+ Reply to Thread
Results 31 to 53 of 53
Last edited by dellsam34; 23rd Mar 2019 at 23:36. Reason: miss spell
I swear, this is the crap that takes all the fun out of video editing. I think I'm going to just pay someone to do my project for me instead of wasting time reading a bunch of sh__t that I have no knowledge about.
I've almost never found editing to be fun. Restoring, sure. Conversion, sure. Watching, definitely. But editing is akin to a visit to the dentist's office (and you don't even get the consolation prize of having a cute dental assistant).
480 does not contain that, ever wonered why the 525 line NTSC was always 480 lines in digital realm? Well now you know, 480 is the image, the rest is the rest you metiones. Similarly 625 and 576 in PAL.
And it's lovely that people still know the tricks of analog capture, especially not to resize 720 to 640. And those additional pixels are called nominal analog blanking.
The ITU (cannot remember number) that specified digital SD television with backwards compatibility with analog specified the length of the scanline as 53,333... µs, in PAL the actual image is in 52 µs, in NTSC it is a fraction more, thus when sampled at standard 13,5MHz, the 53,3 µs results in 720 samples, while the active 52 µs comes as 702, similar to NTSC.
Thus, never trust people who use 1,067 PAR. Never trust VLC that will play 720x576 as 768x576, or worst of all 720x480 as 720x540 stretching vertically...
2. I was replying to the false statements about what is included in the 480 lines.
3. What that I have said was misleading?
4. I have already read a lot of forums, guides and articles ranging from the 80s to most recent, believe, that is NOT common knowledge for some people, if it is for you, then good for you.
For NTSC video images, the SMPTE 259M professional standard specifies that the 525 lines be represented as 720 × 486—that is, 720 horizontal pixels by 486 vertical pixels. This default video size is commonly known as D1. Capturing footage with most modern video capture cards from a professional BetaSP or Digital Betacam source results in a D1-sized frame. Capturing footage from a DV (digital video) source, however, yields a 720 × 480 frame. The difference between the D1 specification and the DV specification is only 6 vertical pixels.
The D1/DV NTSC and PAL specifications specify non-square pixels (often called D1 aspect ratio), while computer monitor pixels are square. In fact, DV/D1 NTSC pixels have a 0.91 pixel aspect ratio (PAR), which means they're tall and skinny; DV/D1 PAL pixels have a PAR of 1.09, so they're short and squat. For this reason, when you look at a D1 video image on a computer monitor, the images appear to be distorted (see Figure 1).
1. If one wants to create a perfectly legal, based on specs, digital ntsc file, the you'll need to use the horizontal resolution of 720 pixels.
Yes, you can play 640x### etc resolution files on any device that outputs to the tv or monitor that file full screen (by stretching or shrinking it to fit), but just because you can play a 240x resolution file full screen ntsc doesn't mean it's a legal to specs file.
2. 486 or 480?
Not much you can do about that.
dv to ntsc and you'll either have black lobes above and below, or a forced stretch.
Broadcast ntsc to dv and you'll either have lines cut off top and bottom, or a squish.
So, pick based on the expected output you'll use.
E.g.. If you're outputting to pc monitor, phone, and dvd, simply use 720x480 rather than 720x486 which is used for broadcast.
Yes, annoying when standards bodies can't decide on a standard ntsc resolution.
3. If you're capture from the digitizer is not a standard size, eg 640x480, it's not a standard resolution.
You can get software on the pc to display it correctly (force 4:3), but it is not the resolution you'll need for other devices.
E.g.. Drop that 640x480 onto a dvd and the software will either maintain 1:1 pixel relationship by centering it on black 720x480, or stretch pixels to make 640 pixels fit the 720 width.
IF the software is smart enough, the aspect ratio will be retained and all is good. If not, it'll look stretched and you'll have to process the video yourself beforehand.
4. 640x480 with black edges.
Not much you can do about overscan captured on these consumer devices. If you go vhs to dvd (on double drive recorders), or into dv/hdv (either two decks or into a camcorder with analog input), the edges are usually cropped and you'll get a full picture.
Not always since that's the nature of analog input formats like vhs.
5. If you crop but don't change the resolution by stretching, eg640x becomes 625x after cropping, you'll likely have to manually calculate the aspect ratio for the video player to display it correctly.
The video player may be very confused - is this a 1:1 ratio video file? 4:3 but you're feeding me fewer pixels than expected?? Etc
6. Ideally, you can adjust the size of the input to fully fit without black and the resolution of the capture (720x480), but if you've got 640x480, black, then you'll need to feed through virtualdub, crop, resize to 720x480 and embed 4:3 aspect ratio flag to be absolutely "proper".
Unfortunately, the resize does alter the quality, so this may be bad for your desired final output.
7. The solution is to recapture everything to a device that outputs 720x and where it either properly crops black entirely, or you can adjust the overscan area manually (broadcast equipment).
Then, you'll get what you want with good 1:1 pixel capture and quality from the original with no further stretching or cropping to mess up the image.
Such devices include the Canon hv10 and newer HDV tape camcorders that have analog input, or better, the Canopus avcd 100, 110, 300....
Or dual deck dvd vhs recorder like the Panasonic ez47, then rip dvds to pc into mpeg-2 files.
Hdv file format > dv file format > mpeg-2/dvd file format for capture quality from vhs tapes.
For your 640x480 captures, leaving it as is and merely making sure it has the proper display ratio flag is the least destructive way to maintain the original source quality. At this resolution level, stretching pixels through resampling isn't good at all - not enough pixels to make anything look good, only worse.
Last edited by babygdav; 8th Jan 2020 at 21:50.
The 480 lines contain active image only, no "head switch lines, Vertical burst, Caption, Macrovision ...etc", which I believe I have said then and there.
What is DVD section of this site.
NTSC (NTSC Film)
Up to 9.8 Mbit/s* (9800 Kbit/s*) MPEG2 video
Up to 1.856 Mbit/s (1856 Kbit/s) MPEG1 video
720 x 480 pixels MPEG2 (Called Full-D1)
704 x 480 pixels MPEG2
352 x 480 pixels MPEG2 (Called Half-D1, same as the CVD Standard)
352 x 240 pixels MPEG2
352 x 240 pixels MPEG1 (Same as the VCD Standard)
23,976 fps with 3:2 pulldown = 29,97 playback fps (NTSC Film, this is only supported by MPEG2 video)
16:9 Anamorphic (only supported by 720x480)
When I talk about NTSC, that is broadcast (over-the-air) TV standard resolution.
DVD standard is not a NTSC broadcast standard, so it specifies and allows many more resolutions, including 704x and lower.
A NTSC DVD is merely mastered for correct playback on NTSC TV sets and players, not for NTSC broadcast.
I'm talking about NTSC, broadcast-safe (ready/ formatted correctly / to specs) resolution:
Thus, 720x (although it does allow 640x if you take out overscan information, etc).
Now, analog broadcasts don't have pixels - they are continuous analog signals. There is the concept of scan lines, but no concept of an individual pixel.
Rec. 601 defined the digital "equivalent" of such worldwide, and it's been the standard in use for the transition of standard definition analog to digital.
Again, 720x is specified and is compliant with both the NTSC and Rec. 601 standards.
Pretty much all SD devices in broadcast use 720x, so while 640x might be legal, it's not the de facto standard.
740 vs 704 = 5% less data/picture information/detail recorded.
740 vs 640 = 15% less detail recorded.
While you can certainly capture VHS tapes to 640x, why if you can use 720x? You gain an additional 15% in detail (roughly since it's not exactly more pixels equals more details), so it's a better analog to digital conversion.
DV? Sure (but NTSC only).
Pretty much anything else? Nope.
E.g.. If you're outputting to pc monitor, phone, and dvd, simply use 720x480
3. If you're capture from the digitizer is not a standard size, eg 640x480, it's not a standard resolution.
E.g.. Drop that 640x480 onto a dvd and the software
What exactly is a "DVD" in this context. Just data on a disc? DVD-Video?
... and I just quit reading. Nothing in your post makes a lick of sense.
FYI, that's the same sideways logic that, 20 years ago, had people parroting the myth that "VCD = VHS quality", all because VHS has "240 lines" and VCD was x240 (NTSC). What's even dumber is that those numbers reference different axis.
FYI, that's the same sideways logic that, 20 years ago, had people parroting the myth that "VCD = VHS quality", all because VHS has "240 lines" and VCD was x240 (NTSC). What's even dumber is that those numbers reference different axis.[/QUOTE]
Yeah, more or less, that's rather my thought abbreviation reffering to the active image in the scan lines. I might have incorrectly used the specific terms reffering to analog lines and digital resolution. And I have never considered VCD to be an equal of the VHS, VCD's best use is as a beermat, I'd rather be watching and rewinding VHS.
Maybe video capture cards are not alike:
The technical definition of "active video area" doesn't exclude head-switching noise as you've marked.
Now the "usable and watchable" part of the image is something else entirely.
Head-switching noise is two-fold. The playback VCR, and the source tape. The noise is an interaction between these two things.
This is why some tapes are excessive, and others not, even on the same VCR.
Yes, you can reduce it in the VCR, but it's often simply a matter of calibrating the alignment and heads. There is no "proper equipment" to speak of, unless you're just referring to using another VCR entirely.
The noise profile is different not just brand to brand, or model to model, but unit to unit of the same brand/model.
And overscan is part of the active picture.
It's in the overscan anyway. Mask and move on. It's not worth discussing unless size exceeds the overscan area.
The best results I have seen were 2-3 lines of noise and the most bottom part of the image was as stable as the main image. If one has lots of spare time and patience it's not impossible to align them maually for each frame.
This is an example of what I squeezed out of ordinary VHS-C:
2 rows of pixels make up the whole head-swithing noise. And with some deinterlacing it disappears most of the time.
But the main reason it might look better is because it's PAL.
And I completely did not notice this:
If you don't bother, then don't bother even discussing.
Show me please an example, where I "just talk people into submission".
Last edited by SF01; 10th Jan 2020 at 09:05.