you compress and decompress a movie that is totally white frames, you will have very good compression and zero loss when you decompress it. If you compress and decompress a movie that actually has typical video content, you will get a larger file size but will end up having data loss when you decompress it. Within a normal movie, there will be scenes that compress very well and lose little or nothing and those that compress less well and lose data, and every combination of compression efficiency and data loss possible. Given that, the loss per frame will not be constant. It will depend on the content of that frame. My jpg/tiff experiment demonstrated just that.
                          I do think I'm getting your point now, but I'm still not sure I'm willing to agree with it yet. But since I've just finished watching Hell's Kitchen and almost ready to pass out now. I need to continue this tomorrow, after I get a chance to reload my guns.

                          Until then Ado. I had fun.

                          Comment

                            #88
                            Re: User-requested enhancements & features?

                            Originally posted by HeadHodge View Post
                            I don't have a reference for an example, but there are lot's of examples of phrases that are out there that demostrate you can trick your brain into thinking one thing but literally is not true from the actual text.
                            Tinhk aoubt how you are rdeanig tihs sntecne. The hmuan bairn olny nedes the frsit and lsat ltetres of a wrod to be in the crorcet pacle. You raed wrdos not lrettes.

                            OK, back to normal. Yes, your brain recognizes the words based on the first and last letters. The middle ones don't matter how they are arranged except for when you are in English class. The middle stuff is all translated by the brain.

                            By the way, I'm splitting this whole discussion into its own thread...

                            Comment

                              #89
                              Re: User-requested enhancements & features?

                              Originally posted by NA9D View Post
                              Tinhk aoubt how you are rdeanig tihs sntecne. The hmuan bairn olny nedes the frsit and lsat ltetres of a wrod to be in the crorcet pacle. You raed wrdos not lrettes.

                              OK, back to normal. Yes, your brain recognizes the words based on the first and last letters. The middle ones don't matter how they are arranged except for when you are in English class. The middle stuff is all translated by the brain.

                              By the way, I'm splitting this whole discussion into its own thread...
                              Sometimes I just can't sleep at night because my brain goes into overdrive wondering where my posts are going to end up the next day!! It's the thrill that just keeps my juices flowing.

                              Comment

                                #90
                                Re: Compression Discussion

                                You guys are scaring me. Is this forum going to turn into AVS now?

                                Comment

                                Previous 1 3 4 5 6 7 8 template Next
                                Working...
                                X