Quantcast KVCD Article Posted at DVD-SVCD-Forum.de - Need Translator! - digitalFAQ.com Forums [Archives]
06-10-2002, 01:51 AM
kwag kwag is offline
Free Member
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Hello everyone:

An article about Q matrixes and some information on KVCD templates has been posted by Andreas, one of the moderators in the dvd-svcd forums in Germany, and the person who created the Quantization Matrix that is currently used in the KVCD templates. ( I can't thank him enough )

The article is in German. So I am looking for someone who would be kind and translate the article to English, so that I can post it here.

Here is the link of the article:

Reply With Quote
Someday, 12:01 PM
admin's Avatar
Site Staff / Ad Manager
Join Date: Dec 2002
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
06-10-2002, 03:04 AM
slab slab is offline
Free Member
Join Date: Apr 2002
Location: Chicago
Posts: 43
Thanks: 0
Thanked 0 Times in 0 Posts
Kwag, This is from a web based on-line translator...hope it helps...


Hello forum

Reference taking on the principle article: "a film on a CD, principles", second contribution, and the newly arisen discussion (GOP 1-14-3-1) I approached myself once and reckoned new MPEG1 Streams with the new TMPEG version 2.54plus. Also the forum KVCD. net (homepage) was very helpful. Here one very prettily can pursue, in how far it can be possible to generate a film, that suits on a CD. Also the basic TMPEG Template of this page and the moderator kwag is performed here very much preparatory work in order to have a start extension. Unfortunately the head attention is directed on the NTSC pseudoAVIs and are some attitudes not so good so that for our case PAL the results in the first moment were not so very promising. With our prior knowledge, we found however a solution, that I hold thoroughly for usefully. The result is on each case better than a VHS video cassette especially since one can use the advantages of the SVCD/xVCD with its chapter functions.

I incorporated to the left into the contribution, with which one can load himself to each KVCD test once an example file down, in order to got a graphic impression of the solution the single variants. Quiet examples are in order to be able to judge the sharpness. Actionsequenzen become yet nachgereicht. All files are as a SVCD gemuxt, what means so much that these files with one burn program can be burned directly as a SVCD. First however two new matrices in TMPEG should be bound. All information for the attitude of TMPEG are to be found in this left (Matritzen for TMPEG2.54). The TMPEG. ini against the new exchange and already function also the new matrix.

To be able to generate around example files characteristic with the respectively correct attitudes must following sequence in TMPEG compelling kept become:

1. Under WOULD FILE, Open Project that project file load *. tpr
Load 2. The film in Tmpeg in video source
Load 3. The sound in Tmepg in Audio source
4. Under Load (in the lower right in Settings) that *. mcf would file load

Therewith all attitudes should be undertaken up to the attitudes that of each before newly make must:

1St Source range, length the to encodenden film
2Nd Clip frame in order to adjust the correct aspect ratio

If one is certain himself, that everything it can agree go off.

Tja, and there also the question stands how one wants to designate then such a CD with film. Everything what deviates by the standard becomes somehow with a "x" marked. Only a xSVCD can be somehow everything, of 720X576 VBR MPEG2 with 4500Kbps MPEG5.1 multichannel until there to 352X288 CBR MPEG1 750Kbps mono.
I perceive to keep it therefore better an arranged „standard“, could have among other things the advantage that such a CD is to be played back on the Standalone Player, especially since one better can assign data and facts. Receive why therefore not the label of the KVCD Board? Moreover there are several Intros on the KVCD start page for the CD, that one can bind as a preliminaries into the CD. Well yes, looks simply entirely witty. Deliver Gegenstimmen please in the forum in the subsequent Diskussionsthread.


We come to the CD. It names herself KVCD and contains characteristic, interesting specification. To receive nevertheless, a given solution, without testing detailed, hold mounted I in the quality to be expected with the connected risks (an entire film and then yet to 120min game film length) for not so especially since yes the KVCD-Template mistake had. Fundamentally for the KVCD, a MPEG1 Stream is used, that deviates however by the standard because it becomes as a MPEG1VBR encoded. An extremely long GOP is used the expanse, that is not so in this type standard. Also the sound has only 128 Kbps. We had then not consistently further pursued Tja, these fundamental considerations also already, only we that. It is very clearly a special matter, that is not to be utilized for the Hobbyfilmer, because the quality becomes only well, if one has clean, progressive video material (pseudoAVI). The solution lies exactly between the VCD and the SVCD. Therewith the picture format leans against the DVD standard and is the so-called ½ D1 format with 352X576 PAL. We gather that „standard value“ the KVCD:


MPEG1 VBR, in the means about 900 until 1000 Kbps videos
Solution ½ D1 with 352X576
Sound with 128 Kbps Audio and 44,1 KHz
GOP 1/15/3/1 open or closed GOP
special KVCD matrix

We go stepped stepped for on the single head points on in order better to be able to understand subsequently the tests.

MPEG1VBR with on the average 900 to 1000 Kbps videos
No new matter, but very interestingly. Of the SVCD, we know yes the MPEG2VBR. MPEG2 is thought however in the reason interpretation for higher data ratio and also interlaced encoding. If the data ratio become for the picture so low, arise with MPEG2 problems and here lets now well the MPEG1 use itself. MPEG1 needs already once less data advice of the reason here in order to produce a comparable picture. The picture in the direct comparison is most a little fuzzyer because we have however a slight solution and in addition yet these little data advice, is it nevertheless a very good solution.
The data advice again represents new problems. In order to get a feeling for the orders of magnitude, here a little list of that to be expected middle video data ratio if it concerns a 90min CD:

Specification for total data advice:
Data ratio in the means in Kbps + Audio 128 Kbps + 71 Kbps Muxingrate for SVCD

80min film = ß1375 Kbps
90min film = ß1200 Kbps
100min film = ß1050 Kbps
110min film = ß 950 Kbps
120min film = ß 875 Kbps
130min film = ß 800 Kbps

One sees quite clearly how slight the data ratio become. Technically now the problem arises that there are some DVD-Player, that correctly function from a lower data advice no longer. My Philips has considerable problems with the Synchronität and with Tonaussetzern under 600 Kbps videos.
In my attempts, I indicated 650 Kbps videos as lower data advice. It would be well if it would test some of you once, whether that can be confirmed in your Playern also. Then that so would remain on all cases in the Template (test with a quiet place in the film).

Solution½ D1, 352X576
We already saw also that. Our MPEG2 interlaced videos often so are generated as a SVCD. It sets gets to be sure ahead, that the Player the stretching into the width around the factor 2 correctly there to change the height without also simultaneously. The solution is therefore such a Zwischending of VCD and SVCD. Has the advantage, that in the height more picture information is than in the VCD and therewith the sharpness is higher than that of the VCD, through the stretching in the width around the factor 2 is however in the relation to the SVCD with the factor 1,5 the picture somewhat fuzzyer. A compromise therefore, that felled or also not. For SVCD interlaced films, this 352x576er solution is almost „standard“in order to got one half-way sehbare quality. Go why for the progressive pseudoAVI nevertheless not directly on the SVCD solution?

We look at ourselves the ratio from data advice to picture solution once more exactly. In the example, we take a 110 min film and the Tool FitCD of shh that decreases us the calculating. We want looked observe the specific data advice per Macroblock, bps/in order to get an aussagekräftiges, comparable result.

First we observe the standard-VCD. Here we have a CBR with 1150Kbps video and a solution of 352x288. in a Letterbox 2,35:1er film (320x160 motion area) come therefore 1037 bps/looked and a compression of 1:18,5 out. A respectable result. Because however the solution is very slight, often the VCD anamorph to the insert comes. Herewith then the result (320x224 motion area is looked) 740 bps/with a compression of 1:25,9.

In a standard-SVCD with 224 Kbps Audio and 55 min film on the 90 min CD (2017Kbps middle data advice video) comes for a letterbox 2,35:1er film round 700 bps/looked and a compression of 1:27,5 out. As an anamorphe SVCD it 439 bps are looked/and a 1:43,8er compression. In an Audiodatenrate of 192 or quite 160 Kbps raises itself the specific data advice by 1,5% and/or around 3,2%.

VCD letterbox 1037 bps/looked 320x160 = 51200 Pixel
VCD anamorph 740 bps/looked 320x224 = 71680 Pixel
SVCD letterbox 700 bps/looked 448x320=143360 Pixel
SVCD anamorph 439 bps/looked 448x432=193536 Pixel

It is to be recognized here clearly, in which areas one moves. In order to come now again on the KVCD, we will calculate the solution 352X576 with letterbox and anamorpher solution and that of all in a 110 min film on a 90 min CD.

In a KVCD with a 2,35:1er film and 320X320 motion area it 413 bps are looked are looked/with a compression of 1:46,5 and in an anamorphen KVCD it 302 bps/and 1:63,5er compression.

KVCD letterbox 413 bps/looked 320x320=102400 Pixel
KVCD anamorph 302 bps/looked 320x432=138240 Pixel

For a KVCD with SVCD solution would become it be:

KVCD_svcd letterbox 295 bps/448x320=143360 Pixel (SVCD) looked

One can recognize very well how far one moves after below. In order to obtain herewith generally yet good results, one must grasp proper into the trick box.
Between the KVCD anamorph and the KVCD after SVCD picture solution, it is not much, however of KVCD letterbox to KVCD with SVCD letterbox picture solution. So slight data ratio approach easily at DivX-data ratio and equalize already. Also here the picture quality is not berauschend, but the DivX-Codec better can compress. One can calculate himself wherewith one gets to do it.

Sound with 128 Kbps Audio and 44,1 KHz
In order to save data advice and to make available simultaneously data advice for the video, must decide one in the sound. For our intention are max. To utilize 160 Kbps in 44,1 KHz. It is better however to work with 128 Kbps with 44,1 KHz. Unfortunately here the air becomes very, very thinly. One certainly can imagine this if the original with 384 Kbps comes 48 KHz DD therefore. A Downsampling on dolby surround results to be sure, but it is available nevertheless yet as a stereo trace. A comparison to MP3 and these data advice says yet a few things. Such an Einschnürung of the MP2 sound has therefore sequences. The total Tonempfinden becomes m. E. flatter. So readily one can begin there with TMPEG nothing more, for the result is for my taste no longer sufficiently. Obviously here Tmepg must yet strongly nachlegen. Or, one uses another Tool what converts then the Sound.

The Tool DVD2AVI laid out itself as very good to generate around a downmix on dolby surround in 48 KHz. In DVD2AVI, the Tool WAVEFS44 is incorporated, which is described in its function herehttp://arbor.ee.ntu.edu.tw/~jackei/dvd2avi/doc/audio.html , (). For a 384 Kbps, 48 KHz => 128 Kbps, 44,1 KHz-intention I use then again Plugins in TMPEG. On the page of MB1, the inserting of the Proggis is describedhttp://members.s2u.org/hobby/mb1svcd/audio/audio1.htm very exactly (). The further compressing of the Tondatei goes then quite orderly. Therefore one gets a Soundfile that thoroughly yet can be listened to. Signal of codified and with a dolby sorround is that also just yet o. k. It goes also with separate solutions, however I not yet tried that. I prefer makes before rather that „in one all“ solution if the quality agrees.

GOP 1/15/3/1 open or closed GOP
Here we come to an interesting subject. The GOP, Group Of Picture, describes the picture construct of the film in the three picture reason type I, P and B-frame. He who wants to know more over these pictures, should look around in the network. Basic idea is here the possibility the single pictures differently to compress. An I-frame lets very badly compress is itself to compress in contrast to the B-frame that very strongly. The P-frame lies any where in the middle. Now kwag selected here a very, very large GOP of 1/15/3/1 (I / P / B / x frame). What sticks behind it?

We have here round 60 pictures in a GOP, and therefrom only an I-frame. Normally one would have to stand here 1/4(5)/2/1, would be in a closed GOP that 13 of pictures. For 60 pictures in sum, round 4 I-frames would come together, can be compressed that however only contingently. Therewith then again an increased data advice connected in order yet respectably to be able to indicate the picture. Moreover we have only 2 very well komprimierbare B-frames. In sum not so well can be condensed in order to save place and therewith also data advice. Through the gigantic GOP, one can exploit now however this advantage very well.

Unfortunately this procedure has however also its hooks. First, these large GOPs are not standard in conformity. Although we have one ½ D1-Auflösung with VBR MPEG1, this Stream cannot be bound into a DVD. The standard allows there only a GOP of 15. A strong compression of also losses in the picture to the sequence has the expanse. The picture will be missing unclear, details, colors become compressed bad, simply the entire palette of strongly pictures. Also comes a more or less visible pumping of the picture to the light. If one good fortune one not sees has it on the television. It is just so that in the MPEG compression an advantage becomes loads of another disadvantage erkauft.

Also this GOP with the slight data advice and the normal matrix functions no longer so good. Large Blockartefakte and „Moskitoschwärme“ around edges are the result. Therefore finally also the matrix must be adapted.

Moreover the scene detection is to be recommended to turn on. Because the GOP is very long, it can occur thoroughly once that there are several scenes changes in a GOP. Therefore it is better if there is for each new scene a good start picture in order to avoid mistake in the compression. Glosed GOP is sooner uninteresting in this length and has distinguished yielded in my attempts none in the picture.

To hold around the strong pumping and the Farbausfälle in boundary I am parallel at the trying as well as the GOP with the distribution 1/3/3/1 makes myself. Results lie could be outlined not yet before however here an alternative to minimize around the itemized disadvantages.

Special KVCD matrix
So, and there we would have arrived again in our „favorite subject“. How pretty it could be if one for everyone count became. But nö, that is not so. The tests started with the standard matrix of TMPEG (default) and the attitudes out of the KVCD Board because the people are light on enthused of the good quality. Based on the special requests, I from there have to optimize out of tried. Unfortunately appeared that for the different test solutions and CQ-requests thoroughly different matrices are required and the proposed attitudes not correspond to the picture, how one expects it. As an apology, I would like to count let the NTSC. NTSC has a slighter solution and a picture fewer to encoden in order to bring it on the CD.

My Testviewer is the software Player of ATI because I have an ATI GraKa.
It is very honest, and much more important without Bildverbesserer spielbar.
Therefore one can see also fine differences in the picture, that disappear on the television or are covered by other Softplayern. Therefore the Referenzierung results first at the PC and later at the television. First if the PC-picture corresponds to my presentations, that comes compare to the television.

The matrices that turned out have, can divide one in three categories: high, middle and slight solution. Also the total system with the connection of the encoder method (CQ, 2pass) must be observed. Because I take at present CQ, also here the CQ-value for carrying comes.

The tests

I began with the highly possible solutions. How further above carried out, that the KVCD with SVCD-solution, Letterbox and an anamorphe KVCD are. Were the test sequences first once fine structures (frames 128000 to 129500 out of? The almost and furious?, then Action (frames 129000 to 143000 see previously), then water (capes 15 out of? the Thomas Crown affair) and to the end KAP36 of Blade 1. It became with the attitude CQ of 70 to 60 in 2er steps and the values

Max. 2000, min 650, P and B frame with 0, Audio 128 Kbps and 44,1 KHz, PAL

Tested. For characteristic comparisons would be here on the Template out of the KVCD Board referred. Here each sequence self once one in the comparison can calculate. The examples, that one can load further below down, are all than a SVCD gemuxt. One needs to take it only and to burn for example with the VCDeasy as a SVCD. Insert, play back and compare then in the Player. Or one looks at himself such files directly at the PC and assesses self once the differences between picture at the PC and TV.

1St test, KVCD with 448X320 film picture, KVCD_svcd

The KVCD_svcd is therefore interesting because it concerns in sum the picture with most information. Otherwise the solutions become only more slightly and therewith also theoretically more film suits on the CD. Here I tried to go successiv of CQ70 on CQ60 down. Unfortunately only with moderate success. In a motion area of 448X320 is in CQ65 end with jolly. I hold the picture mistakes, that are visible then at the PC screen, for too largely. At the television is in ca. CQ62 end. Nevertheless I put then the lower boundary for the picture on CQ65. That distinguished is in the file size m. E. too slight, the picture quality however much worse. Therefore I am somewhat certain also at the PC. In good Softplayern (for example sasami 2k, freeware), one can eliminate let the mistakes and is the picture then thoroughly also at the PC good.

The matrix in addition

The KVCD_448x320_CQ65 matrix is altered to this extent when that I took the Defaultmatrix of TMPEG and changed only the non-intra block. Soften looked is noise z. Zt. turned off. Out a matrix that I enclosed the Stream came exactly so like the Template. The Verblockungen not to are got at the PC very away. Only if the soften noise in the non-intra looked looked raised becomes at 30 to 50, is it somewhat better, but the movements become yet fuzzyer. Without the soften, noise looked is to be taken at the Fernsehschirm nevertheless a verb curl scarcely more aware.

KVCD_448x320_CQ65, download testMPEG,. trp and. mcf

2Nd test, KVCD with320X432 film picture, KVCD_anamorph

As second, I calculated, the „standards“ because of, the KVCD_anamorph. The picture is stretched in anamorpher technology (like yes everyone know left Pixelbeitrag) around the factor 1,3333 in the height. The Pixelunterschied to the KVCD_svcd amounts to round -5%.
The result is astonishing. Because the picture is pulled in the playing back in addition around the factor 2 into the width, Kantenrauschen comes and looked noising substantially more strongly to the validity than in the KVCD_svcd variant. Because to the end the picture is compressed, in order to play back it verzerrungsfrei, has one to be sure again the sharpness of the KVCD_svcd, but mistakes become through it more strongly visible because the lines interval is diminished.

The matrix in addition

It was compressed emerge now necessary to find a new matrix out, that the unpleasant Moskitoschwärme around an object around, like it in strongly films, reduces. Out the KVCD_320x432_CQ65 matrix which constructs on the preceding came. Simultaneously it should not bring in too much Unschärfe, for less Moskitos means z. Zt. so well like always also a fuzzyer picture. In the attempts appeared, changed will must that now also the intra matrix.

KVCD_320x432_CQ65, download testMPEG,. trp and. mcf

3Rd test, KVCD with320X320 film picture, KVCD_standard

Now it becomes exciting. We come now to the KVCD_standard out of the KVCD. net forum. With this it possible should become is better an entire film (PAL) with to 130min game duration in good quality, that than that of the VCD and the video VHS-cassette to got on a CD. The solution is the normal ½ D1, therefore 352x576, with a motion area of 320x320.
One can erahnen how the picture behaves. In the height, it remains because the format is not stretched like the VCD around the factor 2. Therefore therefore the solution in the height is twice as greatly. However becomes also here into the width concerned the factor 2, like in the VCD. We had that already in the KVCD_anamorph. Picture mistakes make itself however not very so strongly noticeably because the Stauchung stays away vertical.
If we look at ourselves the Bitrate per Bildpixel, we get about 1/3 of more data advice in same Bitrate in contrast to the KVCD_anamorph per Pixel. If therefore that the physical is distinguished, should to reduce therefore also the data advice around 1/3 be or climb the picture quality (less Verblockungen). So much to the theory.

In the practice, another picture appears. The base was also here the Standardtemplate out of the KVCD forum. With this reference (CQ65 calculated), a picture, that I perceived as not so crazy, arose. The pictures calculated previously were better. Now the attempts began, the picture rauschfreier and without producing Moskitoschwärme around the objects. Fehlanzeige. All attempts yielded a worse or equivalent picture. In spite of the insert of the filter noise reduction, I was able to assess scarcely visible improvement of the picture only that now the computation time increased around the factor 4. Therefore also that was no alternative. The Moskitoschwärme to not to confuse with a Verblockung in Actionszenen, remain. Obviously the matrix used at present is the goal aim ends. Not to my actual contentment.

The matrix

Attempts with a higher data advice had a better picture as a consequence, however also the file size increases a few. Finally I have looked taken TMPEGs default matrix with the soften noise von40/20 and CQ on 68 hochgesetzt. From this attitude, I was able to say that the picture is right good. In the relation to the KVCD forum, the quality is a better because CQ is higher, therefore the Quantisierung and the LOWER data advice, a little. Does also do what did here? The soften noise looked its contribution to the picture. In order to got generally a comparison, I have the standard-KVCD in this download out of the KVCD forum beige place.

KVCD_320x320_CQ68, download testMPEG,. trp and. mcf

4Th test, KVCD with320X224 film picture, KVCD_vcd_anamorph

We come to the long-known VCD-variant, but with increased solution therefore to the KVCD_VCD_anamorph. Does this have a solution of 352X288 with a motion area of 320x224. why actually anamorph? Now, the VCD has a so slight solution that the picture becomes already condemned fuzzy. A pure VCD letterbox has one something quality better than a VHS video cassette, only through the circumstance that the typical Kantenflimmern is to be seen, how it in analogous film, falls.
So that the picture becomes somewhat better, I would recommend each, such a solution only anamorph to encoden. To compress the picture then at the television and to get now one somewhat sehbare picture quality.

The matrix

In order to received a few more sharpness, I took here the given TMPEG matrix MPEG standard and worked in addition with the filter sharpen edge and placed both values on 20/20. With the quality CQ65, I was also here not satisfied so that now again CQ68 was adjusted. That is a somewhat good compromise between picture quality and file size. All other attitudes are had how.

KVCD_320x224_CQ68, download testMPEG,. trp and. mcf

5Th test, KVCD with320X160 film picture, KVCD_vcd

To the end now yetthe KVCD_VCD_letterbox, that I received only because of the totality, comes. It is into no manner an alternative to the other forms of representations because it has really a very bad solution. One can bring perhaps the sharpness regulator yet on 35/35, only has that altogether little effect because simple the picture is already too spongy at itself. Otherwise are it exact the same attitudes as well as before, only with a motion area of 320x160.

KVCD_320x160_CQ68, download testMPEG,. trp and. mcf


Data ratio = film length
My attempts showed, fundamentally possible is to be burned that it an entire PAL-film on a 90min CD. However such a film will not be allowed to be rise justifies are to be obtained scarcely longer than a 100 min in order to got a picture quality, that that „“, for with a well made, standardmäßigen !!, anamorphen SVCD thoroughly 90% DVD picture quality (on two CDS). I could not unfortunately the results out of the KVCD forum nachvollziehen especially since yes into its Template certainly unfavorable attitudes are available.

Large technical difficulties result again and again from the slight data ratios whereby it comes then in addition, that picture and sound synchronically run no longer. There it utilizes often also not newly to muxen, for the drive itself with the internal Playersoftware is not to be braked in the situation so strongly in order to select the data. That, and the bad picture quality in slight data ratios, was the reason for a lower data ratio boundary of 650 Kbps videos (attempts at the Philips751). Simultaneously the meaning has that I not so highly must screw CQ in order to come below around yet on sizable picture results. For only with a considerably larger CQ value than a CQ65, a good, lower picture quality would have been available, made would have that again also nothing other when the data advice to raise. At the same time raises however very strongly the middle data advice and therefore suits itself again less film to the CD.
It arrives the expanse naturally also always on the film even how largely final the file can become in order to get picture quality. The film plays much in the dark or much in the desert or, or, or. The parameters for this are considerable and therefore is no 100% given to ige guarantee that all, picture quality suit and game film time, together.

The primary KVCD forum trial was undertaken obvious at NTSC-films. In progressive NTSC DVD-game film, however only 24 pictures are per second on the disk and yield in a 90 minute films 135000 frames. If the same film in PAL would be played back, it 25 pictures would be with one around 20% higher picture information. This Mehrinformation simply must be processed and so comes it also there that in a clean PAL-conversion round 20% (120 min NTSC vis-á-vis 100 min PAL) less film suit on the disk. An exit is yet, to use a 99 min CD, however it can come here then to lesetechnischen play back problems.

KVCD picture and Tonqualität
Here it becomes difficult. The KVCD is especially susceptible to furnish in slight data ratios a yet sufficient picture quality. Under a certain value should the data advice already therefore on no case go. To hold around the quality m. E are. 650 Kbps videos the lower boundary, in addition to the already mentioned play back problematic nature. Towards the top theoretically the SVCD-boundary can be taken, however here then the film becomes the larger. In how far there can be an approach at the upper SVCD-data advice, must show yet more detailed tests. Are momentary max. 2000 Kbps set and therewith I got thoroughly good results in Action loaded films.
Generally the picture the sharpness is missing because the solution not reaches once the value of the SVCD. A possibility is, that anamorphe KVCD can recommend that I to each, for one delay himself here again in one „standard“ and can estimate the results. Also one approaches himself the SVCD solution strongly so that one certainly sharpness is reached. Also the KVCD Letterbox is lies yet in order and the picture clearly over a VHS-video cassette especially since then more film suits on the CD.

Through the large GOP, prinzipbedingt is to be recognized a pumping by the PC monitor, disappears will let that to the good fortune at the TV screen however the total picture somewhat fuzzyer. Also the scenes recognition helps in order to start in a cut with a new, good picture quality into the GOP.

In the sound, one must automatically the expansion with the Audio programs, tooLame and rrsc o. ä. distinguished because TMPEG is not even in the situation to transfer these strong requests. Therefore one gets a Tonfile that can be heard thoroughly also yet on a HiFi unit. The attitudes result had how, make the remainder then the plugin programs.

Terminating I am the opinion that the KVCD is an interesting alternative, if it concerns films, with which one on very good picture quality renounce wants/can, only will get one 120 min not yet in sufficient quality on the CD. Optimization possibilities are yet some available especially since at the matrix another little more can be gotten out. Therewith therefore yet Potenzial remains in order to bring perhaps once 110 min game film in good quality on the CD.


Diskussions thread in the TMPEG area
Reply With Quote
06-10-2002, 03:26 AM
kwag kwag is offline
Free Member
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
Thanks Slab!

I think I understand maybe 40% of it. These on-line translators do a so so job.
But this one is better than altavista's babelfish

Reply With Quote
07-22-2002, 10:33 AM
Free Member
Join Date: Jul 2002
Posts: 7
Thanks: 0
Thanked 0 Times in 0 Posts
Hello kwag, hello everyone,

well, in the last weeks I´m read only the forum because I test, test, test material to the second part of my KVCD article into the DVD-SVCD-Forum.de. I was not realy happy with the quality of the encoded sequences in PAL. Kwag have found out better and better results for the NTSC films but with PAL movies the results are one or two steps more worse. But the second part of the article are ready.
At this time I´m looking for a matrix it redused the bigness of the whole file by the same picture quality. Maybe it is possible to take 3 to 5 minutes more on a CD.


second part of KVCD article
Reply With Quote

Similar Threads
Thread Thread Starter Forum Replies Last Post
KVCD Official Specifications are now posted! kwag Video Encoding and Conversion 0 06-27-2003 10:58 PM
How can I take screenshots and posted in forum??? MUS Off-topic Lounge 8 02-02-2003 08:53 PM
svcd vs. kvcd? Do we just abandon SVCD? Mojo Video Encoding and Conversion 6 09-08-2002 10:16 AM
New KVCD template posted kwag Video Encoding and Conversion 22 05-24-2002 04:31 PM
New Revised KVCD Templates Posted. kwag Video Encoding and Conversion 0 04-27-2002 06:22 PM

Thread Tools

All times are GMT -5. The time now is 01:22 AM  —  vBulletin © Jelsoft Enterprises Ltd