Improving Betacam video resolution

[carlmart]

Yeah, but even with the levels adjusted, that looks like crap compared to some BCSP footage I've personally worked with (sorry, not to disparage your stuff...). BC really should be better than that on average. Are you sure about the generations/lineage of this? And just to clarify, this isn't Betamax, right? 'Cuz if it was, then it would be in line with what I was expecting. Even so, there is a WHOLE lot of grain/noise on that hillside shot. Low level, gain bumped up?

Would there even be any way of (you) VIEWING the Beta masters that these DVs were Xferred from (to compare)?

I thought it was a joke about this being Betamax, that's why I didn't answer that before. It was Betacam, shot with an Ikegami camera. We never had Betamax in Brasil, only VHS.

The hillside was noisier than I expected it to be. It was very early in the morning, and I probably boosted the camera gain.

Unfortunately it's a long time since I don't view the original Betacam tape, which was the original edit. It shouldn't be more than one or two generations away, at most, from the original shooting tapes.

On the studio editing monitors, probably Sony, the image looked nice. It wouldn't pass my present standards, of course. That's why I'm discussing it here.

[filmboss80]

At this point, we have to suspect that something went terribly awry in the Betacam-to-MiniDV transfer. Ikegami cameras were about the best you could do back then, and the quality loss of 2nd generation Betacam was nominal. I would tend to rule out tape aging, as it would manifest other problems.

And no, we were not joking about Betamax. That is what the image looked like, and we were just trying to eliminate any possible confusion.

[edDV]

A simple 10 IRE (percent) black level shift down gets this.

>

Here is a simple before-after HSL (hue saturation luminance) adjustment.

>

You should make these adjustments while watching a calibrated broadcast monitor display.

Thanks for the suggestion, but the idea on color hue was just to be yellowish, almost like candle-light. The lighter areas are also too bright.

Rough adjustments were

1. -10% black stretch
2. -50% Chroma saturation
3. +19% Luminance Gain

After that hue was ok or only needed a slight adjustment. A full color corrector would allow separate color balance settings for dark, mid tone and bright areas.

My best guess is during capture to DV format, blanking went to digital luma 16 instead of setup causing black to go above digital 32. Digital conversion should place black at level 16 and 100 IRE luma white at level 235. The analog levels used would have been appropriate for European PAL.

I can't guess why chroma saturation was so hot.

[jagabo]

I fiddled with it a little bit yesterday but deleted everything. As I recall, ColorYUV(off_y=-24, gain_y=16) got the levels close to right. But that was a compromise because the outdoor and indoor shots had different levels. I don't know what you want to do with the color.

[carlmart]

I wonder why this topic stopped in black or color adjustments, when I said what I wanted more to correct are those "fading repeated vertical lines" (I don't know how that is called) to the left of most vertical objects.

How can I improve on that?

[jagabo]

Those are ringing artifacts from a sharpening filter. Try disabling any sharpening in the tape player. But it appears you have several tape-to-tape generations of this so simply fixing the last one won't help much.

A few Blur(1.0, 0) filters in AviSynth removes most of it but blurs the image a bit too.

Before:


After:

[edDV]

If at all possible, you need to go back to the original tape and recapture. There are a few generations of loss.

[poisondeathray]

yes, recapture if you can - the quality is udderly poor

[jagabo]

Manure really milking that one.

[poisondeathray]

just a minor point, but you could mooove the picture left a bit (left black border is larger than the right, so crop them and add them back correctly)

it might be hard to recapture , since the equipment has probably been put out to pasture



you could try other halo removers too, like yahr(), dehalo_alpha(), blind_dehalo3 . dehalo_alpha has a x and y strength rx, ry you can adjust . You will have to apply them on separate fields to keep the interlacing