Newsmeldung von slashCAM: FFmpeg braodcast for professional use - HOWTO of rudi - 20 Nov 2007 18:44:00 The (extremely practical) codec library transforms always more to the universal all-rounder. Unfortunately, the legal situation to this program so far from clear, because in some countries, so theoretically some software patents could hurt. In addition, of course, have any questions about any royalties, which discourages them from the programmers, FFmpeg as a binary version available. Nevertheless, now seems to be synonymous in FFmpeg braodcast area to make popular. The linked document shows how FFmpeg synonymous to en-and decode braodcast familiar format that can be used. Among others, these are SonyIMX (D10), DVCAM, DVCPRO, DVCPRO50 and even Avid DNxHD (!!)). We recommend: Article print and always next to the editing system are palpable.
This is an auto-generated entry
Antwort von Bernd E.:
Braodcast ... ... Is this the same as before the spelling reform Slashcam-"broadcast" was? ;-)
Gruß Bernd E.
Antwort von ronsei:
Hmm apparently the file was removed ...
Here you will find the text still seems (2.Post):
Antwort von rudi:
Were a small database problem when uploading.
The link should work now, but now left in the braodcast Slashcam-German, because otherwise the comments from the forums articles move (to be indicated on the title).
Complaints in writing, please s.die Slashcam-editing, keyword "bread-casting" ...
Many greetings
Rudi
Antwort von WoWu:
.... the FFmpeg codecs are a group of clusters, z.Tl. re-engineerter codecs, Donation to the base in the Internet have been asked.
Comparisons with professional codecs have such a poor image quality revealed that one of a codec can not really talk. Good codecs achieve a coding efficiency of approximately 75-80%. The codecs have compared it just to approx. 40% brought even fewer individual codecs.
Only the already freely available codecs have shown no differences (why synonymous, they are simply copied).
So if you absolutely need to fall back on should the lousy pictures will not complain.
Antwort von urmel:
Comparisons with professional codecs have such a poor image quality revealed that one of a codec can not really talk. Good codecs achieve a coding efficiency of approximately 75-80%. The codecs have compared it just to approx. 40% brought even fewer individual codecs. Sounds so superinteressantem number after work, I would be very interested to know what the coding efficiency "is accurate. Do you have a link as synonymous to your statements?
My experience with ffmpeg is not so bad, you have to do is know what you do with it. If you eg create h264 wants to get with the default values, not very good results. However, is because with the x264 in it, and if the correct parameters then feeds pretty good quality out. If you for example taking the time to read this you'll see that here with x264 MainConcept to the top buhlt.
http://compression.ru/video/codec_comparison/pdf/msu_mpeg_4_avc_h264_codec_comparison_2006_eng.pdf...and with ffmpeg hast du vollen Zugriff auf die x264-Funktionen. Wenn du wissen willst, wie man den etwas zähmt lies z.B. hier:
http://rob.opendot.cl/index.php/useful-stuff/ffmpeg-x264-encoding-guide/ I can not now be included on each codec say something, but the ingetrierte Lossless codec Logarith is also erste Sahne. More personally, I do not need (a codec to work and one for distribution).
And then of course always ffmpeg is unbeatable when you receive material that nobody else aufbekommt. Especially the "professional codecs. As is often professionally with proprietary confused what anyone uses. But enough geschwätzt ...
Rudi
Antwort von PowerMac:
Too much Gefrickel ...
Antwort von Andreas_Kiel:
(...) Z.Tl. re-engineerter codecs, Donation to the base in the Internet have been asked. (...) Is this a euphemism for copying, disassembled, renamed, re-compiled and brought under the Public?
: - /
Andreas
Antwort von WoWu:
... Jau Andreas ... it sat on the point ...
... as well as Power Macs.
Short, concise but in Black's.
@ All the codec issue is not interested in:
Stop reading NOW!
And later with rum-fold Chinese polemisieren.
@ Rudi
I sit straight on the subject s.einer preparing ... but how would you, yes Encodieralgorithmen are not part of the standard but only the MPEG bitstream. The implementation is the 100% quality of an algorithm. In MPEG4, the Grope already on the z.Tl. lousy implementations in MPEG2 or synonymous with DV were already responding and where it was possible to fix Floating Point Point exchanged, but synonymous others like implementations made for weaknesses.
I can only respond with PM, if you s.dem codec rumschrauben only need to get a result, although it meets the need of programming enthusiasts, but I do not synonymous. Just as the codec (especially DVCPro) but are worth it backwards and not forwards.
Data areas are as I said clumsily chosen by quantization in the coding part Entrophy insufficiently filled and AC coefficients are dumped, which leads to significant artifacts. Furthermore, no optimization of Videosegmante instead. And much more ... The problem is just that good algorithms are now really all patent-protected and the re-engineering often only to s.gewisse borders is gone. But that's just with today's standards, often no longer enough.
With respect to MPEG-4 is the base of the V-1, although an early draft of H.264 equivalent, but not the standard, then yes yet been developed over years and are now in various score of V-1 differs . So that extent for commercial application is not interesting and not synonymous next subject of investigation ...
Résumé ... who work intensively with codecs company wants is to be a playground ... Who but with good quality video would like to deliver, should be the best implementations of draw.
The PDF I am in peace times, but when you first fly over looks, depending on a paper, that the Russian colleagues have prepared ... My search for the test conditions were synonymous then promptly failed because it is with all due respect for their colleagues are always a bit careful. But as I climb a next time .. Perhaps I have overlooked something.
@ PM .. what are your travel plans? Gruss W.
Antwort von WoWu:
@ Rudi ...
... only to general amusement ... There is a Chinese synonymous HD Standard:
AVS 1.0 - The Chinese Next-Generation Video Coding Standard.
Dept. of Electronic Eng. , Tsinghua University.
Institute of Information and Communication Engineering, Zhejiang University.
NLPR, Institute of Automation, Chinese Academy of Sciences.
Department of Computer Science and Technology, Tsinghua Un
Beijing Media Works Co. Ltd, Beijing, China.
After all, respectable addresses ... But if you look at the content of times there seems ..... there is 100% MPEG-4 there.
So now is the time with the re-engineering. It is so synonymous firms, with the old Labor-operation system of DEC on the way have become rich.
Antwort von PowerMac:
@ WoWu:
The plans are still only ideas. But we shall certainly have to take Käffchen in South America. Do not be now. Liebe Grüße
Antwort von WoWu:
@ Rudi
... it has a few days, but I have looked at the pdf views and in fact lacks the basic parameters for comparison. If I already create a test, do I need for identical conditions. this is sometimes not easy, because different tools to describe various systems. Most take the tests are then forced to the lowest common denominator here.
I do not especially indicate which codec was used for H.264, Joined whether a model, a reference implementation ... or what? The quality of a procedure and is covered with the codec. Also, as already mentioned, the tools were used ... What intra-coding has been set? How many B Prames between P frames? To what extent was the movement estimate set? Which search area has been discontinued? Has long-range control code used? What kind of Quantisierungsparameter etc. etc.
Perhaps there is still a paper in which all that is specified. This gives the pdf, however, not her.
But only by an identical comparison can only be effectively values. While this is not the case, the test is not worth the paper on which it is printed.
As I've already written, with the Russian colleagues, I would be synonymous in this critical. I have a whole time in the "Academic" with very good colleagues together, but was always surprised how lax such issues have been seen ... The tests were usually result-oriented (mostly knew beforehand what will come out (should be).
Therefore, with all due respect for the reinvented codec ... this test is so worth nothing.
Are there more accurate material?
Antwort von Wotan:
Unfortunately, the legal situation to this program so far from clear, because in some countries, so theoretically some software patents could hurt. In addition, of course, have any questions about any royalties, which discourages them from the programmers, FFmpeg as a binary version available.
Tjo then retrieves MPEG LA times s.and verklickert where that its now the Manufacturer of the codec you yes because you have the thing compiling and therefore synonymous her s.jetzt the license fees for their use would pay. So do everything legally ...
Since then the answer is not something that is offset because not fit into the business because it's not really true Manufacturer seid. One would have with the producers tried to link, but it has not worked quite so.
In other words: we can do nothing against failed so the thing to use. The industry wants to simply abolish opensource ...
In this sense, does it broadcasts what keeps the stuff. Builds Upscale with FFDShow. However, no one can ever say something!
