Quality comparison: Brotherhood of the Wolf

[windtrader]

switch from DVD2SVCD to RB-Procoder?

@UncasMS,

I've been reading a few of your threads on the various testing of various encoders/transcoders. Over the past few years I have used a number of the same tools. Like many here, I use one of the transcoders if the little compression is required and encoders if more compression is needed.

I use DVD2SVCD, D2SRoba, and CCE for the encoding process and am quite happy as the combination provides solid quality encodes, often in one pass. Encodes take 2-3 hours.

The encode times you list for Procoder are pretty long, one in the order of 4-6 hours. Given your tests show CCE and Procoder both perform solid encodes, what reason(s) would I switch from DVD2SVCD, especially given that DVD2SVCD is free and the Rb-procoder solution takes longer?

Also, have you done time and quality comparisons between the various quality settings (high-speed to Mastering) in Procoder? I'm sure there are some guidelines around which setting to use for various source material.

thx

[UncasMS]

dvd2svcd does NOT handle menue - something i wouldnt want to do without nowadays

next thing is: procoder is much better in lower bitrate - c.f. my other comparisons


so yes: given high(er) bitrates like 3.500+ cce or hc will be fine using only an OPV mode and that may take 60-90 min in many cases

but i do want to keep extra material in addition to the main movie (again dvd2svcd wont be able to achieve this) and then the avertage bitrate drops down to a level where cce wouldnt be my first choice any longer and thus i have no problem with some 3-6hrs of encoding when the final results is better

[windtrader]

dvd2svcd does NOT handle menue - something i wouldnt want to do without nowadays

I see. For me, I always want the feature movie to have maximum quality (movie only) and minimum user inteface (start movie immediately).

For DVDs having extra material that I want to backup, I use CloneDVD2 to select everything except the main title then back this up which create a fully uncompressed DVD with all material except the main title.

For me this process is what I prefer as it gives me the movie only disc and another with the extra materials including the menus.

given high(er) bitrates like 3.500+ cce or hc will be fine using only an OPV mode and that may take 60-90 min in many cases

Pardon me if this has been previously answered but what method(s) do you use to determine the average bitrate before starting a backup job?

Lastly, could you kindly comment on your perceived differences in quality between the various Procoder quality settings? Do you use different settings based on some criteria or always select mastering mode?

[UncasMS]

Pardon me if this has been previously answered but what method(s) do you use to determine the average bitrate before starting a backup job?

dvd rebuilder states the bitrate after the preparation phase

concerning procoder's settings:
i dont use anything but mastering quality as i want best quality - time is no issue for me

[dravenar]

UncasMS, which software do yo use to make the captures?

[UncasMS]

the screenshots of the different frames were made with "virtualdubmod"

- load vobs
- use ctrl+g to jump to a certain frame or use the slider
- shift+1 ro save a snapshot of the sourceframe (to png in my case)

- finally converting the png to jpg with not too heavy compression

[windtrader]

next thing is: procoder is much better in lower bitrate - c.f. my other comparisons

so yes: given high(er) bitrates like 3.500+ cce or hc will be fine using only an OPV mode and that may take 60-90 min in many cases

[16:16:46] Phase I, PREPARATION started.
- "Movie Only" mode is enabled.
- VTS_01: 3,078,564 sectors.
-- Scanning and writing .D2V & .AVS files
-- Processed 187,276 frames.
-- Building .AVS and .ECL files
- Reduction Level for DVD-5: 70.8%
- Overall Bitrate : 5,329/4,263Kbs
- Space for Video : 4,064,888KB
- HIGH/LOW/TYPICAL Bitrates: 5,610/3,605/4,263 Kbs
[16:31:06] Phase I, PREPARATION completed in 15 minutes.

I ran RB and it shows an overall bitrate as 5,329/4,263Kbs. What do the two different numbers tell me?

You mentioned higher bitrate streasm can be handled by CCE. Would you expect both CCE and Procoder encodes to produce similar quality results?

thx

[UncasMS]

HIGH/LOW/TYPICAL Bitrates: 5,610/3,605/4,263 Kbs

this line may give you the answer as to the question of the different numbers

highest, lowest and average bitrates are stated here


and yes, i'd expect very similar results with cce, hc enc or procoder with these numbers but cce being fastest, followed by hc enc and procoder will take the most time of all encoders

[windtrader]

I did see the line with
HIGH/LOW/TYPICAL Bitrates: 5,610/3,605/4,263 Kbs

That's why I ask the question. I see that 4263 is the same as typical - that number makes sense but the 5329 number doesn't match any others. Also, I would expect one number to represent the overall bit rate, not two. In the log that was pasted in the front of this thread, there is only one number for overall bitrate. not two. So - what is the first number, 5329, trying to communicate?

thx

[conair]

@ windtrader

I have never seen a log file with a similar entry “Overall Bitrate : 5,329/4,263Kbs”
There is always one number that matches the “TYPICAL Bitrate”

[UncasMS]

comparing frames of HC 017 vs 018 can be found here:

http://forum.digital-digest.com/showthread.php?p=391464#post391464

[windtrader]

any ideas what this means?

[20:36:33] One Click encoding activated...
-----------------
[20:36:33] Phase I, PREPARATION started.
- "Movie Only" mode is enabled.
- VTS_01: 2,851,752 sectors.
-- Scanning and writing .D2V & .AVS files
-- Processed 147,231 frames.
-- Building .AVS and .ECL files
- Reduction Level for DVD-5: 77.4%
- Overall Bitrate : 7,035/5,628Kbs
- Space for Video : 4,218,654KB
- HIGH/LOW/TYPICAL Bitrates: 5,650/4,991/5,628 Kbs
[20:40:17] Phase I, PREPARATION completed in 4 minutes.

Here is another example of strange overall bitrate being reported. I have no idea what RB is computing and reporting (7,035/5,628kbs).

[jdobbs]

The reported bitrates you are quoting are the average bitrate for the entire stream. Since you are seeing two number you are very likely an NTSC user. The first number (7,035Kbs) represents the actual playback bitrate applied to the stream and the second number (5,628Kbs) is that bitrate used to encode at 23.976fps.

The first bitrate is the one typically reported by other software packages or your standalone player when it is played back. On most movies (film sources) there are only 23.976 true frames encoded for every second of video and it is interpreted to 29.97fps (the NTSC playback standard) during playback. So the bitrate being pushed out of the player is 25% higher than the the bitrate at which it is stored.

The first number was added so I would stop getting questions saying "but my other program showed a higher bitrate so it must be better" -- the fact is that the other program is not showing actual encoded bitrates. Sometimes it's hard to convince people that for any set of video that occupies the same space on the disc at the same frame rate you always get the same bitrate -- no matter what program you're using... that's just the physics of it.

The three numbers you see across the bottom are the HIGH/LOW/TYPICAL bitrates that have been applied to individual segments of the encode (e.g. cells). Each cell has a different level of demand depending upon the complexity of the picture -- so naturally it will have a different resulting overall bitrate.

[windtrader]

jdobbs - thanks for the reply as it is exactly as you state. A suggestion to make this clearer might be to note in the log in parenthesis that the higher rate is the NTSC playback rate.

thx again

[rack04]

the screenshots of the different frames were made with "virtualdubmod"

- load vobs
- use ctrl+g to jump to a certain frame or use the slider
- shift+1 ro save a snapshot of the sourceframe (to png in my case)

- finally converting the png to jpg with not too heavy compression

I'm trying to do a quality test of my own but I'm having problems with the screenshots. I tried the process you stated above for screenshots of different frames but what I'm finding is that, for example VTS_10_4.VOB frame 8471 in CCE is different than VTS_10_4.VOB frame 8471 in HC Enc. Have you ever experienced this before? So far I've encoded "Hoodwinked" using CCE SP Trial, HC Enc 0.18, Procoder 2, and QuEnc 0.7. Thanks.