Removing overscan on bottom lines during capture?
 

Hello,

I am in the process of capturing a bunch of video8 and VHS tapes using huffyuv. All the tapes I have encountered have overscan lines at the bottom of the screens. I have been removing the lines during capture in the ways recommended on this website as outlined here using the resize filter to crop and size to 720x480:

https://www.digitalfaq.com/forum/vid...erly-crop.html

I personally find the lines slightly distracting when viewing back footage which is why I have removed them, but is there a downside to removing the overscan? Should I be doing this? Most of the time the videos will be viewed on a PC using VLC or equivalent, but if I want to re-encode the original huffyuv archives to a newer codec to view on a modern 1080p/4K/8K 16:9 TV someday down the line using DVD/bluray/or some other means (maybe streaming from PC), will not having the overscan lines mean the TV will cut off part of the actual useful image towards the bottom when viewed on a television? In other words, if overscan intentionally gets cut off during capture/edit, and you play it on a TV, do you see less useful picture (non-overscan) information from the bottom because the overscan is missing from the source while the TV might be expecting it?

On that note, if removing overscan (really any kind of edits including trimming) from huffyuv encoded files post capture using v-dub, is there any visual quality loss for removing those lines using the full processing mode options or direct stream copy for trimming, or are all these fully lossless operations?

Thanks in advance.

Overscan lines on the bottom is perfectly normal, it is best to mask them after capturing is done and you are onto the editing phase, not while capturing.

You see them because a capture card captures the whole frame, including parts of the image that wouldn't be visible on a TV. Computer screens don't have overscan, but TVs still do to this very day. You'd be able to see those lines on a TV if you were to change the overscan settings to display the full frame.

To answer your question, yes cutting them out first would result in the TV cutting off more parts of the image unless you change the settings to display the whole frame.

If you crop out the head switching noise from the bottom of the video, resize the video back to the original resolution, author it to a video DVD and play that on a DVD player connected to a modern digital TV, I think it probably removes the bottom few lines again. In that case, it's better to mask the head switching noise to black so when the TV removes the bottom lines, nothing is lost.

For VLC, Plex/Kodi etc., I think it's better to crop out the bottom lines so the video fills more of the screen when viewed in fullscreen mode.

So there you have two different answers for two different use cases.

With HuffYUV, edits are fully lossless.

I followed this how-to guide:

https://www.digitalfaq.com/forum/vid...erly-crop.html

How do I mask the overscan to black?

Masking is often an exercise of cropping (by 2, even numbers only), down to a non-standard resolution. Then padding it back to legal resolution by adding pixels. The extra benefit is that it centers the good portion of the image, as the overscan was generally offset.

That guide is correct, but we all use 720x480 now (lossless AVI era), not really 352x480 (MPEG era).

There are certain broadcast format converters that can mask any/all edges to whatever color you want (black in this case). The ones I have experience with are the Snell and Wilcox CVR series. While they are format converters, they still do their processing even if you keep the format NTSC->NTSC for example.

I think I have seen "live masking" in certain scalers as well like the DVDO iScan series (for capture, you'd tell them not to actually scale). The retrotink 5x and 4k likely can do live masking as well, but I haven't tried it myself.

None of these are "recommended" devices here, so you'll have to determine what is best for you.

To further complicate things, NTSC actually has 486 as a vertical resolution (well, two fields of 242.5 lines), so which 480 lines get captured by different cards varies unless you're capturing full frame 720x486 which I'm only aware of SDI being able to do using products like Blackmagic/AJA which also aren't recommended here haha.

That's why you'll see some capture cards show less head switching noise visible than others even when recording from the same source - the 480 lines they are capturing are higher up in the frame. There is no "standard" for which block of 480 lines are captured by different cards and it varies by manufacturer. None let you actually choose which vertical segment of 480 lines get captured to my knowledge.

Several versions of ATI MMC live masked. But the feature isn't worth using MMC over VirtualDub.