Ripping PORN movies?

[metrum_eng]

I would like to add some porn movies to my collection of encoded movies, but cannot seem to find specific titles at my local video stores. The titles I'm seeking are of the fetish nature and I can only find these titles at the online dvd rental web sites. My questions are #1, can anyone recommend some good dvd porn rental sites? And #2, if anyone rips porn dvd's, I would like to hear how the results are, because I once read that most porn dvd's are interlaced and in order to encode it properly you would need a de-interlace filter? Not that it would be hard or a problem (since I'm pretty much a pro at encoding) but I'm just curious to hear the results of someone who has already encoded some porn dvd rips.

I prefer divx myself, and I usually encode my movies to fit on at least 2 cd-r's so the quality is dvd quality. I use the divx 4.11 codec, 2 pass vbr, and GordianKnot as the shell program. I have about 90 movies that I encoded myself, perhaps if someone is interested in discussing encoding and specifically encoding porn movies you can contact me at: trinidad_hbo@yahoo.com, or simply reply to this message thread.

I also have some vcd movies but I prefer divx, however, I would be interested in both formats if they are the porn movies I am seeking lol
Sammy

[hakko504]

Can't help you with #1, but in my experience almost all porn is made with normal videocameras and therefor fully interlaced. Usually though I don't like to waste time on these encodes so I only use this script:LoadPlugin("mpeg2dec.dll")
MpegSource("clip.d2v")
Separatefields().SelectEven()
BicubicResize(384,288) # PAL, for NTSC use 320,240
Otherwise, I usually do a compression test with GKnot to select de-interlacing method: I don't like the blending effects of fielddeinterlace and smartdeinterlace so I always set the blend parameter to false. In these cases fielddeinterlace always gives the worst compression ratio! Most of the time GreedyHMA will produce both the most eyepleasing result but sometimes smart-deinterlace has been the better choise.

Interesting footnote : the last two adult movies I made were both widescreen, progressive PAL movies where the only thing I needed to do was use Telecide to fix switching field order. It looks like even this kind of film is starting to take full advantage of the DVD media at last,not only by cramming as much as possible onto one DVD.

Edit: The fast method should of course keep the original framerate by ony using half the fields. I forgot selectEven.

[Stabmaster-Arson]

Remeber if you make it vcd (352x240 or 352 x 288) then you do not have to deinterlace. Svcd will require deinterlace and so will x 480 size divx rip. Also do not force file or use any iVTC.

[diji1]

Hi there :),

I cannot help you with question no.1 - here in Australia it is ( more or less ) illegal to hire out xxx rated material but in so called "adult shops" it's pretty easy to hire "under the counter" ... very commonplace in fact ... i think the shop owners budget the yearly fines in as a business expense.

I have encoded 34 dvds of that nature so far, all PAL. hakko is right about field deinterlace and the compression comment ( terrible reults before decomb came out ... ), and yes, generally they all need deinterlacing for divx. i have found that Decomb's Telecide filter works better than anything else i have tried ( smart, field deinterlace and greedyHMA ), usually with default settings.

A lot of adult dvds have fairly bad production values. There is often a lot of noise that translates really badly in adult material, therefore i have always experimented with noise filters. Filtering if aplied too heavily makes human skin tones look very unnatural as u prolly know so i suggest using very light applications. Usually i'll use temp smoother (2,1) but occasionally ill need somethin else ... sometimes tempsoften2 by itself or in combination. Filtering also helps lessen the visible deinterlacing "blur" ( hey, telecide's the best i found but it's not perfect ).

I also thought i'd mention that for the majority of adult dvd's are not very compressible - handheld cams and bright lighting etc. - and they are also usually more than 1.5 hours at least So you may need more than 2 cds to achieve high quality encodes.

Final note - for my taste reducing res for quality when determing ur encoding resolution leads to better results.

[hakko504]

First I mostly agree to everything diji1 said, even though Don Graft himself said somewhere that pure interlaced material should not have pure telecide applied to it. In bad cases it can make a stuttering clip so therefor any of the pure de-interlacer will give better results.

Also 1.5h?: Maybe it's my choise of films, but average seems to be like 3h. Worst case: 6h on a single sided DVD.

And lastly: Some of the best stuff I've ever saw was produced in Australia...

[diji1]

Also 1.5h?: Maybe it's my choise of films, but average seems to be like 3h.
Well, ur right about the lengths :) ... 6 hours eh ? Now *that's* value for money!

re: Telecide - i remember reading somewhere that someone had good results using telecide on pure interlaced, but i wasn't convinced until i tried it. the results ( only on pr0n ) were out of sight of anything else i have tried.

[hakko504]

Originally posted by diji1

Well, ur right about the lengths :) ... 6 hours eh ? Now *that's* value for money!

Length: Yes.
Quality: No.

Originally posted by diji1
re: Telecide - i remember reading somewhere that someone had good results using telecide on pure interlaced, but i wasn't convinced until i tried it. the results ( only on pr0n ) were out of sight of anything else i have tried.
I did try telecide once with OK results, but at the time I put it down to the postprocessing, i.e. fielddeinterlace. Maybe it's time to do some serious testing on this subject again.

[manono]

Yeah-I also tried Telecide without Decimate once on an Interlaced source. It looked OK except for the slight jerkiness every so often. The algo is looking for field matches when there aren't any to be had, so it can't help but mess up.

[trbarry]

I first suggested processing porn using GreedyHMA() with auto reverse pulldown back in this thread (http://rilanparty.com/vbb/showthread.php?s=&threadid=11130&highlight=porn).

The idea is that there are many stationary scenes with fixed tripod mounted cameras and maybe very slow motion in parts of the scene, so pulldown-like field matching will give the most detail, at least on clean source. So auto pulldown in GreedyHMA can work, you just don't want to do the decimate part because it would create the 3:2 judder if you reduced the frame rate.

I assume Telecide could do similar things.

- Tom

[hakko504]

Since we are all striving for perfection :devil: and telecide seems to be the best field matcher and decimater whereas Greedy is the best de-interlacer maybe the following script could be really useful:Telecide(post=false) #Do not let telecide de-interlace what is left,
GreedyHMA(1,0,0,0,0,0,0,0) #let Greedy do that.
#Decimate(cycle=5) #uncheck for IVTC

Update: I made a quick compression test comparing the above script and using only Greedy as de-interlacer, and Greedy only had 82.8% and combined with Telecide I got 84.8%. Movie 1h56m NTSC-interlaced, only other filter a resize. As expected, the above combination took about 10% longer time (18m became 21m).

[trbarry]

Movie 1h56m NTSC-interlaced, only other filter a resize.

If you are also going to be downsizing later then you might (or not) get smoother deinterlaced results with the Greedy vertical filter turned on:

GreedyHMA(1,0,0,0,1,0,0,0)

But you can get it mostly for free, there's no significant performance penalty and it will probably compress slightly better.

- Tom

[hakko504]

Thanks Tom: Adding the Greedy vertical filter bumped up the compression ratio to 90.7% and it took about a minute longer to do the compression test. However I was doing some other things too with my computer so that would account for some of the speed loss. This is absolutely something to look out for in the future.

[angelfly]

In gknot 0.28 w/ avisynth 2.5, GreedyHMA is gone.
What would be the replacement?

Originally posted by hakko504
Telecide(post=false) #Do not let telecide de-interlace what is left,
GreedyHMA(1,0,0,0,0,0,0,0) #let Greedy do that.
#Decimate(cycle=5) #uncheck for IVTC

[hakko504]

Originally posted by angelfly
In gknot 0.28 w/ avisynth 2.5, GreedyHMA is gone.
What would be the replacement? The only native YV12 solutions are fielddeinterlace from decomb and TomsMoComp. They both have their advantages and disadvantages. Remember that if you use Telecide you should use the guide parameter to minimize jerky video.

[dani82]

http://dir.yahoo.com/Business_and_Economy/Shopping_and_Services/Sex/Adult_Video_and_DVD/Rental_By_Mail/ (you probably already found that link)

can't really recommend any of those since i'd never join any of them (really, i haven't), all i rip is anime and some movies; i just can't see myself ripping and watching porn for the next month or too, hentai is a different story.

[waldok]

What about acting in your own home porn instead of watching it on TV ? You could even tape it to DV and producing DVDs from it would be a breathe :D

I'm personally fed up by seing guys on the screen with a bigger hose than mine. My wife starts to ask questions :D

Waldok:cool:

[duartix]

Man,
Telecide, GreedyHMA, Decimate, Decomb, TomsMoComp, Separatefields,
SelectEven, :eek: :eek: :eek: WTF?

You are SUCH perfectionists with your image quality!!! :scared: :scared: :scared:
I always though that watching pr0n would make you go blind... :devil: :devil: :devil:
... or was it something else? :confused: :confused: :confused:

[hakko504]

Originally posted by duartix
You are SUCH perfectionists with your image quality!!! No, no, no, no, NO! We're just trying to cram as much filth onto a single CD (or DVD) as possible :D

[trbarry]

You are SUCH perfectionists with your image quality!!!
I always though that watching pr0n would make you go blind...

Watching pr0n only makes you go blind if you are careless on your encoding parms, especially deinterlacing. :devil:

And I'd now mostly recommend TomsMoComp(0,5,1) or TomsMoComp(1,5,1) for adult material, instead of GreedyHMA.

If the .d2v file is mostly 0's use (0,5,1). If it's mostly 2's then it's top first, use (1,5,1).

- Tom

[hakko504]

Currently I do almost all my pr0n with TMC(x,5,1) because it increases compressibility more than any other deinterlacer - and it's native A2.5/YV12 so it's fast too.

[wing1]

#1 location: NASA-> are we talking about US space program here? if u are, then you should not have any problem seeking porn materials at all in either states that you are located in. At least for me I can usually finding anykind of porn materials. The only thing I failed to find are old materials (classics) :D

#2 Why are porn being ripped at VCD resolution? can it not be done at full resolution? or is it not worth doing so? Most porn materials I found floating around the net are all in that formats. Once in a while I do see some amazing full res format, but they are rare. My personal opinion is: If you are going to do a rip, then treat it as equal to any rip that you have done. No materials should be treated different from another (especially from a DVD for that matter).

watchin porns does make me go blind :D It is not because of the contents, but rather because of all the freakin low grades FX that are being incorporated into them: Since when porns need FX?

[leadman584]

I have a bud who wanted to learn to use GKnot, just for porn. I hear he has seen more interlaced content in a month than I have in 2 years and 150+ encodes. He seems pretty happy with BOB for deinterlace. The videos he has shared for quality verification, are getting much better, and seem to handle this deinterlacing fine. The original DVD's are extremely noisy, probably because they try to smash 6 hours of video on some of the discs.

Why in the heck would anyone want 6 hours of porn? Why in the heck would they put it all on one DVD? I'm absolutely positive it is not for plot developement.

[bananix]

anyone have updates to this thread? :)

here's the info that dgindex shows of my dvd
NTSC 720x480 4:3 Interlaced field order is bottom, frame struct is frame, colorimetry is SMPTE 170M and the .d2v shows FILM 0.00%

i want to do same resolution xvid and i used TomsMoComp but it smoothens the image a lot and so too much details are lost. if i'd use Inverse Telecine option (i understand that then i have to change the fps to 23.976 in the gknot window and would be no sense to keep it at 29.970)

[Chainmax]

First of all, grab the latest Decomb and soak up on its readmes and tutorials. If the DVD is telecined (unlikely), use Decomb or TIVTC. If the DVD is interlaced (more likely) use TDeint or LeakKernelDeint. You might also want to use FFT3DFilter(sigma=1,bw=32,bh=32,bt=3,ow=16,oh=16) in order to clean up the noise or make it more compressible.

[bananix]

so i used dgindex with Honor Pulldown Flags open the .d2v (the .d2v file is here http://qget.shitforbrains.org/d2v.txt i don't know what numbers in it the guides tell to check in there to identify some method of telecine) in gknot and preview (gknot window) motion parts frame by frame but all i see is frames with lining so i can't use Forced Film i belive.

now if in gknot i press edit and only add these there and press preview

LoadPlugin("DGDecode.dll")
mpeg2source("clip.d2v")
AssumeTFF().SeparateFields()

and it jumps back and forward and with AssumeBFF.SeparateFields() it plays normal so it's BFF (if DGIndex shows field order bottom is it always as valid as this test?)

next i try with these and don't think there's any patterns (the guide tells to serve the clip raw so does this mean i need to use Ignore Pulldown Flags in DGIndex?)

LoadPlugin("DGDecode.dll")
LoadPlugin("decomb.dll")
mpeg2source("clip.d2v")
Telecide(order=0,guide=2)

next i change the Telecide line to this

Telecide(order=0,guide=1,post=0,show=true)

but can't see any metrics displayed on the frame...

aslo it shows it as 16:9 (or is that supposed to be viewed like that when inspecting?) in wmp 6.4 so i replaced it with bsplayer (mpc can't correctly go frame by frame) to be able view it with 4:3 aspect ratio. in order to do that i had to run this .bat file so windows doesn't automatically replace the mplayer2.exe when i try to replace/remove/rename it with copies found in the other paths. or perhaps i should use LanczosResize(640,480)...


copy /y "C:\Program Files\Windows Media Player\bsplayer.exe" "C:\WINDOWS\ServicePackFiles\i386\mplayer2.exe"
copy /y "C:\Program Files\Windows Media Player\bsplayer.exe" "C:\WINDOWS\system32\dllcache\mplayer2.exe"
copy /y "C:\Program Files\Windows Media Player\bsplayer.exe" "C:\Program Files\Windows Media Player\mplayer2.exe"

[bananix]

i tried FFT3DFilter and downloaded aslo required FFTW3 but i get error "fftw3.dll is not and AviSynth 2.5 plugin" can't it be used with latest gknot?

should i check IVTC in avs -> correct Frame Count. always when deinterlacing?

[buzzqw]

"fftw3.dll" isn't a avisynth plugin ! so there is no need to load it

put this dll into your windows/system32

BHH

[Chainmax]

Do not mistake Deinterlacing and Inverse Telecining, they are two very different processes.

When you create the d2v file with DGIndex, always use "Honor Pulldown Flags". If I'm not mistaken, only few DVDs with b0rked flags will benefit from the "ignore" option. Besides, I don't think this option affects interlaced videos anyway as pulldown is a sort of mock-up telecine, i.é: pulldown is applied to a FILM (23.976fps) stream in order to get the player to display it at 29.97fps.

I never used the Force FILM option as it's a very simple operation that might yield some jerkyness or leftover combing. Besides, like it says in the IVTC TUT link, DGIndex has a particular way of expressing things: some times it might say the source is Interlaced when it's perfectly IVTCable. If there's need for IVTC I just use TIVTC or DeComb.

If you use TDeint for deinterlacing or TIVTC for Inverse Telecining, you don't have to find the field order as these filters do it automatically. If you want to use DeComb (which is a fine choice as well), then what you describe is correct. I never used DGIndex's info on that regard as I don't know wether it's correct or not.

Serving the raw clip means that you only have to put LoadPlugin("DGDecode.dll")
mpeg2source("clip.d2v")
in your script. Remember that there is no pattern on an interlaced source. I don't know why no metrics are displayed, you should ask neuron2 (the creator) about that.

About Aspect Ratio, it's usually on the back of the DVD case. If you can't find it, try Amazon or similar online stores that sell DVDs, they usually have the Aspect Ratio somewhere in between the description of the contents.

[Guest]

next i change the Telecide line to this

Telecide(order=0,guide=1,post=0,show=true)

but can't see any metrics displayed on the frame...
Like I'm going to read a thread on porn?! :)

Set post=1 to enable metrics calculations.

It's almost certain that your flick is pure interlaced video. You do not need Telecide. Use a good deinterlacer if you want progressive output. You can try LeakKernelDeint(), TDeint(), FieldDeinterlace(), etc. I currently am very pleased with TDeint()+EEDI2(), but it's very slow and a bit tricky for naive users.