Features Overview of DVD2SVCD

[mrbass]

Features Overview of DVD2SVCD

- Automates entire process, just need to press Go!
- Automatically selects the .ifo with longest time length
- Enables manual selection of .ifo for .vobs on harddisk
- Allows deselecting chapters such as credits
- Aspect Ratio can be specified
- Choose multiple audio tracks
- Automatically Shutdowns computer at end of process
- Various Deinterlacing options for PAL, Anime
- NTSC or PAL selection
- Load and save profiles of different parameters
- Set priority all programs
- Crash recovery to resume saving time
- Advanced batch operations
- CD Label naming the cd
- Title Picture can be downloaded from imdb.com
- ChangeCD picture display at end of cd
- svcd.txt contains encoding parameters can be put on svcd
- Fixed chapters can be set to any time length
- Original DVD Chapters auto created
- Workarounds for non-compliant standalone DVD players
- Various methods of authoring image files (I-Author, VCDImager, VCDXBuild)
- Creates .bin .cue image files to be burned to cd
- Auto splits movie into 2,3,4 cds if needed
- Subtitles (CVD or SVCD) or I-Author (CVD)
- Subtitle location is adjustable
- Subtitle palette makes choosing custom colors a cinch
- Create an overlap of the movie between cds
- Parameters to fix audio synch issues with dvd players
- Built in ripping of vobs to hd with either vstrip or internal routines
- Automatic NTSC Field Operation detection (On, Off, IVTC)
- Autodetec sound gain level to avoid distortion and achieve max sound level
- Besweet converts ac3 to mp2 on-the-fly requiring 1-2GB less hd space
- Audio downsampling option from 48Khz to 44.1Khz
- Advanced parameters of avisynth resizing filters and others
- Custom editing of .avs file
- Bitrates can be set for CCE or TMPGEnc with accurate size prediction
- CCE encoding options CBR, 1pass VBR, Mutlipass VBR
- CCE advanced parameters can be set
- TMPGEnc can be used instead of CCE for encoding
- Advanced TMPGEnc encoding options can be set (CBR, VBR, CQ), etc.
- Automatic Pulldown executed for NTSC films


I thought perhaps this might be a quick glance at the various features DVD2SVCD has to offer. Obviously if someone has already downloaded DVD2SVCD then they can just flip through the options and find out themselves. This can help if they're debating whether or not to try out the program for themselves or not. Just a thought.

[DSPguru]

you might also wanna mention that dvd2svcd it the only package-tool that its audio conversion process (ac3->mp2) is fully floating-pointable.

[DDogg]

...is fully floating-pointable.

Could you explain a bit more what that means and how it is a good thing. :)

[DSPguru]

sure !
i'll try not tot get to technical. in case i fail, feel free to translate :D.

let's think of pi. it's value is 3.141592653589793...
in computers we use memory space in order to save the values of our audio samples.
now, if we were to save the value of pi in our computer, ANY computer in the world would fail, 'cause the value is infinte.
so what we do is limit ourselves to a finite set of values, and a well-defined percision. for example, we'll take only 3.141 for pi, and 1.414 for sqrt(2), etc'..
the meaning of this rounding process is that we get ERRORS, 'cause the sample 3.1412 or 3.1413 or.. 3.1414 would all be treated as pi.
we won't be able to distiguish between those samples. the only thing we know is that the precision is 0.001.
now, think about the whole processes involves in BeSweet - decoding the ac3, normalizing, compressing the dynamic range, sample-rate conversion, frame-rate conversion, encoding to mp2. think about the ERRORS you gain in each process, and you'll find out that your accuracy drops to much less than 0.001.
this sort of fuck-up the quality :devil:.

so what can we do ? not much.. :), just start with a huge initial precision, so we can end up with a good one too.

most encoders use 16bit integers to represent the audio samples. this gives us precision of about 1e-5 (0.00001). in BeSweet i'm using 32bits floating-point to represent the audio samples. this gives us precision up to about 1e-85.

yes, floating-point arithmetics consumes more memory and more running time to caclucate, but since there's no intermediate-file in BeSweet, and since it's sooo fast comparing to the old ways, i decided to go all the way in order to gain higher quality.

judging by the feedbacks i get, this was a good decision :cool: .

Dg.

[DDogg]

Wow! For a numbers junky antennae head you actually did a great job explaining that in layman's terms :p :)