Selur's Little Message Board

Hello,

I'm having the common green strip issue on Sony PAL camcorders, in the right overscan region.

I understood that not much could be done hardware wise, it seems to be caused by an inbuilt chroma filter, the V channel is inverted.

So I searched for software solutions.

I was using Balanced borders / edge fixer, with relatively good results.

But I could still clearly see the filtering on the right when I watched my videos (because I knew it was there, and it was annoying).

I ended up cropping the right 12px and adjusting the format for peace of mind.

I just read a Hybrid chromashift script on the digitalfaq forums which seemed quite interesting, didn't know about the chromashift filter honestly:

Code:
import chromashift

import havsfunc

b = chromashift.ChromaShift(clip=clip, shift_right=14).std.Crop(left=694).std.AddBorders(left=694)

c = clip.std.Crop(right=26)

clip = havsfunc.Overlay(b,c,0,0,opacity=1)

It is from Selur which I believe is the Author of Hybrid himself.

However I get a havsfunc error when I try to preview. I did a quick search here and apparently havsfunc is no longer in use in Hybrid 2025.07.27.1

Is there a way to do it without using havsfunc, or is there another way of fixing this common issue that I'm not aware of?

I joined a small 720*576 PAL .avi lags file that I captured with a Hauppauge USBLive2 capture device in Virtualdub2 build 44282.

The green "chromastrip" is very strong with warm/yellowish artificial lights. Sometimes the strip goes purple when the video is blue (sky, sea...).

I'm not experienced at all, both in scripting and Hybrid, but I'm always learning and I think I can follow a step by step procedure.

Any help is highly appreciated thank you guys.

havsfunc isn't needed anymore, remove it
side note: chroma shift is not the right solution
instead, use:

Code:
clip = core.std.Crop(clip=clip, left=4, right=10, top=8, bottom=8)# cropping to 706x560

# Fix bright and dark line artifacts near the border of an image using BalanceBorders

clip = edge.bbmod(c=clip,cLeft=0,cTop=0,cRight=16,cBottom=0)

which can be easily done in Hybrid, by:
1. setting the crop values
2. moving BalanceBorders below the Crop filter (Vapoursynth->Misc->Filter Order/Queue)
3. configuring the BalanceBorders filter (Vapoursynth->Frame->Edge->Balanced Borders)

Cu Selur

Hello, Selur,

Thanks for the promp response.

Yes, that's about what I was doing. I cropped 26 right, directly in VS-frame-edge tab, after deinterlacing, without changing any order.

I'll do it your way.

I guess I'm a little obsessive about that. I can still see it but it's objectively a million times better, barely visible when considering the position and the entire frame.

I thought shifting the chroma would be unnoticeable, but when I think about it: what about right pans with the camera?

Thank you very much for your help, much appreciated.

GT

Shifting the chroma would shift the chroma of the whole image, which is wrong for your case since the chroma problem is just localized.
One could write a script which only shifts the chroma on the edge, but then you have something similar to balance borders and similar. Smile

You can also adjust the BalanceBorder settings (maybe also try cRight=18) and or apply it multiple times, or additionally use other filters.
=> simply play around a bit with it, but global chroma shifting is the wrong approach.

Cu Selur

From what I understand, it was meant to blend with an overlay so that the rest of the image is bypassed.

https://www.digitalfaq.com/forum/video-r...lored.html

The problem of objects coming from the right persists though.

Yeah it's more of an OCD problem from my side Smile

Balanced Borders is doing a great job.

I'll play around, depending on the footage.

Thank you,

GT

Ah okay, yes, you can use chroma shift and apply it only to a part of the image, but that only is useful on a few rare cases.
BalancedBorders&Co are usually the better approach.

Trying to fix an overscan issue that was not even an issue back in the days, for a few pixels out of a 30yo ccd sensor.

You know what, I'll crop it off old school style Big Grin

we have better things to do with our lives.

All the best, and thank you for Hybrid, it is awesome.

GT

GT_22

Selur

GT_22

Selur

GT_22

Selur

GT_22