More and More people have started to use their PCs for Multimedia Playback and Content creation, it is hard to miss how much Video is being shared today by people online.
Most of us have a lot of Movie DVD's ( Originals ) with us that we want to backup, however most people tend to use lousy Software solutions for this, so keeping this in mind I have decided to post a Tutorial how to get good quality backups out of your collection.. I will not explain how to rip copyright protection as it is against forum rules, if you want to learn that find it elsewhere on the web.
This tutorial is for complete beginners who have no clue about video encoding, advance users stay away..
NOTE: This tutorial is for backups of DVDs and Video in general. Blu-ray encoding is not covered because it little more complex. That will be dealt later...
SOFTWARE USED: Handbrake, Why?
Reasons: - Open Source
- Completely Free
- Simple to use, ideal for beginners
- Bloat-free, doesn't come unnecessary crap
- You can get very high quality, with right settings
- You don't need to install anything extra for it to work
- Uses high quality x264 library.
- Available for Windows, Mac, Linux ( Fedora and Ubuntu )
Download stable builds from here: ( 32 and 64-bit )
HandBrake
However I recommend that you download the latest nightly build as they use the latest versions of x264, don't worry I have never faced a single crash myself they work very well... ( 32 and 64-bit )
Download from here:
https://build.handbrake.fr/
Windows users also download .NET Package 3.5 from here ( I am not sure if it is required but the developers recommend it:
Download: .NET Framework 3.5 Service pack 1 - Microsoft Download Center - Download Details
--------------------------------------------------------------------------
Install and Launch Handbrake
Note. On windows Handbrake minimizes to system tray, this can be changed
in Tools->Options->Advanced
--------------------------------------------------------------------------
--------------------------------------------------------------------------
Step 1.)
Add Source File to Handbrake
- Click Source, Video File ( Ctrl+O )
** Handbrake may throw a dialog, close it **
Step 2.) In Destination, click browse and select name and destination of your file.
In Presets click High Profile ( Important )
Step 3.) In Output settings, select
MKV Container.
** Container: Contains your Audio, Video, Chapters, etc, MKV is recommended since it supports the largest collection of codecs **
Step 4.)
In Pictures tab
- Select Anamorphic to Loose
- Handbrake automatically crops your Video, to remove the ugly bars on top and bottom of your source ( if present ), in rare cases you may have to do it manually.
- Click on Manual button to crop.
Step 5.)
Video Filters tab
- Change Detelecine and Decomb to default if they are not
- Decomb will usually get rid of interlacing in your video, if your video has excessive interlacing ( Horizontal lines ) use Deinterlace.
- Detelecine will fix NTSC DVDs framerate ( Rare in India ).
Step 6.)
Video Tab
- Select Video Codec to H.264 ( Gives the best quality )
- Leave framerate at Same as source
- If you have a NTSC DVD source select Variable Framerate else select Constant Framerate.
- Select Avg Bitrate, enter bitrate of your choice ( Higher the better )
- Check Two-pass encoding and Turbo First-pass.
** For Quality Backups Select a bitrate between 1500 or 2000, this will help you get a good quality, don't go beyond 2500 for DVDs, it is just wasteful **
** Higher bitrate doesn't always mean better quality, though it is a fair indicator, your encoding settings matter the most **
Step 7.)
Audio Tab
- Select Audio Codec to be AAC ( ffmpeg ) or AC3 ( ffmpeg ), for best quality use passthru versions.
- Select Audio bitrate, higher the better. ( 192-320 Recommended)
- Leave the rest as it is.
Step 8.)
Subtitles Tab
- If your source contains Subtitle, you will see them here, you can chose to remove them if you want ( Recommended )
Step 9.) Chapters Tab - Really no need to get into this, skip...
Step 10.)
Advanced Tab
-
Reference Frames - Higher the better, don't go over 6
* Recommended Value - 6 *
-
Maximum B-Frames - Higher the better, keep between 4-6
* Recommended Value - 6 *
- Check CABAC Entropy Coding, 8x8 Transform and Weighted P-frames
-
Pyradimal B-Frames - Normal ( Default )
-
Adaptive B-frames - Change to Optimal
-
Adaptive Direct Mode - Change to Automatic
-
Motion Estimation Method - Change to Uneven-Multihexagon
* Don't go beyond Unveven Mult-Hex, excruciatingly slow - no visible quality gain*
-
Subpixel ME & Mode Decison - Keep between 7 & 11
* Recommended - 10 or 11 *
* Stable builds don't have 11 *
-
Motion Estimation Range - Keep between 16 & 32
* Recommended - 32 *
-
Adaptive Quantization Strength - Leave it as it is
-
Psychovisual Rate Distortion - Leave as it is
-
Psychovisual Trellis - Leave as it is or try between 0.15-0.20
* Recommended - 0.15 will do for variety of sources *
* To see the change, see the change in the text box below *
== It should read psy-rd:1.00,0.15 == ( psy-rd is experimental )
-
Partition Type - Change to all
-
Trellis - Change to Always
-
Deblocking - If you want to
- Sharpen the video a bit: -2 , -1
- Smooth out ugly blocks: 1 , 2
- General use: Default, Default
* Don't go lower than -2 or higher than 3 *
* Keep No-DCT Decimate unchecked *
NOTES: The recommended settings are meant for maximum quality, but will come at cost of speed, so tinker around a bit.
Review your settings again.
HIT START and leave your computer alone, encoding is time and resource consuming so be patient. Quality Encodes require time.
Happy Encoding!
Ask your queries if you want..
================================================== ================================================== ======================================
================================================== ================================================== ======================================
ADVANCED SETTINGS IN DETAIL:
----------------------------------------
Video is made of frames, like 1 second of video contains 24 frames.....
When these frames, are played back at high speed it gives an illusion of movement which we call video..
Now these frames are not the same...
Modern video consists of 3 types of frames
- Intra ( I ) frames, also called keyframes
- B - Frames or Bidirectional frames
- P - Frames or Predictive frames
The way modern video is reproduced is such that frames reference each other... i.e... they look into the differences between them and are then recreated...
let me explain more clearly... suppose you have a 1 second video of an apple moving across a static background.....
Now out of the 24 frames of video, one frames let us say X1 will have entire data about the frame ( i.e apple, screen )....
The next frames will however will only contain information about the apple... when these frames are accessed then the image is recreated using this frame + from the information from X1 frame...
The frames which are referenced are called reference frames... obviously this is highly simplified here.. in reality its really more complex...
Anyways - These 3 frames differ from each other in some ways...
1.)
I-Frames: The I frame stores the entire image and so are the least compressible. Video seeking requires I-frames. I-Frames are referenced the most in video..
2.)
P-Frames: P-frames store the difference between itself and an I-frame or other P-frames. P-Frames can be referenced however they don't store as much as data as I-Frames because they are compressed more than I-Frames.
3.)
B-Frames: B frames also store the differences between itself and other frames as well but they can't be referenced so they are the most compressible. x264 has a feature though called B-Pyramid though through which B-Frames can reference other B-Frames... more on that later....
So as we can see... in nutshell.
Reference frames: Higher you have better quality you have.. however it is pointless to go over the 8 because only finite number of frames can be referenced..
More you have lesser compressed the video will be and slower will be the compression..
B-Frames: Same as above... don't go over 6 as B-frames are the most compressed higher values will increase encoding time exponentially with no quality gains..
CABAC Entropy Coding: This is one of the last stages in encoding... After a video is encoded it can be compressed further by applying lossless compression just like zip files....
Now this can be done in two ways either by using CABAC or CAVLC.. CABAC is extremely efficient it brings down your video size by at least 15%... but it is highly resource intensive...
Think of the difference between CABAC and CAVLC as the difference between 7z and Zip format however more complex..
Pyramidal B-Frames: Turns on B-frame Pyramid option, which allows B-frames to be used as references for other B-frames. Atleast 2 B-frames are required... Normal option gives more flexibility to the encoder to use B-Frames reference wherever required therefore this is the recommended option...
Weighted B-Frames: Improves compression and fades in scenes with gradual transition.. keep this on at all times for quality....
Adaptive B-Frames: This allows x264 to determine the number of B-frames to use (within the set B-frames limit ). Setting to Optimal means x264 makes more accurate predictions... keep it atleast to On.
Adaptive Direct Mode: While x264 analyses frames it does so in either two ways.. Spatially which means it searches for differences in current frame itself or searches for differences between other frames called Temporal.. because of the way video is composed.. spatial is used in 90% cases .... it is good idea to keep it at Auto ( hardly any loss of speed ) and let x264 decide from scene to scene... The result in most cases will be like Temporal- 15%, Spatial- 85%.
Motion Estimation: This option selects the way motion is detected by x264. Motion is what compression codecs are all about, tracking differences between scenes to allocate the various frame types and bitrates. These are of currently 5 type which decide which algorithm to use
Diamond: Very fast but bad quality.. use it for quick dirty encodes..
Hexagon: Nice quality and good speed.... use this if you want to have decent quality video available.
Uneven-Multi Hex: Brilliant quality.. but it is quite slow.. use this if you want great quality encodes...
Exhaustive & Transformed Exhaustive: These are incredibly slow and will make you tear your hair out... In this case instead of Block search, pixel search is used and therefore it is very slow.. The difference between these and UMH is almost negligible and the speed cost is very high.. These two are used for testing purposes..
Subpixel Motion Estimation: A very important option that determines how x264 makes decisions about motion estimation. 1 is fastest and dirtiest and 11 is slowest and of best quality..
Don't go below 6 as below 6 it will disable some important features of x264 which preserve quality... 11 or 10 gives best quality...
How frames are encoded ( Simple ): Frames are encoded by dividing the frame into small square blocks, and then Motion estimation is done on these blocks to do prediction...
These blocks are called Macroblocks.. you might have noticed that in cases of badly encoded video what you always see are ugly square shaped artifacts cropping up in video .... these are the edges of the macroblocks when they were being encoded... this happens because algorithms for motion estimation work on separate blocks without sharing info with other block ( hope this is changed in H.265 ), so the edges sometimes get defined when bad encoding techniques are used.. Notes on this topic later...
Motion Estimation Range: This setting defines how many pixels are analyzed for motion estimation. Higher range values result in a more accurate analysis, but will also slow down the encoding speed significantly... 16 - 32 are great values.. you can go higher but it won't give more quality...
Partitions: These options determine the partition search types. Enable all to improve quality.. what this means is that x264 will divide video in smaller macroblocks for Motion Estimation.. at the expense of speed.. turn it "All" to improve quality... these are P4x4,P8x8
Trellis: This option improves the rounding of Transform coefficients... This is very resource intensive but also gives great quality... Use only in multi-pass encoding and use atleast 1.
( Transform Coefficients: This is tough to explain without maths ... All current encoders work on the basis of an algorithm called DCT ( Direct Cosine Transform ) and these coefficients are part of this process... Discussion of this just off the scope so lets leave it at that.
8x8 Transform: This option has been separated from others because it has a different command line flag. This enable 8x8 transforms... highly recommended to leave this on. Adaptive DCT should be on.. in handbrake its on by default...
Psycho-visually optimized RDO & Trellis: The human eye doesn't just want the image to look similar to the original, it wants the image to have similar complexity. Therefore, we would rather see a somewhat distorted but still detailed block than a non-distorted but completely blurred block. The result is a bias towards a detailed and/or grainy output image, a bit like xvid except that its actual detail rather than ugly blocking. The purpose of Psy RDO is to keep the complexity of an encoded block similar to the complexity of the original block. This way Psy RDO produces an image that looks much sharper and more detailed in many cases. It also helps to preserve film grain greatly! In addition to Psy RDO there also is Psy-Trellis now. This is still considered an “experimental” feature and disable by default, but it seems to help greatly for retaining textures in the video. Note that Psy Trellis is based on Trellis quantization.
Psy RDO Strength: This setting controls the strength of Psy RDO. For normal stuff use 1.0 ( Psy-RDO is now automatic, strength just limits it ), For Animation it should be lowered to 0 or 0.2.. Why? Difficult to explain here without maths...Just accept it as a tricky beast.. ( Note: When I say animation I mean 2d cartoons like Japanese Animie not Pixar movies )
Psy Trellis Strength: This setting controls the strength of Psy Trellis. The default value is 0.0 currently, so Psy Trellis will be disabled by default. For most sources 0.15 to 0.20 will do.. don't use for animation though... I really don't understand this feature properly myself ......Sorry! I can't explain more...
Adaptive Quantization Strength: This feature tells x264 how to distribute bits... default is 1.0.. .. This is a psy-feature.. Example- Suppose we have a scene where we have a character illuminated by torchlight, his face is visible greatly but his surroundings are darker.... What Adaptive Quantization does it that it decides that a viewer's focus will be on the face rather that the shadows so it assigns more bits to the face... Cool isn't it .. The strength factor decides this 0 means bits are distributed evenly... 1.0 is default and is good for almost all sources.. most video has portion with contrasting lighting....don't touch this if you don't understand it... for animie lower it to 0.2 or 0.. most animie has simple color grading, virtually no contrasting lighting if think of it....
Deblocking: Mandatory feature of H.264, always enabled in handbrake......
Its defined like this: 1,0,0 ( On, Strength, Threshold )... Just think of a filter which decides how much smoothness to apply to video.. ( Note this is not a correct description of deblocking, its complete discussion is beyond the scope of this text, Amazingly many encoders*People* make this mistake and crank up deblocking to negative values thinking it increases sharpness, ending with artifacts and longer encoding times which could have been avoided... Yeah I am talking about all your Secetmyth, HD-lite stuff.... )
In handbrake it is always turned on....no matter what.. you can't turn it of neither you should....
Strength: This setting is also called “Alpha Deblocking”. It controls how much the Deblocking filter will smooth the video, so it has an important effect on the overall sharpness of your video. The default value is 0 and is good enough for everything. Negative values will give a more sharp video, but they will also increases the danger of visible block artifacts! In contrast positive values will result in a smoother video, but they will also remove more details.
Threshold: This setting is also called “Beta Deblocking” and it's more difficult to handle than Alpha Deblocking. It controls the threshold for block detection. The default value is 0 and should be enough to detect all blocks in your video. Negative values will “save” more details, but more blocks might slip through. In contrast positive values will remove more details and catch more blocks.
In general 0,0 is always good.... for all purposes... don't touch it unnecessarily if you don't need it as it greatly affects speed and quallity...
No DCT Decimation: Just don't touch this.. keep it unchecked... In rarest of rare cases some blocks appear in sources with straight gradients and checking it avoids it but may produce artifacts elsewhere in the video .. meaning just don't touch it...........
P
hew! This covers the options exposed by Handbrake... Many more incredibly important options which usually expert encoders tweak are not shown by Handbrake as this program is meant for general purpose encoding.... To tweak these use the x264 CLI directly or MeGUI.... On these way later...
Note all explanations given above, I tried to make them as simple as I could, but in reality these things highly complex and explaining them like that will fill tomes... so I have avoided them...
I Hope this helps..... Good Luck.... Please provide feedback to improve his article....
Video Encoding: For Beginners
Re: Video Encoding: For Beginners
Till date, I used mediacoder 
Linear Mode
