Best quality choice

vbimport

#1

Constant o Variable bitrate??

I can’t able to see the quality differences between the two compression tecniques…

I’m not interested at the transcoding time…, please don’t answer me thats depends to long or short movies. Needs only the way to obtain the best quality w/ DVD2one 1.2.

Thanks


#2

Which is best, an orange or a peach?

You ask a question with no answers. The two engines use different techniques, one of which may be better for a particular DVD than the other.

You admit that you cannot tell the difference, but want someone else to make a choice for you. And how would they choose?

It can’t be done. Sorry, but there we are.

-Pete


#3

Originally posted by 0nion
[B]Constant o Variable bitrate??

I can’t able to see the quality differences between the two compression tecniques…

I’m not interested at the transcoding time…, please don’t answer me thats depends to long or short movies. Needs only the way to obtain the best quality w/ DVD2one 1.2.

Thanks [/B]

yeah as Peter McCall said its hard to tell you what you need if we dont have info about the movie your trying to backup :slight_smile:

whats the name of the movie ?

if its short movies IE: under 1 half hours then use variable ratio if its over 1 half hours use constant ratio !

thats the method i use. and its for FULL movies All Extras All Menus.

use constant ratio for movie only variable ratio for Full MOvie.

its hard to tell you but this is the methods i use. :slight_smile:


#4

Tell me something. Why, if ‘Constant’ mode is good for long movies, is it only so for ‘Movie Only’?

-Pete


#5

I’ve tried Aliens: Special Edition (movie-only), approx. 154 minutes. Variable bit rate looked much better than constant. In my opinion, variable is the best choice for movie-only back-ups. Allocating more bits in scenes that need it seems to be a better choice than having a constant allocation throughout the whole movie, resulting in all scenes having a lesser quality. You don’t encode/transcode your audio at a constant bit rate, do you? Variable all the way. Naturally, someone may say I’m crazy and explain why constant is the way to go, but hey…this is just my 2 cents. :smiley:


#6

I’ve done the VARIABLE vs. CONSTANT test, and here’s what I’ve found. I used the movie UNBREAKABLE. In that movie there’s a lot of scenes where the camera doesn’t move at all for long periods of time, then it was a lot easier to see the differences. The differences are not very big, but this is what I noticed. I used movie-only and select all the audio tracks and all of the subtitles so I could have the maximum information to compress. In VARIABLE mode, I can see in the background scenes some blockiness happening once in a while. At the same time (when the blocks are appearing), the picture is less clear in the background for a fraction of a second. If you look carefully, you’ll see that quite easily. In the CONSTANT mode, well the blockiness is hardly noticeable if at all, I would say it’s almost like the original and the thing is that the picture stay sharp. Did someone else notice that? My opinion would be that big movies (6-8GB) are better in CONSTANT mode than in VARIABLE mode… I did the test with CATCH ME IF YOU CAN, and it’s the same thing. The difference is not that big though, but it is there. After 3 hours of test, that’s what I found. For smaller movies, I didn’t try any test yet.


#7

Has anyone else tried Saving Private Ryan (movie-only, one audio stream, no subtitles)? Output wasn’t very good in either mode for me (blocky artifacts seen quite often, with a strange unstable sharpness in many scenes) I’m not going to complain. I’ve always had problems with this movie, and I’m not expecting any miracles.


#8

Originally posted by DVDTekno
I’ve done the VARIABLE vs. CONSTANT test, and here’s what I’ve found. I used the movie UNBREAKABLE. In that movie there’s a lot of scenes where the camera doesn’t move at all for long periods of time, then it was a lot easier to see the differences. The differences are not very big, but this is what I noticed. I used movie-only and select all the audio tracks and all of the subtitles so I could have the maximum information to compress. In VARIABLE mode, I can see in the background scenes some blockiness happening once in a while. At the same time (when the blocks are appearing), the picture is less clear in the background for a fraction of a second. If you look carefully, you’ll see that quite easily. In the CONSTANT mode, well the blockiness is hardly noticeable if at all, I would say it’s almost like the original and the thing is that the picture stay sharp.

the fact that you see blockiness in a variable bitrate is probably the result of a poor data allocation algorithm. A good algorithm would show very little or no errors of such! And in general an variable bitrate is much better than the constant bitrate (as long as you have a good data allocation algorithm!)

greetz


#9

Originally posted by theunbeatable
[B]

the fact that you see blockiness in a variable bitrate is probably the result of a poor data allocation algorithm. A good algorithm would show very little or no errors of such! And in general an variable bitrate is much better than the constant bitrate (as long as you have a good data allocation algorithm!)

greetz [/B]

Okay I’m not a video expert. So what your saying is DTOX new engine is greatly affected by the source material encoding quality? Has anyone tried a comparison then with the same DVD on the previous version vs this new version? Also how would one figure out if the orignal DVD material has poor encoding? Is there a software than one can run to test it? That way you can pre-test before setting your DTOX to crunch and be able to pick constant or variable.


#10

How about using your eyes?

-Pete


#11

Originally posted by Peter McCall
[B]How about using your eyes?

-Pete [/B]

I’m not stupid. I’ve also read some of your other post. Why are you so rude to people. Many are new users like me and need help. We’re not experts like you, so give us some slack and not make us feel stupid for posting questions, that may seem silly to an expert like you.

Many DVDs may look fantastic on the TV, but from what the previous post says, it may still look bad when shrunk with DVD2One because of its orginal encoding. Thats why I asked that question if there is such a software that checks that, the eyes can’t judge that. If that question was not valid, all you have to do is say that its not possible. Maybe your trying to be funny?..if so them I’m sorry if I offended you in any way. :wink:


#12

Originally posted by Peter McCall
[B]How about using your eyes?

-Pete [/B]

Ok… undestood, but the way isn’t so simple.
Make an example…,
(hope to be clear but, as we see i have some troubles w/ english ;). Sorry 4 that.)

I take a movie…, make a copy.
U take the same movie and make the copy, same software, same compression, same all settings…
Then, you watch the backup that you do and so, i watch the mine

Television? Monitor? …who have a reference monitor? (…something like Barco’s monitors).
The monitors aren’t =, same way 4 the televisons.
Do u use a pc to establish the quality of the movie? What player? Hw acceleration?
Too many variables to wach and judge a quality of a backup…

…no other ways?? Only eyes and personal taste?

Thank so much 4 discussion…


#13

I was only being half ironical in my reply.

You have to realise that with the possible exception of InstantCopy, these transcoders compress. Some make a choice of how much to compress, and where. Some even allow you make that choice. But compress is ultimately all they do. Their ability to influence quality by their own excellence is limited - and ultra dependent on the quality of the original MPEG encoders used by the studios, which are not, I might add, of uniform quality. In practice, you wouldn’t want your life to depend on being able to distinguish the output of trancoder ‘A’ from transcoder ‘B’.

And yet, and yet… People will declare to you how brilliant one is and how lousy the other is. This is usually based on an examination of still frames from movies - an uncertain principle in any case. I’ve even seen comparisons drawn where the frames where actually different! It’s ludicrous.

In the end only you can tell. Your playback equipment, the quality of your screen, its size and, yes, the acuity of your vision, and God knows what else, all come together in a unique combination. It matters not if we amass an array of bitrate analyzers, 'scopes, fast Fourier transformations and all the rest if you don’t like it. They’re your copies/backups and they only have to please you, no one else.

So try them all and try to reach your own conclusions. There are no shortcuts.

There, is that better? Does that help:)

-Pete

Onion replied while I was writing this reply. As you can see, he says much the same thing, only he is much more polite…


#14

Okay, I understand then. There are just too many variables that the studios do.

Like I’ve mentioned to Erwin on an email conversation we had, I should keep the old DVD2One as well as the new one and try it on each. The engine is so fast, I can afford to spend the time to get the best results to make my choice on which results to use. Thanks for the help.


#15

Hmm. I don’t know about ‘keep the old DVD2One’

As far as I’m aware, the variable mode of the current version (1.2) is the old engine - with a rather more accurate size guestimation routine.

-P

By the way, it’s not the variables that the studios use! It’s the variations in your viewing setup and personal expectations that play the major role in influencing your perception of whether the output is good, bad or indifferent.


#16

I think there’s a bit of misurderstanding about VARIABLE and CONSTANT. We’re not talking about re-encoding as far as DVD2One mission is, but rather transcoding (or compressing). Like it’s explained in the URL (http://www.dvd2one.com/ratio.php) the CONSTANT ratio is just keeping the original variable bit rate but at a lower level, that’s all. I’m sure that keeping the exact same curve as the original is better than changing it like the VARIABLE ratio mode do. There are some scenes that risk to be too much compressed that way. I doubt that ReneB and Erwin worked as hard for the past 2 months for nothing. There should be an improved difference in CONSTANT ratio over VARIABLE ratio. I’ve seen it myself. Perhaps they could give us their own impressions about their work on that.


#17

Yes but if you use a constant [compression] rate and by doing so reduce an original bit rate from 2Mbits/sec, say, to 500Mbits what then?
The original profile might well be fine because it’s upshifted and at no point does it fall into the blockiness trap.
But, when you downshift it - without regard to this phenomenon - you can end up with areas of the movie that don’t have enough bits to display smoothly.
The idea of variable compression is to recognise this and not compress those parts as strongly as those with more bits to start off with.
All that ‘constant’ does is (a) give you excellent eventual size calculations and (b) get rid of the pulsing or pumping effects some have complained of - caused by varying the compression. The downside is, as I say, you can have the bitrate fall too low with an adverse quality impact.
Whether this will actually happen depends on the original you’re working from and so I suggest (again) that you really do have to try both methods every time to see which works best. There’s no hard and fast rule for all DVDs.

-Pete


#18

I understand your point of view. There must be a trick though for deciding which method is best to use. Maybe with a bitrate viewer or calculator over the original vob files first to give us an idea or any other tool. No one will have the time, the energy and the willing to do the two methods for each backup they want to make.


#19

UNBREAKABLE is a rather slow movie. CONSTANT ratio gave me a better result. But what about a movie like TITANIC (not anamorphic and not so good original) or SAVING PRIVATE RYAN (anamorphic and super quality original) or any others… it’s really hard to know. We thought that DVD2One was an easy way but now it looks like a harder way…


#20

tried about 10 movies so far with v1.2.0 (movie-only) and all i can say is: constant ratio looks better, on the tv AND the tft, no matter how long the movie. barely any blockiness and no color pumping at all. period.