CQM comparison - V4 - Finished

[Soulhunter]

Once again...

Now with full length samples !!!


1. How is it supposed to work ???


I did a encode of this source...

The Matrix (German/PAL/R2) - The lobby shoot-out (03:08 min.)


With this AVS...

LoadPlugin("C:\Stuff\MPEG2DEC3.DLL")

Mpeg2Source("C:\Stuff\Lobby.d2v", idct=2)

Crop(0,78,720,420).LanczosResize(720,288)


And this settings...

XviD v.1.0.3 (Koepis) / Defaults / 2Pass @ 1100kbps (VHQ4)


With 5 different quantization matrices !!!


Then I renamed the files like this:

CLIP_NO_003.avi



2. But how to vote ???


Simple, watch the samples under your "usually" playback conditions...

Note: Please make sure that all post-processing is disabled !!!


If you are not able to see a difference between the samples...

- Reduce the distance to your screen

- Raise the brightness & contrast


If you are still not able to see a difference...

- Reduce the playback speed



When you got a good impression, write down a number between 1-5 !!!

1 = Crap
2 = Bad
3 = Tolerable
4 = Good
5 = Perfect


Then send me your votes like this...

CQM_NO_001 = 2.0

CQM_NO_002 = 1.5

CQM_NO_003 = 3.0

etc...


And after some time (~8 weeks), I post the results !!!



3. How to get the samples ???


1st. Click here... (http://service.gmx.net/mc/JG8CzluxLaM8SDMfHCBqKYDZFqoFLI)

2nd. Press this button -> http://img60.exs.cx/img60/6854/bt_gmxmediacenterstarten.gif

3rd. Download all 5 zip files from the folder and unpack them !!!



4. Finished...

The results !!! (http://forum.doom9.org/showthread.php?s=&postid=648109#post648109)



Thx n' Bye

[Teegedeck]

I'm glad you struggle on with another blindtest! :)

And now that you're closing in to medium bitrates it is becoming much more interesting for me! Thanks.

[CruNcher]

my decission has allready fallen i'll use Sharktooths V3ULR matrix in the next EDP build :P but i atend as allways :D

[Teegedeck]

Ow, and I had so hoped not to see something low quality, this time. :(
But at 1000 kbps, what ever did I expect? How long will it take till we get to a 2000 kbps roundup? ;)

[Soulhunter]

Originally posted by Teegedeck

How long will it take till we get to a 2000 kbps roundup? ;)

You mean the "lost" 2200kbps round I have skipped ???

Well, I gonna post it next after this round... ;)


Bye

[Teegedeck]

Voting will take some time, certainly not easy to judge...

[*.mp4 guy]

Yes they are all quite similar. Except for one that seems quite a bit more smoothed out then the others too me...

[Soulhunter]

Btw, what do you think about my new "less matrices but full length samples" concept ???


Bye

[Didée]

Generally I like the concept very much.

It's only a pity that, due to the fact i live in hi-tec country Germany which CANNOT SUPPLY ME WITH ANYTHING FASTER THAN 56k DIAL-UP, I'll not be able to participate in this round, again.

[*.mp4 guy]

Originally posted by Didée
Generally I like the concept very much.

It's only a pity that, due to the fact i live in hi-tec country Germany which CANNOT SUPPLY ME WITH ANYTHING FASTER THAN 56k DIAL-UP, I'll not be able to participate in this round, again.

:(, I've used dialup before. Definately not fun.

It was nice too have the full clip, however it didn't really seem necessary to rate the one that looked best. I think maybe 30 seconds with them all in a 7-zip file might be more conveniant for most peaople. Testing a smaller number of cqms at once tho seems like a good idea to me.

[Soulhunter]

@ Didèe

Hehe, hi-tec country... :D

Whats with a DSL connection via ASTRA (http://www.ses-astra.com/market/deutschland/internetviasat/nutzen/startseitenutzen.htm) ???


@ All

The next round is 4x50MB huge... :\

Would a additional "4x10MB" dial-up zip do it ???


Bye

[*.mp4 guy]

Imo 40 mb will take hours with dialup. Also 7-zip usually gets a lot more compresion then zip does, fyi. If for some reason you dont like 7-zip... well never mind.:D

[Soulhunter]

Originally posted by *.mp4 guy

Also 7-zip usually gets a lot more compression then zip does, fyi...

Well, not for AVI files... :\

http://img90.exs.cx/img90/9900/zip2ki.png (http://www.imageshack.us)

http://img90.exs.cx/img90/8013/7zip4pf.png (http://www.imageshack.us)

Btw, I zip'ed the files not to make them smaller...

I did this to prevent that you review b0rked files !!!


Bye

[Richard Berg]

600KB/s, nice server! I'll let you know in a few.

[Soulhunter]

Originally posted by Richard Berg

600KB/s, nice server! I'll let you know in a few.

And now imagine the upload speed... :D :D :D

If you know a better free hosting solution, plz tell me !!!


Bye

[*.mp4 guy]

:confused: ...I thought 7-zip was better for everything, I'll have to look into this. Also why would non-zipped files be borked? does zip have error protection?

[Sharktooth]

It's really hard this time. Detailwise they're all there, except one clip.
Can you elaborate some more on the voting?

[Joe Fenton]

Originally posted by *.mp4 guy
:confused: ...I thought 7-zip was better for everything, I'll have to look into this. Also why would non-zipped files be borked? does zip have error protection?

All archive formats have error detection, and many have the ability to include a certain percentage of redundancy for error correction as well.

No archiving format can be as good at all file types as some format is at one file type. If an archiver is designed for certain types of files, it normally outperforms any other archiver on those files.

[*.mp4 guy]

hmmmm. I just made a 7-zip of an avi that was smaller then the equivelent zip file...However I used the PPMd algorithm instead of LZMA
1 xvid in avifile test:

Original: 8,065,314 bytes
Zip: 7,869,867 bytes
LZMA: 7,871,227 bytes
PPMd: 7,822,962 bytes
Deflate64:7,872,532 bytes

2 xvids and 4 other avifiles, 40mb:

Original: 42,422,830 bytes
Zip: 41,585,185 bytes
LZMA: 41,611,095 bytes
PPMd: 41,698,635 bytes
deflate64:41,576,376 bytes

[Soulhunter]

Originally posted by Sharktooth

Can you elaborate some more on the voting?

Yeah, the difference is small...

Should I allow 1/4 playback speed ???


@ *.mp4 guy

Joe Fenton is right !!!

A bad AVI you could watch without noticing that its b0rked...

But a bad ZIP would show you a error message n' tell ya that its b0rked !!!


Bye

[Sharktooth]

Originally posted by Soulhunter
Yeah, the difference is small...

Should I allow 1/4 playback speed ???
Well, dont know if it is "correct", coz you dont watch a movie at 1/4.

[Soulhunter]

Originally posted by Sharktooth

Well, dont know if it is "correct", coz you dont watch a movie at 1/4.

Maybe I should change my 1st post like this... :\


2. But how to vote ???


Simple, watch the samples under your "usually" playback conditions...

Note: Please make sure that all post-processing is disabled !!!


If you are not able to see a difference between the samples...

- Reduce the distance to your screen

- Raise the brightness & contrast


If you are still not able to see a difference...

- Reduce the playback speed



When you got a good impression, write down a number between 1-5 !!!

1 = Crap
2 = Bad
3 = Tolerable
4 = Good
5 = Perfect

[...]

[Sharktooth]

Rise the brightness AND the contrast.

[Soulhunter]

Originally posted by Sharktooth

Rise the brightness AND the contrast.

Yeah, sure... :o

I have edited both posts !!!


Bye

[Sharktooth]

Now, can you send me the source at the usual gmail address?
I wanna play with it and x264...

[Soulhunter]

Originally posted by Sharktooth

Now, can you send me the source at the usual gmail address?
I wanna play with it and x264...

I gonna send you a pm after uploading the vfs files... ;)


Bye

[Sharktooth]

vfs?

[Soulhunter]

Originally posted by Sharktooth

vfs?

Mail providers dont like huge files, you remember ???

Hope you still have this VFS tool... :rolleyes:


Bye

[Sharktooth]

... no ...
multipart .zip or .rar (with no compression)?

[Soulhunter]

Ok, I gonna upload this VFS tool as well !!!

Hope you remember how to use it... ^^


Bye

[Sharktooth]

:scared:

[Richard Berg]

I know there's more detail in this scene (at least on the NTSC R1 disc) but I can't tell the difference between them except for the smoothed one.

[Sharktooth]

It's not easy to spot differences when there's high motion.
That's why they seem all to be the same...

[*.mp4 guy]

look at textures during pans etc to look for sliding. also look for ringing around edges. Not that you really have to look, those artifacts are present in all the encodes, just look for the one in whitch the presence of those artifacts is least offensive to your eyes. Also smoke can be a good indication of detail preservation.

[*.mp4 guy]

Soulhunter, how many people have given there results?

[Soulhunter]

Originally posted by *.mp4 guy

Soulhunter, how many people have given there results?

Except you ???

Zero... :(


Bye

[kurt]

don't be sad, soulhunter....
my votings will follow in the next week :)

[Sharktooth]

mine too. i just need some time to analyze the clips again and again... :)

[Spyn]

compare the vote result with SSIM result could be interesting...

[Shinigami-Sama]

Originally posted by Soulhunter
3. How to get the samples ???
1st. Click here... (http://service.gmx.net/mc/GEmrN9RdeNc4xlIJ5VVZe07lX4JziJ)
2nd. Press this button -> http://img60.exs.cx/img60/6854/bt_gmxmediacenterstarten.gif
3rd. Download all 5 zip files from the folder and unpack them !!!
Tia n' Bye
if only it would let me dl
or I could raed german to tell me what I'm doing wrong:rolleyes:

[Soulhunter]

Originally posted by Shinigami-Sama

if only it would let me dl
or I could raed german to tell me what I'm doing wrong:rolleyes:

Erm, whats your actual problem ???

Can't you reach this (http://i158.exs.cx/img158/4927/113zp.png) page after following step 1&2 ???

Can't you download the zip's via right-click -> "save target as" ???


Bye

[Shinigami-Sama]

ehh?
I think I just hit a 5min hiccup on the server -_-
now it;s letting me dl...:)

[Soulhunter]

Originally posted by Spyn

compare the vote result with SSIM result could be interesting...

Remembers me my first (http://forum.doom9.org/showthread.php?s=&threadid=76021) CQM odyssey... :D


Bye

[Shinigami-Sama]

ok
now that I;ve watched after that weird error >_>

opps
sorry
I was tired when I posted this
^^;

[kurt]

i think it would be better not to post here your results ---> pm soulhunter instead....
people should rate the clips by their own...

[*.mp4 guy]

Yes please remove your results, it could interfere with the comparisons "double blind" nature.

[Shinigami-Sama]

sorry about that
I guess I shouldn't start testing stuff at night anymore

[kurt]

check your inbox :D

[Soulhunter]

Ok, this makes 3 votes now !!!

Keep it coming guys... :)


Bye

[Soulhunter]

Just noticed the 4 weeks are over... ^^;

So, should I prolong the deadline another 4 weeks ???


Bye

[Sharktooth]

yes please

[Soulhunter]

Originally posted by Sharktooth

yes please

Okay...

Busy, huh?


Bye

[Shinigami-Sama]

kind of sad to see so few votes :(
I mean it only took about an hour or less ot do this test
dam lazy people and their "tried and tested" mentalities eh?

[Sharktooth]

Originally posted by Soulhunter
Okay...

Busy, huh?


Bye
Really too much busy.
I hope i can send you my votes for the weekend.

[*.mp4 guy]

Ok, I'm not trying to troll or anything, But I'm a little upset. Soulhunter put a lot of work into this test, and most people can't even find enough time to take it?! Thats just rediculouse. In the amount of time it takes to run DVDdecryptor and DVDtoAVIDG you could easily finnish this comparison and vote.

[Shinigami-Sama]

Originally posted by *.mp4 guy
Ok, I'm not trying to troll or anything, But I'm a little upset. Soulhunter put a lot of work into this test, and most people can't even find enough time to take it?! Thats just rediculouse. In the amount of time it takes to run DVDdecryptor and DVDtoAVIDG you could easily finnish this comparison and vote.
my thoughts exactly
it took less than an hour to do the test
with postpressecing on and off....

[Soulhunter]

Bump!

Six votes now... :)


Bye

[obieobieobie]

Just sent in my votes. :)

[708145]

Originally posted by Soulhunter
Just noticed the 4 weeks are over... ^^;

So, should I prolong the deadline another 4 weeks ???

Bye

Yes please :)
I'm taking it now...

bis besser,
Tobias

[JasonFly]

Soulhunter, have you set an other deadline to this comparison? The 4 weeks seems to be exceeded since a long time :) If i remember correctly, I even sent my votes after the initial deadline.

6 votes are not much and I don't know if the results can be considered seriouly in these conditions. Thinking of this low number of votes I thougt that this could be an idea to talk to doom9 in order to make an annoucement for this thread in the news page (for example, a day when he has nothing to put in the news, as it seems to happen sometimes). I know that the could lead to a debate such as "Have we the rigth to point a thread in the news? and not an other...

Anyway, that was just a thougth. I think it's a pity that so few people participate to this CQM comparison. The results could be much more "accurate" with more people.

[Shinigami-Sama]

Originally posted by JasonFly

6 votes are not much and I don't know if the results can be considered seriouly in these conditions. Thinking of this low number of votes I thougt that this could be an idea to talk to doom9 in order to make an annoucement for this thread in the news page (for example, a day when he has nothing to put in the news, as it seems to happen sometimes). I know that the could lead to a debate such as "Have we the rigth to point a thread in the news? and not an other...

Anyway, that was just a thougth. I think it's a pity that so few people participate to this CQM comparison. The results could be much more "accurate" with more people.
I wopuldn't mind seeing announcment in hte news about things going on in the forum on snow days:rolleyes:
but Doom9 seems to be a bit stressed latly so I don't I'd ask quite yet..

anyways
it is sad to that only six people voted. I mean hte test doesn't that much out of your day does it?
and hour and half tops?

[Soulhunter]

Originally posted by JasonFly

Soulhunter, have you set an other deadline to this comparison? The 4 weeks seems to be exceeded since a long time :) If i remember correctly, I even sent my votes after the initial deadline...

I extended the deadline 4 weeks (1st May) !!!

Btw, now we have 8 votes... :)


Bye

[niamh]

I voted :)
I find that on a lcd screen, they do show differences, mainly in mosquito noise/detail loss, and crispness. Didn't take me long to make up my mind ;)
2 of them I would scrap :eek: , and I found one to be quite superior to the others (not giving names :p).
On a crt, it might be hard work to spot the differences though, I admit.

[Soulhunter]

Yay, 11 votes... :)

Tia n' Bye

[Soulhunter]

Matrices and appropriate numbers


CLIP_NO_001 = Hybrid8AQ
CLIP_NO_002 = h.263
CLIP_NO_003 = Jawors 1CD
CLIP_NO_004 = Soulhunters v8
CLIP_NO_005 = EQM v3 ULR



Results from best to worst


3.03 = EQM v3 ULR
2.97 = Hybrid8AQ
2.95 = Soulhunters v8
2.91 = Jawors 1CD
2.59 = h.263



Are you ready for the next round?

-> 2200kbps / 4x 50MB... :D


Bye

[*.mp4 guy]

[edit] Heh I didn't think my matrix was that bad, it actually did quite well. :)

[niamh]

So it was h263 my favourite one? :D I can't bear ringing so it's logical..

Any chance to include SixofNine HVS in the 2200 kbps poll? it might be a little on the low side for it for an action scene, but all the same, it would be nice to see how it does.

will you snail mail the samples for 56k users in the future? :D

[Soulhunter]

Originally posted by niamh

Any chance to include SixofNine HVS in the 2200 kbps poll? it might be a little on the low side for it for an action scene, but all the same, it would be nice to see how it does.


Who knows, maybe its already included... ;)


Bye

[niamh]

By the way, you say results form best to worse, and I think there is a typo in the numbers
I noticed this thread has 3700 views, and less than 15 people or so voted :rolleyes:

[Sharktooth]

Yes, results are badly ordered
However im glad to see ULR won again:)

[Soulhunter]

Indeed, I accidentally mixed-up the results of h.263 and Hybrid8AQ !!!

So h.263 was in the right row, just with the wrong result... :o

Its corrected now, sorry for the confusion !!!


Bye

[kurt]

hehe, and i thought i had an eye_disease http://www.cheesebuerger.de/images/smilie/konfus/a105.gif

[IgorC]

1. What is Hybrid8AQ matrix? Is it AQ Xvid´s codec option? (Adaptive Quantization on)
2. Xvid 1.0.3 is the last stable version, but Xvid 1.1 is quite stable too. Why don´t use it?
3. bitrate 1100 kbit/s. Which is purpose? (1CD rip or 2cd rip ... )

I´m downloading samples. Later is better than never

[Soulhunter]

1. Is a matrix by *.mp4 guy -> link! (http://forum.doom9.org/showthread.php?s=&threadid=90520)

2. Mainly coz I did all this encodes ~2 months ago...

3. Its roughly the bitrate for a 2CD rip of the PAL version!


Bye

[MacAddict]

I've been out of the loop on CQM's for quiet some time unfortunately. Does that explain why I dont see the HVS-Best matrix used in this testing?

Thanks to all who were involved, interesting information.

[Sharktooth]

coz the clip is a very high motion part and hvs-best is a mid-high bitrate matrix (the choosen bitrate was 1100kbps).

[TheUnforgiven]

is it too much to ask for all the 11 results to be posted rather than the average. it would be useful to see which matrices most of the guys agree on. thanx for the good job

[CrashUK]

I did think EQM v3 ULR matrix was going to win I use it for every thing and at every bitrate

[Soulhunter]

Originally posted by TheUnforgiven

is it too much to ask for all the 11 results to be posted rather than the average. it would be useful to see which matrices most of the guys agree on. thanx for the good job

Well, I still have all the mails...

But I have to ask all participants for their permission to publish the mails !!!

Maybe I should add some sort of "agreement declaration" to my next comparison ???


Bye

[708145]

Originally posted by Soulhunter
Well, I still have all the mails...

But I have to ask all participants for their permission to publish the mails !!!



You may publish mine.

bis besser,
Tobias

[Sharktooth]

You can publish mine too.

edit: it's here

Ok, finally here are my votes

Clip01 -> 2.5 : crisp and deailed but too much artifacts
Clip02 -> 1 : Washed out, blocks everywhere
Clip03 -> 3.9 : Like clip05 but with few more details... and blocks
Clip04 -> 4.5 : Nice details and some blocks but i like how it looks
Clip05 -> 4 : Good compromise between details and blocks

[kurt]

I have no objections. :cool:

[niamh]

I don't think I gave any details myself, but go ahead... I wish I had now, because it seems my assessment is very different from others. I saw fewer blocks in what I know now to be h263, and horrendous mosquito noise/ringing in some others.(A friend preferred it too, though he forgot to vote :rolleyes: )
As well, it would be nice to know if you have a crt/lcd secreen, at what resolution, and what do you use to decode? I use ffdshow (on a 1280x1024 lcd 18") and I know xvid decoder looks different.

[*.mp4 guy]

I viewed the encodes an a 17" crt at 1024*768 resolution, and got these results
They were all very similar. except number 2 which was too smoothed out for my tastes.

Clip-1:3.9
Clip-2:3.2
Clip-3:3.6
Clip-4:3.7
Clip-5:3.8

[Soulhunter]

Backdoor... ^^

Originally posted by Soulhunter

And after some time (~8 weeks), I post the results !!!


So it should be ok to publish the numbers without the surrounding text, no?


*.mp4 guy wrote on 1st March 2005 18:11:

Clip-1:3.9
Clip-2:3.2
Clip-3:3.6
Clip-4:3.7
Clip-5:3.8



Shinigami-Sama wrote on 16th March 2005 20:49:

clip1=4
clip2=1
clip3=3
clip4=2
clip5=5



kurt wrote on 23rd March 2005 21:20:

CLIP_NO_001 - 4
CLIP_NO_002 - 2,5
CLIP_NO_003 - 3
CLIP_NO_004 - 2
CLIP_NO_005 - 3,5



McGee wrote on 28th March 2005 12:22:

CLIP_NO_001 = 3
CLIP_NO_002 = 2.5
CLIP_NO_003 = 3.5
CLIP_NO_004 = 2
CLIP_NO_005 = 2.5



JasonFly wrote on 5th April 2005 22:10:

CQM_NO_001 = 2.0
CQM_NO_002 = 1.5
CQM_NO_003 = 2.5
CQM_NO_004 = 3.0
CQM_NO_005 = 2.0



dr.Prozac wrote on 17th April 2005 01:53:

CLIP_NO_001 = 2,5

CLIP_NO_002 = 4,5

CLIP_NO_003 = 3

CLIP_NO_004 = 4

CLIP_NO_005 = 2



obieobieobie wrote on 17th April 2005 12:19:

CQM_NO_001 = 1.3
CQM_NO_002 = 3.8
CQM_NO_003 = 2.0
CQM_NO_004 = 3.3
CQM_NO_005 = 1.5



Sharktooth wrote on 18th April 2005 15:13:

Clip01 -> 2.5
Clip02 -> 1
Clip03 -> 3.9
Clip04 -> 4.5
Clip05 -> 4



708145 wrote on 26th April 2005 16:12:

CQM_NO_001 = 2.5
CQM_NO_002 = 2.5
CQM_NO_003 = 2.5
CQM_NO_004 = 2.5
CQM_NO_005 = 2.5



niamh wrote on 26th April 2005 16:53:

CQM 001 = 3
CQM 002 = 4
CQM 003 = 2
CQM 004 = 2
CQM 005 = 3



<dread@[...].pl> 04/23/2005 - 12:24

CQM_NO_001 = 4.0
CQM_NO_002 = 2.0
CQM_NO_003 = 3.0
CQM_NO_004 = 3.5
CQM_NO_005 = 3.5


Bye

[Soulhunter]

Btw, here is a "votes -vs- psnr" comparison...

CLIP_NO_001: Result = 2.97 / Average PSNR = 43.6605 / Overall PSNR = 43.1309
CLIP_NO_002: Result = 2.59 / Average PSNR = 44.0550 / Overall PSNR = 43.5215
CLIP_NO_003: Result = 2.91 / Average PSNR = 43.8578 / Overall PSNR = 43.3039
CLIP_NO_004: Result = 2.95 / Average PSNR = 43.5369 / Overall PSNR = 42.9814
CLIP_NO_005: Result = 3.03 / Average PSNR = 43.8485 / Overall PSNR = 43.3116


Bye

[Sharktooth]

once again PSNR does not correspond to what human eyes see...

[Shinigami-Sama]

Originally posted by Sharktooth
once again PSNR does not correspond to what human eyes see...
though it it usualy give us a rough idea of the relitive quality
I'm looking forwards to nice<2200 yay> ones hunter :D

[708145]

I was looking at the samples with totem (no pp) 2x original size on a 20" 1600x1200 LCD with a viewing distance of approx. 1m.

I'm surprised that some people voted that some of the clips looked 4 or better which means "good".

For me a 4 is transparent to typical DVD in my typical viewing setting. Perfect would be if even stills looked... well... perfect.

What I'm getting at is: I think different people annotated different meanings to the different grades => It's hard to compute an average out of numbers which use a different scale.

bis besser,
Tobias

[niamh]

What I'm getting at is: I think different people annotated different meanings to the different grades => It's hard to compute an average out of numbers which use a different scale.

True, I spread the numbers from 1 to 5, sort of. On a real scale, it would have been much lower. I hadn't thought of it this way.

[*.mp4 guy]

I took 3.5 as a starting point, because that is an "average ok" or tolerable score, not good, not too bad either. Some of the matrices got close to looking "good" to me, as in some artifacts but overall a vissually pleasing picture, none of them were quite there though. I took 5 to be "as good as the source", obviously none of the matrices were anywhere near to being transparent at this bitrate so I didnt score anything neer 5. Imo all of the encodes were definately tolerable, Although h.263 was getting dangerously close to bad for me.

[Shinigami-Sama]

I just rated based on relitive to the next
1being the worst of the group and 5 being hte best

[Brother John]

As I’m starting to learn about CQMs I guess participating in the next test wouldn’t be a bad idea.

To avoid different interpretation of the scale, why not define a scale like this:
5 = best matrix in test
4 = above the average
3 = roughly in the middle between 1 and 5
2 = below the average
1 = worst matrix in test

Would this be too complicated to judge?

[dr.Prozac]

I think next time it should be clearly described in the first post of a new comparison. Something like Brother John has just written above.

[Shinigami-Sama]

Originally posted by dr.Prozac
I think next time it should be clearly described in the first post of a new comparison. Something like Brother John has just written above.
yeah soulhunters was kinda confusing 5 samples and five numbers
soul hunter
When you got a good impression, write down a number between 1-5 !!!

1 = Crap
2 = Bad
3 = Tolerable
4 = Good
5 = Perfect

so it's kinda open and the fact that I was rating these near the end of my night didn't help my reading comprehention either, thouhg it was an accutare test seeing as how I watch most stuff before I go to bed lol

[Soulhunter]

Hmm, was the description really so inaccurate ???

Maybe I should change it like this...

1 = Looks like crap
2 = Looks bad
3 = Looks tolerable
4 = Looks good
5 = Looks perfect


Bye

[Poutnik]

Originally posted by Soulhunter
1 = Looks like crap
2 = Looks bad
3 = Looks tolerable
4 = Looks good
5 = Looks perfect
Because of subjectivity, perfection hunting man can sign the clip set in range 1-3, but tolerable man can mark the same set as 3-5 .... So best matrix can has a mark 4, and worst a mark 2.
The suggested evaluation ( 1-worst, 2-below avg,3-average,4-above avg, 5-best) has autoranging properties.
IMHO the essencial thing is, if you are looking for absolute quality (crap-perfect), or relative quality (best-worst).

[Soulhunter]

Ok, you convinced me... :)


Bye

[Soulhunter]

-> CQM comparison V5 (http://forum.doom9.org/showthread.php?s=&threadid=94222) !!!


Bye

[obieobieobie]

Interesting results. Seems like I really like h263 :)