Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.compression > #203 > unrolled thread
| Started by | Phil Carmody <thefatphil_demunged@yahoo.co.uk> |
|---|---|
| First post | 2011-05-03 13:32 +0300 |
| Last post | 2011-05-12 20:15 +0200 |
| Articles | 20 on this page of 63 — 9 participants |
Back to article view | Back to comp.compression
removing predictable crap from poor quality jpegs Phil Carmody <thefatphil_demunged@yahoo.co.uk> - 2011-05-03 13:32 +0300
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-03 13:53 +0200
Re: removing predictable crap from poor quality jpegs Phil Carmody <thefatphil_demunged@yahoo.co.uk> - 2011-05-03 16:09 +0100
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-03 18:10 +0200
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-03 18:32 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-03 06:22 -0700
Re: removing predictable crap from poor quality jpegs Phil Carmody <thefatphil_demunged@yahoo.co.uk> - 2011-05-03 18:03 +0300
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-05 04:18 -0700
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-05 13:35 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-06 00:23 -0700
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-06 14:24 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-06 20:33 +0200
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-08 09:52 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-08 15:27 +0200
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-08 17:55 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-08 07:56 -0700
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-08 18:44 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-08 10:21 -0700
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-09 10:51 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-09 11:49 +0200
Re: removing predictable crap from poor quality jpegs human <noreply@elfavo.com> - 2011-05-09 12:06 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-09 15:29 +0200
Re: removing predictable crap from poor quality jpegs Guido Vollbeding <gv@uc.ag> - 2011-05-09 16:03 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-09 22:01 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-09 23:24 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-10 16:45 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-09 04:15 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-09 14:26 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-09 15:51 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-09 21:09 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-09 22:28 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-09 23:06 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-10 17:34 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-12 15:01 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-09 09:40 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-09 20:23 +0200
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-09 22:37 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-10 02:27 -0700
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-10 05:01 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-12 15:51 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-13 05:07 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-13 14:39 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-13 15:39 +0200
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-13 16:36 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-13 08:52 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-13 19:00 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-13 10:07 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-13 17:25 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-13 10:06 -0700
Re: removing predictable crap from poor quality jpegs Pete Fraser <pfraser@covad.net> - 2011-05-13 18:17 -0700
Re: removing predictable crap from poor quality jpegs Guido Vollbeding <guido@jpegclub.org> - 2011-05-14 09:16 +0200
Re: removing predictable crap from poor quality jpegs Guido Vollbeding <guido@jpegclub.org> - 2011-05-14 10:07 +0200
Re: removing predictable crap from poor quality jpegs Guido Vollbeding <guido@jpegclub.org> - 2011-05-14 19:52 +0200
Re: removing predictable crap from poor quality jpegs Guido Vollbeding <guido@jpegclub.org> - 2011-05-14 21:08 +0200
Re: removing predictable crap from poor quality jpegs Guido Vollbeding <guido@jpegclub.org> - 2011-05-16 08:43 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-16 05:15 -0700
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-16 17:00 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-16 13:02 -0700
Re: removing predictable crap from poor quality jpegs Thomas Richter <thor@math.tu-berlin.de> - 2011-05-17 09:43 +0200
Re: removing predictable crap from poor quality jpegs Sebastian <s.gesemann@gmail.com> - 2011-05-19 02:55 -0700
Re: removing predictable crap from poor quality jpegs Noob <root@127.0.0.1> - 2011-05-20 13:23 +0200
Re: removing predictable crap from poor quality jpegs Pete Fraser <pfraser@covad.net> - 2011-05-20 05:38 -0700
Re: removing predictable crap from poor quality jpegs human <noreplay@elfavo.com> - 2011-05-12 20:15 +0200
Page 3 of 4 — ← Prev page 1 2 [3] 4 Next page →
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-13 05:07 -0700 |
| Message-ID | <f2651828-5088-4e8a-82d1-7f2c2611c1ad@c1g2000yqe.googlegroups.com> |
| In reply to | #258 |
On 12 Mai, 15:51, Guido Vollbeding wrote: > Sebastian wrote: > > Don't make it sound like others don't have the mental > > capacity to understand what you have to say ... > Actually, your mental structure is too heavy, so that it completely > covers up your consciousness, which is finer and different domain. LOL. You equate education with closed-mindedness. Don't do that. I'd say, it's even the other way around. (Proper) Education broadens the mind and makes it more receptive for other things. We're not talking about memorizing facts. We're talking about learning critical thinking, thinking in terms of abstractions, trying to see the big picture / things from another perspective, ... > Even if I would tell you the discovered DCT properties, you would > not be able to grasp it with your rational mind, > I have done that, trying to explain the fundamental DCT properties > to various kinds of people, and the result is interesting: > The more intellectually educated (=conditioned) the person is, the > less capable to understand. > Look, at least the first fundamental DCT property is published. This interpretation builds on false assumptions. For example, you assume that you have something interesting to share (while talking to people like Thomas!!!). Maybe there is some interesting contet to find at your homepage. VollbSlides.pdf ain't it. It looks like you're just proud to have understood "spectral domain resampling" and feel the need to tell everybody. Well, that's nice and dandy but certainly not news to anyone who knows a thing or two about DSP. The blocking issues still remain and still escape you. > Ask Thomas Richter, he should know better, but he also ignores it, > [...] Maybe because there is nothing special about what you call "fundamental DCT properties" that would be news to him. Of course, I can only speak for myself. BTW, we already had the pleasure of "meeting" here about 8 years ago. If I recall correctly, it was about how to properly do the chroma resampling for JPEG encoding/decoding. You suggested to use 16x16 DCTs and to retain only 64 coefficients. I suggested to use an analysis/ synthesis pair of lowpass filters from a perfect reconstruction two- channel filterbank. Example: analysis, impulse response: [ 1 1 ]./2 synthesis, impulse response: [-1 1 8 8 1 -1]./8 While both approaches satisfy the "lossless roundtrip property" (that is, resizing by factor of two followed by resampling to half the resolution results in the original data) there are differences in terms of artefacts -- basically: blocking artefacts versus artefacts due to the filters' wide transition band. It is a trade-off that still seems to escape you. You're effectivly proposing (perhaps unknowingly) to chop the image into blocks and to use "perfect lowpass filters" (infinite sinc impulse response) independently on these blocks where samples beyond block boundaries are extrapolated by mirroring. I'm saying blocking is generally a bad idea and filters with short impulse responses are preferable -- especially for image processing. But it may take DSP goggles and some filter design experience to understand what I just said. Maybe it is enough to get you motivated to learn more about DSP. Cheers! SG
[toc] | [prev] | [next] | [standalone]
| From | human <noreplay@elfavo.com> |
|---|---|
| Date | 2011-05-13 14:39 +0200 |
| Message-ID | <4DCD267C.B1219CA7@elfavo.com> |
| In reply to | #262 |
Sebastian wrote: > > We're talking about learning critical > thinking, thinking in terms of abstractions, trying to see the big > picture / things from another perspective, ... As said, you can't understand me, because I'm talking about a state of no-thinking, no-mind, that is consciousness, unconditioned consciousness. Because your mind and your thoughts are always conditioned, as said, they cannot enlighten you. Consciousness is BEYOND the mind, BEYOND thinking, and that is the reality. From that perspective, thinking is just dreaming, you are in a dreaming state, you are sleeping, you are not awake, you are not in reality, because reality is always here and now. Find a Zen master, for example, to enlighten you about that. So far, you are totally clueless about that, sorry. > I'm saying blocking is generally a bad idea and filters > with short impulse responses are preferable -- especially > for image processing. You are not aware of the second fundamental DCT property. And, again, I don't speculate, you have no chance. For me this is reality, and it will be manifested for application later, similar as the unique features regarding first fundamental property have been manifested. If you were ready for real insight, I would try to enlighten you, but your statements show that you are not open for reality, you are only ready for intellectual speculation, and you can't get this from me, sorry. ciao human
[toc] | [prev] | [next] | [standalone]
| From | human <noreplay@elfavo.com> |
|---|---|
| Date | 2011-05-13 15:39 +0200 |
| Message-ID | <4DCD34AD.7F5FBC1A@elfavo.com> |
| In reply to | #262 |
I wrote: > > Find a Zen master, for example, to enlighten you about that. OK, I'll give you a hint. For a start watch this: http://www.youtube.com/watch?v=S562reOBgnM (Eckhart Tolle - The Simple Truth) from one of the greatest spiritual masters of our times. If only for a moment (OK, about an hour) you could retract your mind and intellectual reasoning, and just LISTEN to this man, then that might give you a glimpse about what I was talking about. This guy is really remarkable and easy to understand for beginners. There are good explanations, especially at the end about difference in "teaching". Listen carefully, and please don't argue! This is not something to argue about, just to listen to and understand. If you don't get it, let it be. ciao human
[toc] | [prev] | [next] | [standalone]
| From | human <noreplay@elfavo.com> |
|---|---|
| Date | 2011-05-13 16:36 +0200 |
| Message-ID | <4DCD4201.64AD084@elfavo.com> |
| In reply to | #262 |
Sebastian wrote: > > This interpretation builds on false assumptions. For example, you > assume that you have something interesting to share (while talking to > people like Thomas!!!). I do NOT assume that I have something interesting to share with you guys. What could I share with you? You are only interested in intellectual speculation, and I am NOT interested in intellectual speculation! I have real insights and I am responsible, and one can only be responsible if one is conscious, if one sees the reality, and if one doesn't speculate. Please understand that. > Maybe there is some interesting contet to find > at your homepage. VollbSlides.pdf ain't it. You will always find the wrong places, because your mind which wants to speculate prevents you from finding the answer. The answer is there, but YOU are not there (see Eckhart Tolle talk). Look here: http://jpegclub.org/temp/ http://jpegclub.org/temp/ITU-T-JPEG-Plus-Proposal_R3.doc But note, this is a programmatic paper which explains something, and it does not attempt to be a matter of an intellectual speculation, see above - I'm not an intellectual speculator, please note that! ciao human
[toc] | [prev] | [next] | [standalone]
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-13 08:52 -0700 |
| Message-ID | <0b38e630-9f23-4d2d-870a-cca968de2305@m40g2000vbt.googlegroups.com> |
| In reply to | #265 |
On 13 Mai, 16:36, Guido Vollbeding wrote: > [...] > I do NOT assume that I have something interesting to share [...] > What could I share with you? I'm not sure. You keep saying that you understand more about the DCT than others. And you mention this instead of giving a mathematical argument about the (in)validity of my filtering/blocking statement -- even after I explained the statement in detail and provided a sketch of a proof. > Look here: > http://jpegclub.org/temp/ > http://jpegclub.org/temp/ITU-T-JPEG-Plus-Proposal_R3.doc Unfortunately, I'm not finding any surprize or revelation in there. It's the same old story you advocated 7 or 8 years ago (if my memory serves me well). "Frequency-domain resampling" and an alternate transform coefficient scan. What else is new? Also, how is this supposed to invalidate my statement regarding filtering and blocking artefacts? Maybe you have to turn on rational thinking in addition to your consiousness. -- SG
[toc] | [prev] | [next] | [standalone]
| From | human <noreplay@elfavo.com> |
|---|---|
| Date | 2011-05-13 19:00 +0200 |
| Message-ID | <4DCD6398.1590BDFD@elfavo.com> |
| In reply to | #267 |
EOD End Of Discussion I was right: communication is useless. Sorry for waste. Goodby. ciao human
[toc] | [prev] | [next] | [standalone]
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-13 10:07 -0700 |
| Message-ID | <345b00c4-57ff-41d7-94f4-06eb6f4dc6ee@a26g2000vbo.googlegroups.com> |
| In reply to | #268 |
On 13 Mai, 19:00, Guido Vollbeding wrote: > EOD > End Of Discussion > > I was right: communication is useless. The way I see it: You didn't even try. -- SG
[toc] | [prev] | [next] | [standalone]
| From | human <noreplay@elfavo.com> |
|---|---|
| Date | 2011-05-13 17:25 +0200 |
| Message-ID | <4DCD4D4F.C4699498@elfavo.com> |
| In reply to | #262 |
> http://jpegclub.org/temp/ > http://jpegclub.org/temp/ITU-T-JPEG-Plus-Proposal_R3.doc And by the way, on that meeting I personally talked to Ms. Mitchell (etc.), who is editor of the original JPEG standard and "JPEG bible" book author, and I could see that they had no clue about the DCT. That's why half of the JPEG standard is obsolete. See, I know something, that's why I don't need to speculate. You don't know, that's why you speculate. ciao human
[toc] | [prev] | [next] | [standalone]
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-13 10:06 -0700 |
| Message-ID | <9759feb7-36d0-4e50-ab80-c643bb144583@l6g2000vbn.googlegroups.com> |
| In reply to | #266 |
On 13 Mai, 17:25, Guido Vollbeding wrote: > [...] > And by the way, on that meeting I personally talked to Ms. > Mitchell (etc.), who is editor of the original JPEG standard > and "JPEG bible" book author, and I could see that they had > no clue about the DCT. Interesting. That tells me something. Maybe your perception of reality doesn't correspond to reality. Maybe you're suffering from a superiority complex and have trouble regocnizing intelligence in others. But let's not speculate! :-p > See, I know something, that's why I don't need to speculate. > You don't know, that's why you speculate. I claim, explain, and prove. You speculate (about my knowledge). :-p ;-) -- SG
[toc] | [prev] | [next] | [standalone]
| From | Pete Fraser <pfraser@covad.net> |
|---|---|
| Date | 2011-05-13 18:17 -0700 |
| Message-ID | <iqkl6k$d5f$1@dont-email.me> |
| In reply to | #266 |
human wrote: >> http://jpegclub.org/temp/ >> http://jpegclub.org/temp/ITU-T-JPEG-Plus-Proposal_R3.doc > > And by the way, on that meeting I personally talked to Ms. > Mitchell (etc.), who is editor of the original JPEG standard > and "JPEG bible" book author, and I could see that they had > no clue about the DCT. Guido, you've been claiming for about a decade now that you have secret, deep knowledge of the DCT, and that none of us understands it. I've never seen an explanation of your knowledge. Perhaps you could start by pointing to the fundamental errors in Joan Mitchell's book. Then you could try to give us ignorant savages an inkling into the your magic hidden DCT properties. Do you have the Britanik, Yip and Rao book? Are they as clueless as we are? If so, why? > That's why half of the JPEG standard > is obsolete. Most standards become obsolete eventually. Parts of JPEG were / are useful -- other parts not so. The core stuff is not as good as it might be, but is quite adequate for many of its current uses, and has a lot of inertia behind it. Are you saying that, if only Joan Mitchell had understood the DCT as well as you do, that JPEG would have been much better nowadays. > See, I know something, that's why I don't need to speculate. > You don't know, that's why you speculate. Remove our need to speculate. Tell us. Pete
[toc] | [prev] | [next] | [standalone]
| From | Guido Vollbeding <guido@jpegclub.org> |
|---|---|
| Date | 2011-05-14 09:16 +0200 |
| Message-ID | <4DCE2C68.5FFFA3CA@jpegclub.org> |
| In reply to | #271 |
Pete The basic point is CONSCIOUSNESS, because the rational mind has too many limitations, which you can only recognize from that higher state. I have the requirement to stay in that state, so my only option is trying to raise you to the same level. I have done that, and you have seen the result. Perhaps there will be some time later when I will be able to work out more explanations which would satisfy the poor beings who do not yet have realized their consciousness. But now it is not, and I cannot leave my state, because I have important responsibilities. For those who are conscious, there is basic material available to figure out, and they can ask me personally in real life communication to get answers. It is difficult in this medium. For those who are prepared to gain consciousness, I have given a hint. There are people who's primary task it is to enlighten other people about that, these are the spiritual masters/gurus/teachers. I would perhaps also be able to do this, but as said I have other primary tasks, so I just point you to the appropriate sources. Here is another one from the same teacher, for those who understand German (if not you will find something else): http://www.youtube.com/watch?v=9PY4G-1NJ6E (Eckhart Tolle - Freiheit von Gedanken) As said, this process (coming out of your conditioned mind into higher level of consciousness) is essential for understanding, and that is all what I can do for you now, sorry. Regards Guido
[toc] | [prev] | [next] | [standalone]
| From | Guido Vollbeding <guido@jpegclub.org> |
|---|---|
| Date | 2011-05-14 10:07 +0200 |
| Message-ID | <4DCE383A.C26FDA1B@jpegclub.org> |
| In reply to | #271 |
Pete Fraser wrote: > > Are you saying that, if only Joan Mitchell had > understood the DCT as well as you do, that JPEG would > have been much better nowadays. Yes, of course. But "would have been" is speculation, and I don't like that. It IS much better now with JPEG 8 already, unique indeed, and even more potential to be realized, as said. Fortunately, it can all be done with organic evolution of the few good things from the initial JPEG standard, and that is my task... That initial JPEG standard may be considered just as a draft. JPEG 8 is much more mature. I pointed out the lossless example not by accident. It can be easily understood without much consciousness... (although some very ignorant people still have a problem...) The 1-point DCT is basically a no-op in practice, but fits nicely (seamlessly) in the general framework. We have that desired lossless mode now in the code, without any special casing (the same 1-point "DCT" is used for the 1/8 baseline downscale). Only for performance reasons it may be desired to adapt (one thing: calling a function for each image sample in this case is a bit inefficient). Regards Guido
[toc] | [prev] | [next] | [standalone]
| From | Guido Vollbeding <guido@jpegclub.org> |
|---|---|
| Date | 2011-05-14 19:52 +0200 |
| Message-ID | <4DCEC16F.54D7507@jpegclub.org> |
| In reply to | #271 |
Pete Fraser wrote: > > Do you have the Britanik, Yip and Rao book? > Are they as clueless as we are? > If so, why? Yes, I have looked into these books, and the authors are clueless, this is pure speculation without substance (essence). I am wondering whether there was any person who consciously introduced the DCT, or if it was found by someone by accident. In any case, this somebody is unknown. Fortunately, I carry the essence now in my consciousness (or at least close to it, maybe can go even deeper) and manifest it gradually. But this is not uncommon, it is a rather common phenomenon that the greatest sages made little effort to write down their knowledge. For example, the religious prophets: Jesus did not write down anything, the new testament is a collection of notes from his apostels. Mohammed dictated the Koran to a clerk. Krishna (the Hindu Messiah) was heard by Vyasa who wrote down the Bhagavad Gita (Holy Scripture of the Hindus). Other people: Nikola Tesla, inventor of the basic components of the electricity network, said he would go writing a book later at age 150, after finishing his work. Unfortunately, he died at 86. Laotse was forced to write the "Tao Te King" by frontier officers when he left China, so that something precious could be kept for Chinese government (it took him three days). Osho, the greatest sage and spiritual master of our age, only delivered speeches to his disciples, and it was recorded and transliterated in over 400 books. This has a reason: All those people lived in a state of higher consciousness, and this makes it difficult or obsolete to express the knowledge in written scriptures. I can understand that... That's why, like Nikola Tesla, I say I will work out a scripture "later"... I have more important things to do now. Eckhart Tolle wrote few books, but gives many talks. The essential knowledge is hard to express in a book, and few books exist which transmit essential knowledge. Most books are mere speculations or fictions. I can however recommend any book from Osho. For me, this opus is the culmination of knowledge. You can grab any book of him (as said, all transliterated speeches), and it will enlighten you, educate you, and entertain you at the same time. Nothing else comes close to that. Regards Guido
[toc] | [prev] | [next] | [standalone]
| From | Guido Vollbeding <guido@jpegclub.org> |
|---|---|
| Date | 2011-05-14 21:08 +0200 |
| Message-ID | <4DCED33C.37A2539F@jpegclub.org> |
| In reply to | #271 |
> Laotse was forced to write the "Tao Te King" by frontier officers > when he left China, so that something precious could be kept for > Chinese government (it took him three days). And the first sentence of the Tao Te King goes like this: The truth cannot be expressed by words, and anything expressed by words cannot be truth. This is significant. But words can be pointers to the truth, when expressed by sages. Regards Guido
[toc] | [prev] | [next] | [standalone]
| From | Guido Vollbeding <guido@jpegclub.org> |
|---|---|
| Date | 2011-05-16 08:43 +0200 |
| Message-ID | <4DD0C773.7D5E3428@jpegclub.org> |
| In reply to | #271 |
> I can however recommend any book from Osho. For me, this opus > is the culmination of knowledge. You can grab any book of him > (as said, all transliterated speeches), and it will enlighten > you, educate you, and entertain you at the same time. Nothing > else comes close to that. One book which is on-topic would be: Osho - Intuition: Knowing Beyond Logic German: Osho - Intuition: Einsichten jenseits des Verstandes Regards Guido
[toc] | [prev] | [next] | [standalone]
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-16 05:15 -0700 |
| Message-ID | <b57f6706-34fa-46cc-9321-a2f33dc15cce@l6g2000vbn.googlegroups.com> |
| In reply to | #271 |
On 14 Mai, 03:17, Pete Fraser wrote: > Guido Vollbeding wrote: > > http://jpegclub.org/temp/ > > http://jpegclub.org/temp/ITU-T-JPEG-Plus-Proposal_R3.doc > > > > And by the way, on that meeting I personally talked to Ms. > > Mitchell (etc.), who is editor of the original JPEG standard > > and "JPEG bible" book author, and I could see that they had > > no clue about the DCT. > > Guido, you've been claiming for about a decade now that > you have secret, deep knowledge of the DCT, and that > none of us understands it. I've never seen an explanation > of your knowledge. I guess, we've already seen explanations, or attempts thereof. I believe Guido expects someone who fully understands the DCT and its properties to be as enthusiastic as himself about these things. But others are not. Unfortunately, this expectation makes him conclude that others don't understand what he is talking about. The possibility that they do understand but are simply not impressed does not cross his mind. And since Guido came to this conclusion _and_ is well- intentioned, he feels the need to point us into "the right direction" (by telling us all that stuff about rational minds and consiousness). I saw him mentioning the existence of two "fundamental DCT properties". The first one is obviously about -- to use his words -- the "scalability" of the DCT. Since this "scalability" alledgedly corresponds to the "fundamental image compression principle", the DCT is supposed to be the "optimal" transform in image compression. What the second "fundamental property" is supposed to be, I don't know. But I would be surprized if it was any more "revealing" than the first one. > Perhaps you could start by pointing to the fundamental > errors in Joan Mitchell's book. Then you could try to > give us ignorant savages an inkling into the your magic > hidden DCT properties. > > Do you have the Britanik, Yip and Rao book? > Are they as clueless as we are? > If so, why? As for JPEG, if I recall correctly, it has not been specified how chroma subsampling/interpolation is supposed to be done. Since Guido is the guy that he is (someone obsessed with the DCT and its supposeldy optimal properties) he would probably tell you that chroma subsampling/interpolation should have been implemented by exploiting the "first fundamental property" of the DCT. At least I agree with him on two accounts: - subsampling/interpolation should have been specified in the standard - interpolation followed by subsampling should yield the original signal (JP2 overcame these issues) -- SG
[toc] | [prev] | [next] | [standalone]
| From | Thomas Richter <thor@math.tu-berlin.de> |
|---|---|
| Date | 2011-05-16 17:00 +0200 |
| Message-ID | <iqre6v$jcn$1@news.belwue.de> |
| In reply to | #279 |
Am 16.05.2011 14:15, schrieb Sebastian: > I guess, we've already seen explanations, or attempts thereof. I > believe Guido expects someone who fully understands the DCT and its > properties to be as enthusiastic as himself about these things. But > others are not. Unfortunately, this expectation makes him conclude > that others don't understand what he is talking about. The possibility > that they do understand but are simply not impressed does not cross > his mind. And since Guido came to this conclusion _and_ is well- > intentioned, he feels the need to point us into "the right > direction" (by telling us all that stuff about rational minds and > consiousness). > > I saw him mentioning the existence of two "fundamental DCT > properties". The first one is obviously about -- to use his words -- > the "scalability" of the DCT. Since this "scalability" alledgedly > corresponds to the "fundamental image compression principle", the DCT > is supposed to be the "optimal" transform in image compression. What > the second "fundamental property" is supposed to be, I don't know. But > I would be surprized if it was any more "revealing" than the first > one. I assure you that you're correct with the first one. My best bet on the second one is that the DCT diagonalizes band matrices. Or equivalently, that it is the KLT for sources where the correlation is position-independent. (Which, once I thought about this, is just nothing more that Fourier transformation of a convolution is the multiplication with the transformed series, trivial enough). Amusing enough, this proves only optimality for MSE (mean square error), not for "subjective quality", which is something considerably different, and requires knowledge of the visual bands of the HVS, which are *not* DCT bands. If you look from this perspective, the DCT is no longer "ideal". > As for JPEG, if I recall correctly, it has not been specified how > chroma subsampling/interpolation is supposed to be done. Since Guido > is the guy that he is (someone obsessed with the DCT and its > supposeldy optimal properties) he would probably tell you that chroma > subsampling/interpolation should have been implemented by exploiting > the "first fundamental property" of the DCT. > > At least I agree with him on two accounts: > - subsampling/interpolation should have been specified in the standard > - interpolation followed by subsampling should yield the original > signal > > (JP2 overcame these issues) I beg to differ. (-: First of all, JP2 *neither* specifies subsampling or upsampling. However, it provides a nice tool how to avoid subsampling in first place, or rather, get it done by the wavelet, which is more canonical and much more flexible. The reason why JPEG-1 did not specify subsampling is simply that it was designed for - or considered to be applied to - sources that *were* already subsampled. That is, JPEG followed (and still follows) a tool-design where JPEG is just one of the tools applied in a compression/image manipulation chain, and subsampling appears outside of the standard. That said, the natural resolution would have been to specify subsampling in JFIF (JPEG-4 now), not JPEG-1. Unfortunately, this ship sailed away a long time ago, and it is too late to enforce a specific procedure now. The same goes for JPEG-XR where the formulation is along the lines of: "While we don't standardize how upsampling is to be done, you really really want to do it the way it is spelled out in the standard". And the reason for this is again that you cannot fully control the devices or tool chain where the standard is applied to. If you feed your TV with YCbCr signals (ITU-709 actually), it is up to the TV how to do the upsampling, and who is JPEG to tell the TV what to do about the signals. So JP2 offered two mechanism: Re-locate subsampling outside of the standard by specifying non-trivial canvas subsampling factors, or pull it into the standard by letting the wavelet do the job, the second one being the suggested method when using it to compress digital computer images in RGB for viewing them on the computer screen. So that's basically the reason why things are as they are. As said, if you want a standard for digital images on a PC (which JPEG *isn't*, but *JFIF* is - maybe), then yes, you probably should specify it. But JPEG wasn't designed for this application alone. Greetings, Thomas
[toc] | [prev] | [next] | [standalone]
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-16 13:02 -0700 |
| Message-ID | <83b44c6c-6a81-418f-bdc5-e4d98e7c131d@c26g2000vbq.googlegroups.com> |
| In reply to | #280 |
On 16 Mai, 17:00, Thomas Richter wrote: > Am 16.05.2011 14:15, schrieb Sebastian: > > At least I agree with him on two accounts: > > - subsampling/interpolation should have been specified in the standard > > - interpolation followed by subsampling should yield the original > > signal > > (JP2 overcame these issues) > > I beg to differ. (-: First of all, JP2 *neither* specifies subsampling > or upsampling. I did not mean to imply anything else. Maybe "overcame" was not the right word. Perhaps "rendered moot" would have been more appropriate. :-) > [...] > So JP2 offered two mechanism: Re-locate subsampling outside of the > standard by specifying non-trivial canvas subsampling factors, I didn't know that. So, basically, this part is similar to what JPEG-1 does, right? > or pull > it into the standard by letting the wavelet do the job, the second one > being the suggested method when using it to compress digital computer > images in RGB for viewing them on the computer screen. This is the way I was thinking about while writing the previous post. -- SG
[toc] | [prev] | [next] | [standalone]
| From | Thomas Richter <thor@math.tu-berlin.de> |
|---|---|
| Date | 2011-05-17 09:43 +0200 |
| Message-ID | <iqt8us$73b$1@news.belwue.de> |
| In reply to | #281 |
Am 16.05.2011 22:02, schrieb Sebastian: > On 16 Mai, 17:00, Thomas Richter wrote: >> Am 16.05.2011 14:15, schrieb Sebastian: >>> At least I agree with him on two accounts: >>> - subsampling/interpolation should have been specified in the standard >>> - interpolation followed by subsampling should yield the original >>> signal >>> (JP2 overcame these issues) >> >> I beg to differ. (-: First of all, JP2 *neither* specifies subsampling >> or upsampling. > > I did not mean to imply anything else. Maybe "overcame" was not the > right word. Perhaps "rendered moot" would have been more > appropriate. :-) > >> [...] >> So JP2 offered two mechanism: Re-locate subsampling outside of the >> standard by specifying non-trivial canvas subsampling factors, > > I didn't know that. So, basically, this part is similar to what JPEG-1 > does, right? In a nutshell, yes. Except that it is specified in a more canonical way (the subsampling factors are given explicitly, and not in the rather clumbsy way JPEG-1 did it), and the sampling position can be specified as well (unlike JPEG, which had - IIRC - cosited samples). The filters from and to the subsampled domain are not specified, though. >> or pull >> it into the standard by letting the wavelet do the job, the second one >> being the suggested method when using it to compress digital computer >> images in RGB for viewing them on the computer screen. > > This is the way I was thinking about while writing the previous post. > > -- > SG
[toc] | [prev] | [next] | [standalone]
| From | Sebastian <s.gesemann@gmail.com> |
|---|---|
| Date | 2011-05-19 02:55 -0700 |
| Message-ID | <0b91fc85-2d3c-41c2-9dde-fc791b3d6e0e@j28g2000vbp.googlegroups.com> |
| In reply to | #280 |
On 16 Mai, 17:00, Thomas Richter wrote: > Am 16.05.2011 14:15, schrieb Sebastian: > > [...] > > I saw him mentioning the existence of two "fundamental DCT > > properties". The first one is obviously about -- to use his words -- > > the "scalability" of the DCT. Since this "scalability" alledgedly > > corresponds to the "fundamental image compression principle", the DCT > > is supposed to be the "optimal" transform in image compression. What > > the second "fundamental property" is supposed to be, I don't know. But > > I would be surprized if it was any more "revealing" than the first > > one. > > I assure you that you're correct with the first one. My best bet on the > second one is that the DCT diagonalizes band matrices. Or equivalently, > that it is the KLT for sources where the correlation is > position-independent. (Which, once I thought about this, is just nothing > more that Fourier transformation of a convolution is the multiplication > with the transformed series, trivial enough). Amusing enough, this > proves only optimality for MSE (mean square error), not for "subjective > quality", which is something considerably different, and requires > knowledge of the visual bands of the HVS, which are *not* DCT bands. If > you look from this perspective, the DCT is no longer "ideal". Even ignoring the HVS, I'd say it is not ideal since natural images tend not too look like blurry noise but are also made up out of flat areas and localized things like corners and edges. I would expect something like a hierarchical discrete Wavelet transform to lead to a set of coefficients with a higher sparsity in these cases (compared to large block DCTs). Cheers! Sebastian
[toc] | [prev] | [next] | [standalone]
Page 3 of 4 — ← Prev page 1 2 [3] 4 Next page →
Back to top | Article view | comp.compression
csiph-web