KVCD Vs. Ogm - My report
Ok,
I see this forum and i like it very much, in special because it have forum's for the soft i use HeadAC3/DVD2AVI/Avysynth/VDmod. And of course, i went to KVCD and i must admit it was a method i didn't new at all :oops: , I have heard of many formats, tried many but i didn't knew of this one KVCD. So i went to make a test KVCD Vs. OGM (With Xvid Codec) Considerations on this test: I only want ONE CD, no more no less :D I want Good Quality Based On one CD Movie: Final Fantasy 1H 42M 16:9 25fps PAL In XviD i use a Mode2 90minutes 1CD (895mb), and have encoded with the best settings as far as my knowledge (based on koepi's recommendations, xvid developer), and get a good result, good quality, the resolution was 608x320 and the sound was enough (Ogg Vorbis Quality 2 - 96kpbs, with subtitles and chapters) the time to encode it was about 9Hours in my AMD 2.0GHZ, I use my AGP to connect to my TV so the final result I qualify as VERY GOOD. Then i do the same thing but with KVCD, based on the user's Guide with 1CD of 80Minutes MODE2 (795MB) Because my DVD Player don't support 90Minutes Cd's (1st loosing point) I encoded the sound in MP2 With 128kpbs because in 96kpbs it sounds terrible. I choose the resolution 320x450 to fit in one CD, create the avisynth script based on KVCD templates and start encoding, it takes me about 14H in My AMD 2.0GHZ, It went very good on DVD Player but Quality it was ACCEPTABLE. Final: OGM - 90Minutes CD - Mode 2 - 895MB KVCD - 80Minutes CD - Mode 2 - 795MB (Less 100MB) Ogg Vorbis sound - Q2 (96KBS) equal to 128kpbs MP3 - Less Space KVCD MP2 - 128Kpbs (96 sound Very BAD) - More space OGM Resolution - 608x320 - aspect error - 0.6 KVCD Resolution - 320x450 - aspect error - 0.1 OGM Encode time - 9Hours KVCD Encode time - 14Hours Take note that i use the slowest settings in Xvid to get the best of it as VHQ4/Chroma/Ultra High/QTP And the same in KVCD Highest (slowest). Conclusion: I think for one CD Ogm container with Xvid is still the way to go, you can see above why, KVCD gets an acceptable quality in one CD but not so good as in OGM (My opinion). KVCD takes the lead if you go for 2CD's, yes great results, in special for the awesome resolution of 704x576. Waiting for your comments :lol: |
Sniffer,
It's a good idea to do comparissons like that. I myself used to do a lot o DivX encoding before I got my DVD burner and moved to KDVD. Unfortunatelly your comparisson isn't fair because you KVCD had almost 13% less space than Xvid (considering the CD space difference). And resoultions were not the same in both encodes. Did you try 544x480 instead of 528, or encoding using MPEG2 instead of MPEG1, or doing the VCD header trick (or SVCD header trick)? Also about KVCD: did you use the Motion Adaptative script (or any script)? Scripts play a very important paper in KVCD encoding, even from DVD sources. Also the best setting for KVCD is not the highest, incredibilly it produces worst quality than high quality. Use high quality if you're not using the motion adaptative script, or motion estimation if you are using this script. Also using XVid you have the joys of post-processing, i used FFDShow a lot before to add noise and post-process my encodes. Of course this can't be done on stand-alone, altrough some stand-alone do some kind of post-processing. Please fix this stuff and redo comparisson, It's a really interesting subject!!! :D []'s Vmesquita |
Don't Fit
My Sony DVD Player don't read 90minutes CD'S :cry: ....so i can't get A real good Comparation, i'm doing the test regarding my situation.
With a higher resolution i wouldn't be able to fit all my movies in one CD, for instance K-19 Takes about 2H 11Minutes, yes the movie stacker was used, and KVCD template as well with TMPEGEnc. So in my situation i try to tuned at possible the both of methods. :) |
Ok, but of course you could vary the resolution according to the movie... If I remember correcty, Final Fantasy is about 90 min, maybe you could use a higher res for this one...
[]'s Vmesquita |
Re: KVCD Vs. Ogm (My report)
Quote:
The purpose of KVCD is to replace a DVD and so, to be read in standalone player. What you do is comparing a sport car and a station wagon, forgetting they don't have the same goal nor the same drivers. |
Re: KVCD Vs. Ogm (My report)
Quote:
1 - Ofcourse OGM don't run in a DVD Player 8O , altought by AGP you can connect your PC to a TV. 2 - I know they are quite different, but some people as me ask. What format should i choose?? So i make a test to get some results regarding the two of the most important ones VCD Or DivX, some go one way, some go another. 3 - Yes i will Update the test to a more real comparation between OGM and KVCD, I think it's worthy. the same movie :D Best Regards Sniffer :wink: |
Re: KVCD Vs. Ogm (My report)
Quote:
As for my example with cars, you don't ask "what car should I choose" without adding at the end of the question "for the needs I have". Here is the same thing. For the need we have, there is no choice : it's KVCD or KVCD (lets forget VCD and SVCD :-)) |
Re: KVCD Vs. Ogm (My report)
Hi Sniffer,
I think your test is interesting too. But like the other guys already said, you seem to be comparing apples and oranges, and you also seem to be showing a small unfair bias towards your apples (Ogm/xvid) instead of the oranges (kvcd), due to the shortcomings of your particular dvd player. Let's pretend that I was doing a similar comparison like yours...1CD xvid versus 1CD kvcd. It would not be fair for me to say that I only have a Pentium-1 100Mhz PC with 32MB Ram that can't play Ogm/Xvid correctly, so therefore, KVCD is better and it wins since it works on my dvd player. However, that is sort of what you are doing since your DVD player does not play 90minute CDRs and you counted that as as strike against KVCDs. Please keep in mind that some people do have DVD players that do play 90-99minute CDRs and even support 48khz mp2 audio too, which helps the mp2 audio sound even better @ 128kbps (I'm lucky, my player does both of these things!). So when you say... Quote:
You also said that you are pretty new to KVCDs, so before you make your final decision on who wins "KVCD vs Xvid", you should make sure that you are using the exact latest methods (Motion-Adaptive script, etc) that others here are currently using on their KVCDs to ensure that you get the best possible quality. As vmesquita said, you should also probably be able to use a higher resolution for your 1CD kvcd, like 352x576 or 480x576 or maybe even 528/544x576 (but as a PAL user, 528/544x may be too high for you to hope for on a 1hour 42minute movie). I realize that you are stuck using a PAL source and resolution, so you are somewhat limited by it and will never be able to get results quite as good as if they were coming from an NTSC source. Quote:
Anyways, you wanted feedback, so I guess you got it! A real mouthfull! :lol: :wink: Just make sure that your comparisons are done fairly and equally, otherwise they aren't quite as valid (although, whatever works for you is valid for you! :wink: ...every format has some benefits...and also some drawbacks). But your test is still interesting and I look forward to hearing about your new test results. :wink: Although I'm affraid that the only person whose mind you might possibly change here would be yourself. As you can tell, we do love our kvcds here. :lol: :wink: Best regards, -d&c |
My next Test
1 - I'm not trying to make a unfair comparation, telling this is better and that is worst, the test above it's like that regarding my situation although i will make a fair one and will put some images in here :D
My new test Will be this one: FAIR COMPARATION :wink: OGM - 90M CD - MODE2 - 895MB KVCD - 90M CD - MODE2 - 895MB (The DVD from my girlfriend support it, it was a gift from me :D - Support MP3/CD-R/CDRW/VCD/SVCD) OGM Sound - 128kpbs - Surround2 2channels MP2 Sound - 128kpbs - Join Stereo, because dual channel repart the bits for each channel 64kb, wich i don't agree with the guide. OGM Resolution - 608x320 Mpeg4 KVCD Resolution - 528x576 Mpeg2 Obs: In KVCD i will use CCE Because i think it gives the best results (Quality), in many test comparations between TMPEGEnc and CCE, CCE is better. :wink: Any remarks or ideas are welcome. :D I will do a PRO test with images and equal comparations, if you find good enough you can make a stick of it, to members get a good idea of what they are dealing with. |
Re: My next Test
Quote:
Quote:
You plan to have 1.5 more pixels to encode for the KVCD than for the OGm. There again, no need to test, I can predict the result for you : OGM better. Quote:
Quote:
Pro test = EXACT same thing to encode and EXACT same way to watch them (the picture on a TV set is'nt the same than on a PC monitor + the size as to be the same, and that are only some elements) Note: I used to be a part of the test team of MPEG2 when it was designed 15 years ago. I did some "subjective visual tests" whose purpose was to find best ratio between compression and quality loss. |
@DialHot
I guess you misunderstood the resolution thing: he is probably doing 528x480 with black bars. With XVid you can use any res you like (and generally 1:1 aspect) so 608x320 should be without blackbars. So in this case KVCD has an advantage. @Sniffer Use TMpgEnc and MPEG1 in the test also... That would be nice to see... People here generally use Mpeg1 and TmpgEnc. i however use CCE a lot since I do only KDVDs. []'s Vmesquita |
Test
Dialhot, man you are difficult to deal with :lol:
Yes in Xvid It will be without black bars because is one way to give better quality to the image. Yes In Xvid we can put the resolution we want , BUT WE CAN'T make the same in KVCD, so i will put the best KVCD resolution to 895MB CD. I use Ogg vorbis Surround2 2 channels Because it's the best way for this format, and I think in KVCD dual channel isn't the best way because it splits the bitrate , and Join Stereo gives better quality and small file size wich give me space to improve the video (If you like MP3, why DibRom (Lame Developer) always use Join stereo instead of stereo or any other in his lame codec and preset's??) Ok, Vmesquita, TMPEGEnc will be used :wink: , não há problema algum. Again: I'm doing this test to solve some doubts of my own, and i know the formats are not equal, and I can't make evreything equal in both.....That's not possible at all...But we can squezzee the best of the both and compare them in the final result. :D |
What Settings
It will be used for KVCD
TMPGEnc with MPEG1 Resolution see above Latest Script from KWAQ CQMatic will be used as well High Setting because highest is worst (as far as i hear) Sound: Surround Instead of Join Stereo It will be used On OGM Koepi's Codec Ogg Surround 2 VHQ4/6 UltraHigh/Mpeg/QTP/CHROMA And alternative Curve. |
Just a correction:
If you are using the latest kwag Avisynth 2.5 script (the motion adaptative script) use Motion Estimation instead of High Quality. It is reported to produce better quality with this script. And goes faster! Manda ver! :D []'s Vmesquita |
@ All..
Let be nice to one another. This is a FUN place, not to mention, the idea of this (by chance) comparison too :lol: Respect each other :lol: 8) :lol: @ Sniffer.. Quote:
that much of a factor to me, cause in the end, it will all depend on the user's skill level :wink: But, I as I agree w/ Dialhot, I also feel that the resolution should be the same, if you're gonna peform a test comparison, just so that every block is block for block compared (encoded, that is :) ) . . But, unfortunately, w/ divX, you have a Boarders issue to content with. Also, Aspect Ratio plays a VERY important part to this comparison as well :wink: And, if the two are not sync'ronised, well, you know what to expect, or at least to factor in your final analisys. :lol: But, the test should show as close to specs as possible for an as fair an comparison as can be (even though these two formats are Apples and Nuts) Some examples of equality should be: * Audio - - should be the same * resolution - - should be the same * bitrate - - well, should be close 8O * CDr media * DVD player * TV output - - well, this is tricky, cause** ... ** TV output is dificult to compare with because you have TWO items that need to be fed into the TV, and only ONE can, DVD player ( MPEG-1/2 ) while the divX in question can't :( Also, again, becuase of the nature of how DVD player respond to a given MPEG source, you also have to factor into the equation the AR and resizing that the DVD player puts out 8O Even if the user OUTPUTs his/her divX source via TV-OUT it won't be the same because A) there is not proper AR and B) there is no proper Resizing and C) the Color Space output from the graphics card is not the same as the DVD player's or wont be able to properly utilize/set the Color Space for the OUTPUT of the divX source, ..to make a valid assessment of final quality. All in all, the above (Sniffer's comparison) is a fun idea and still worthy of continuing. Well, I've said enough. I'm boring you all w/ my retentativeness hehe.. But, at least give the above some thought, and.. Have a great day :) -vhelp |
Quote:
The XVID movie HAS to be encoded with the exact same black bars and movie area. Else the comparison is impossible. But you're right on one point : the KVCD has the advantage and not OGM on this point. |
Vhelp,
I think sniffer ideia is more like "I don't care if I am going to play using TV-Out or a stand-alone or what resolution/audio I'll use, what will give me more quality in 1-CD when watching on TV"? And I think it's a interesting topic. Of course Ogm and KVCD have it's differences and it's almost impossible to get them in exactly the same conditions. So I guess it makes sense to push them to the limits to see what performs better in 1-cd, using each one's strenghts. Of course, TV-Out will resize, different TV-Outs will have different performance, and depending on the playback filters, you can enhance the ouput of OGM. And some DVD Players give better picture than others. But the idea is not answer "X is better than Y" and that's it, but the question I put in the first paragraph. []'s Vmesquita |
Re: Test
Quote:
Quote:
Quote:
Quote:
Quote:
If you want to compare, use Stereo (not join stereo) for both ! (or Surround2 for both). But there you use the best format for OGM (surround2 and the worst ever format for KVCD (mono excepted !) Quote:
Quote:
Quote:
Edit : after having read the last post of vmesquita, okay, I admit that there are two possible ways to see the question. My way is "is KVCD or XVID the best encoding process ?", and your way "with wich format will I have the biggest fun to watch my movie". That's correct; I appologize ;-) So please, just do your tests with MPEG1 TMPEGEnc and stereo for KVCD because these are the best settings for this. And let see what happens. |
may I give my two cents on this debate:
From the couple of tests I've done with xvid and kvcd (not so controlled), I guess you'll get very different conclusions if your source is animation than if it's live action (in my experience it seems xvid is better for animation and kvcd is better for live action). Maybe it's not the encode (alone), it can be related to the post-processing (dering) in ffdshow so to be absolutely fair you'd have to test the kvcd with a software player (through ffdshow) and the TV out card. |
Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.