what is the max. resolution for an analogue captured source?

[Gawwad]

hi :)
i have this video capture card:
http://www.hauppauge.com/Pages/products/data_pvrusb2.html

and i just connected it to a nicam VCR using composite (red/white/yellow) to scart (http://en.wikipedia.org/wiki/Scart) cable
and i captured a vhs tape to MPEG-2 in the highest quality possible (12MBits/sec)
the captured video, has 720 x 576 resolution, and its full screen
now i want to turn this video to xvid/avi
so what is the maximum resolution that i can go for?
is 576 x 432 too much?
i just read in this thread:
Now, it is a perfect opportunity to clarify a myth here: There is no way to grabb from any analogue source (including TV broadcasts) and with the use of "S-Video in" to get a usefull resolution beyond 384 X 576/480.
http://forum.videohelp.com/viewtopic.php?t=98177

so i am a bit confused now :confused:
can somebody help me here please

thanx in advance

[evilkorn]

Not to sound rude or off topic, but why would you not just burn it to a dvd after authoring it. That way you don't have to worry about max quality, because you have a perfect copy of it.

Perhaps you plan to do that anyway, if so sorry for the rant. To my knowledge an avi file in the xvid codec can be any size. The reason it is in a lower resolution is to decrease the file size. So we really would need to know your target file size to answer your question, as well as the video length and frame rate, which I think you we can assume it's NTSC because of the original file size.

AutoGK is a pretty basic converter that uses the best, in my opinion, free conversion programs for the avi container. I would try that, set the resolution to what you wish, or just set it to auto and see if you like the results for that file size.

[Gawwad]

Not to sound rude or off topic, but why would you not just burn it to a dvd after authoring it. That way you don't have to worry about max quality, because you have a perfect copy of it.

Perhaps you plan to do that anyway, if so sorry for the rant. To my knowledge an avi file in the xvid codec can be any size. The reason it is in a lower resolution is to decrease the file size. So we really would need to know your target file size to answer your question, as well as the video length and frame rate, which I think you we can assume it's NTSC because of the original file size.

AutoGK is a pretty basic converter that uses the best, in my opinion, free conversion programs for the avi container. I would try that, set the resolution to what you wish, or just set it to auto and see if you like the results for that file size.

thanks for the reply :thanks:

but why would you not just burn it to a dvd after authoring it. That way you don't have to worry about max quality, because you have a perfect copy of it.
well, the captured video is about 50 minute long, and the size is 4.5 GB, and its pal
i just want to make it to a cd size (700mb), so i can easily share it with friends (its non-copyrighted stuff)
since the captured video is over 4.37 gb (dvd5 size), that means if i want to convert it to dvd5, i'll have to encode it
so instead of encoding to mpeg2, why not encode in mpeg4 (xvid)
plus i have a stand alone divx player at home :cool:

Perhaps you plan to do that anyway, if so sorry for the rant. To my knowledge an avi file in the xvid codec can be any size. The reason it is in a lower resolution is to decrease the file size. So we really would need to know your target file size to answer your question, as well as the video length and frame rate, which I think you we can assume it's NTSC because of the original file size.
i want to make it the best quality possible in a cd size (700mb)
I've read in "The XviD Releasing Standards 2005":
³ Resolution and Aspect Ratio: ³
³ - Width: 512 - 672 pixels for WS movies (Letterboxed is considered WS). ³
³ 448 - 576 pixels for FS movies (Only 4:3 images). ³

but now i found in this thread (http://forum.videohelp.com/viewtopic.php?t=98177):
Now, it is a perfect opportunity to clarify a myth here: There is no way to grabb from any analogue source (including TV broadcasts) and with the use of "S-Video in" to get a usefull resolution beyond 384 X 576/480.
so i just want to see what is the resolution that i should go for to have all the possible details of the original source

[Blue_MiSfit]

"The XviD Releasing Standards 2005"

Not sure WTF that is, but it sounds like some shady document from a pirate release group. Probably ill-advised to believe it.

If you captured at full PAL resolution (720x576) then just encode at that resolution. If you aren't happy with the results (unacceptable detail loss or blocking compared to the source), then reduce the resolution and or add some filters and re-encode.

~MiSfit

[prOnorama]

Not sure WTF that is, but it sounds like some shady document from a pirate release group. Probably ill-advised to believe it.

Actually I think those rules apply not just for 1 group but all scene groups.

Anyway I wouldn't be bothered too much by those standards, just use your own preferences. Or just use AutoGK which usually does a nice job :)

[foxyshadis]

S-VHS can be full resolution. It's very dependant on your camera/source, your recorder, and the quality of your line. (You don't have to pay top dollar for pedestrian monster cables, but having some shielding is always effective.) Analog TV typically lowpasses a great deal of high frequencies out of their broadcasts, usually to squeeze more channels into the same frequency range, which may be where that statement came from; so do some cameras. Direct output of a good camera, or a console or PC, will be sharp at 720, though the chroma might be a little fuzzy.

If the effective resolution really is below the encoded resolution, it won't lower compressibility that much. After all, a total lack of high frequencies means easier to compress, but the extra blocks and motion offset that. If you can't immediately tell and want to test, you can create a script that performs two resizes, one down (with spline36resize preferably) and one back up to fullscreen, and keep lowering the down until you can see a difference in sharpness. Even if you find out that it is lower than full, you might still want to keep it just to keep the 1:1 aspect, since dealing with AR can be a PITA at times.

Running a compressibility check in Enc is a more direct and useful way of testing, though, if you already have a bitrate in mind. That will tell you whether you need to resize or not, given certain quality constraints. See the xvid presets thread, xvid forum, for more information on this. (Sadly a few pages back, maybe I should sticky it.)

Releasing groups have decent standards - they have a lot of experience, after all, for better or worse - but they also have a habit of relying heavily on older tools, versions, and filters, as do some anime encoders, so it is a good idea to take it all with a grain of salt. Some hang out at this forum and pick up leet skills though. =p

Something that can never hurt - get higher capacity CDs. I've seen 800MB CDs, but they require special burners as well; more typically you can get 730MB on a 700MB CD. Those 30 megs can make a quality difference. You can find out the most you can get with Feurio (http://www.feurio.de), I guess Nero also has a tool to do it now.

Of course, DVDs are less than a penny each in many places (I haven't burned a non-music CD in years), so if you want a real quality boost with no sacrifices.... Since you would have to re-encode even if you chose a video DVD, because you'd need to encode to around 9500kbps (9000 if you care about wide compatibility) for DVD compatibility anyway, you could easily fit it at that maximum bitrate, with extra menus and all, in 4.3GB. For xvid you could use q2, mpeg or v3hr matrix, and still have room for extra material. Hey, up to you.

[LoRd_MuldeR]

I also capture at 720x576 from an analog source, it's composite though.
Since I need to crop the black borders (which differ form capture to capture) I can't encode at 720x576 anyway.
After cropping I usually end up with some ugly non-standard or non-mod16 resolution.
So resizing is mandatory. I usually resize to 640x480 for 4/3 stuff and 640x384 for 16/9 stuff.
That seems to be quite enough for this kind of capture...

[JohnnyMalaria]

http://forum.videohelp.com/viewtopic.php?p=271700#271700

Update
Now, it is a perfect opportunity to clarify a myth here: There is no way to grabb from any analogue source (including TV broadcasts) and with the use of "S-Video in" to get a usefull resolution beyond 384 X 576/480.
The reason for this, is that in analogue technology, there are no horizontal lines like the digital world, but horizontal DOTS. 720 dots in a raw, creates what we call a line. With the analogue transmissions, if a dot is missing in the raw, there gonna be no replacement, just a null dot, a nothing. The lenght of the raw, gonna be always the same with or not some dots.

That's got to be one of the most inaccurate, misleading statements I've read in a long time.

Horizontal lines and dots????

In the analog world, each scan line is a continous signal. The concept of 720 "dots" only comes into play when you digitize that analog signal. The D1 standard (as used by DV) dictates that the continous analog signal is sampled 720 times per scan line - giving you 720 pixels per line. You will get an image whose resolution is 720 x <# of lines> - that's resolution in computer graphics terms.

However, resolution in terms of detail that can be resolved is different. Sampling at 720 points per scan line means - according to the Nyquist theorem - that you can resolve detail equivalent to half the sampling rate. i.e., if you have image comprised of black and white vertical lines, the smallest lines that you can resolve would be one pixel wide - i.e., the scan line would be BWBWBWBWBWBWBWBWBW.... - the data repeat every other pixel (i.e, at half the sampling rate).

If you interpret the misguided advice as meaning that there's no point creating a file with a frame resolution greater than 360 x 576 (or 480), then you will be throwing out detail. (Note, if you are sampling at 768 points per scan line, then the value of 360 should be 384).

[Gawwad]

thank you sooo much guys for the very detailed info :thanks:
its very appreciated

now i just encoded the video to a 700 mb xvid/avi
http://img267.imageshack.us/img267/7486/clipboard01pw7.jpg

but i just want to check something, for the original captured video, GSpost gave this:
http://img300.imageshack.us/img300/8245/clipboard02vs7.jpg

it says that it is an interlaced video, so i just deinterlaced it, but now i found this in "The XviD Releasing Standards 2005 (http://upload.wikimedia.org/wikipedia/en/7/70/Tdx2005.gif)":
- In some cases PAL movies should be ivtc'd (i.e. to 24fps). Therefore using a PAL source is not an excuse for lack of ivtc.

can i ivtc my original captured video, or is it not possible (i.e. it will produce poor frame rate playback)?

thanx in advance

[foxyshadis]

It's very simple to determine that yourself. Start a script with nothing but the source and SeparateFields(). Look over it frame by frame; for interlaced, it'll move forward every frame, for progressive, every pair of frames will seem to match, and for telecined, it's similar to progressive but has occasional sets of 3 matching frames (the third and first will be identical). In NTSC, it's two progressive pairs to two telecined triplets; in PAL, it's usually 11 progressive then one triplet.

[Gawwad]

It's very simple to determine that yourself. Start a script with nothing but the source and SeparateFields(). Look over it frame by frame; for interlaced, it'll move forward every frame, for progressive, every pair of frames will seem to match, and for telecined, it's similar to progressive but has occasional sets of 3 matching frames (the third and first will be identical). In NTSC, it's two progressive pairs to two telecined triplets; in PAL, it's usually 11 progressive then one triplet.

sorry to trouble you again :o
but what should i exactly write in the avs script if my source video called "The Video.mpg"
and what program should i use to check the frames of the .avs script?

[BigDid]

Hi,

Already indicated but as it seems you are not familiar with avisynth scripts ... you may want to try AutoGK on your captures: http://forum.doom9.org/showthread.php?t=64266
A tutorial is included with the install and if you want a quick start, point your .mpg file as source, use 1cd (or 700Mb) as target size for a 2 pass, add job and fire the encode.

The nice thing with AutoGK is that the installation includes a few useful tools like
- xvid (codec)
- avisynth (for avs)
- virtualdubmod (editing and encoding gui) ...

which can be reused manually if you are willing to dig further :)

Did

[Gawwad]

Hi,

Already indicated but as it seems you are not familiar with avisynth scripts ... you may want to try AutoGK on your captures: http://forum.doom9.org/showthread.php?t=64266
A tutorial is included with the install and if you want a quick start, point your .mpg file as source, use 1cd (or 700Mb) as target size for a 2 pass, add job and fire the encode.

The nice thing with AutoGK is that the installation includes a few useful tools like
- xvid (codec)
- avisynth (for avs)
- virtualdubmod (editing and encoding gui) ...

which can be reused manually if you are willing to dig further :)

Did

thanks alot Did for the help :thanks:
i am actually using staxrip
http://forum.doom9.org/showthread.php?t=102652
and i am quite happy with it
i did encode that video (interlaced pal video) to 25fps 700 mb xvid and it was good quality
but when i tried again to encode the video with ivtc instead of deinterlace, it gave me a 20fps (i think) output that obviously had a jerky playback
now i just want to see if it is possible to encode the video to 24fps
because i read in "The XviD Releasing Standards 2005 (http://upload.wikimedia.org/wikipedia/en/7/70/Tdx2005.gif)"
- In some cases PAL movies should be ivtc'd (i.e. to 24fps). Therefore using a PAL source is not an excuse for lack of ivtc.

[jeffy]

sorry to trouble you again :o
but what should i exactly write in the avs script if my source video called "The Video.mpg"
and what program should i use to check the frames of the .avs script?
Find the file DGIndex.exe, located here, run it:
C:\Program Files\AutoGK\DGMPGDec\DGIndex.exe
Drag & Drop your file "The Video.mpg" on it, click File->Save Project. It will create the file "The Video.d2v".

Open a new text document in Notepad, write the following:

LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
mpeg2source("The Video.d2v")
SeparateFields()


and save it to the same folder, where the d2v file is, as "whatever.avs".

Find in the folder where you have installed AutoGK the application named VirtualDubMod.exe (in my case: C:\Program Files\AutoGK\VDubMod\VirtualDubMod.exe) and run it. Drag & drop the file "whatever.avs" onto it and do what foxyshadis suggested above.

[BigDid]

because i read in -
In some cases PAL movies should be ivtc'd (i.e. to 24fps). Therefore using a PAL source is not an excuse for lack of ivtc.
Hi,

From the AutoGK tutorial:
- "Force Normal IVTC" is something you'd better understand before using. Beginners can ignore it. Sometimes (rarely) AutoGK's Analysis step flubs on diagnosing the video content (and resulting AVI may be jerky or shutters a lot), and you might get better results overriding the decision that it's a hybrid, and go for a normal IVTC instead. Same goes for "Force Deinterlace" option (which can be a second choice when dealing with hybrid material)

Did