Thanks, this worked. I ended up only brightening by 15 instead of 40 because the banding artifacts were too much, like I thought.
Wait, so one pixel on the waveform represents an entire horizontal line or one horizontal pixel? I can't imagine the latter because the waveform is only a fraction of the 960 pixels the picture is. And what does the intensity of the waveform represent if the x,y is the brightness? I see bright and dark purple in the right waveform.A waveform monitor of the luma channel is basically a graph of the brightness of each pixel. Consider one horizontal line of pixels. Some are brighter, some are darker. Instead of showing the brightness of each pixel as brightness, transform that line of pixels to a graph where the y axis represents the brightness of the pixel (0 at the bottom, 255 at the top), the x axis the position of the pixel across the line.
Ok so one waveform pixel = one entire line from pic. x,y position is the average brightness. I still don't get why the waveform has a varying brightness then.Now draw the same graph for every scan line of the image, all drawn on top of each other. In the final graph bright areas indicate many pixels had that brightness in that column. Darker areas indicate fewer pixels had that brightness. Black areas indicate no pixels had that brightness.
+ Reply to Thread
Results 61 to 73 of 73
-
-
Some pseudo code will probably make it clearer to you:
Code:uchar image[height][width]; # a luma only image ulong graph[256][width]; # the graph we're building for (y=0; y<height; y++) # for each scan line { for (x=0; x<wdith; x++) # for each pixel of the scanline { uchar brightness = image[y][x]; # brightness of the pixel graph[brightness][x]++; # brightness transformed to height in the graph } }
Last edited by jagabo; 8th Jun 2013 at 08:20.
-
Nope, that's complete Greek to me.
Btw, there's a part filmed in almost pitch black.
http://www.sendspace.com/file/fxfjt5
The quality is so bad that I can't tell if that's the guy or girl, but this is my best restoration attempt.
http://www.sendspace.com/file/jsnmex
What do you all think?Last edited by Mephesto; 7th Jun 2013 at 19:56.
-
Sorry, I thought you some programming.
Did you understand anything I said in post 59? Did you look at the post it linked to? It's actually pretty trivial once you "get it".
Did you understand the part about transforming the brightness of each pixel to a height in the graph? -
Hahaha it's a guy! or a manly woman ... Short hair, guy shorts, triceps shape and "relative" musculature . So you can close the thread now
It's a good clean up, but looks like heavy neat video treatment -
First part that loses me is the ++, what is that?
Did you understand anything I said in post 59? Did you look at the post it linked to? It's actually pretty trivial once you "get it".
I think I get it. Don't see much use for this though. It's confusing and hardly more resourceful than a simple histogram.
Hahaha it's a guy! or a manly woman ... Short hair, guy shorts, triceps shape and "relative" musculature . So you can close the thread now
I didn't think it would make sense for the GIRL to film her partner sleeping, but then again I've noticed women including mine like to observe their sleeping partner. I never understood why. They really are weird sometimes.
It's a good clean up, but looks like heavy neat video treatment -
-
That clip is the entire scene. The first hour looks like the first screenshot I posted, timestamp puts it at 10 PM to 11 PM then it cuts off to 2:30 AM where the guy fell asleep and the girl decided to film him without turning on the night vision on the cam for a few seconds, then the clip ends.
By why do you assume it's a GIRL filming HER partner ? Maybe it's a BOY filming HIS partner
You can be CSI Mephisto -
That is a vector scope plot of U and V. Ie. each pixel has a U and V value, it is plotted on that graph as U horizontally, V vertically. The center of the graph represents low saturation (grey), the edges high saturation of the different hues. A strong line from the center toward an edge indicates all pixels had the same hue. Watch RotatingHue.MKV in post #60.
-
Okay I played around with that Animate() function for brightness, contrast and all kinds of settings while watching videoscope. I get how it is now. It would be more convenient if there was a Vdub filter as good as videoscope so I wouldn't have to keep re-opening AVS scripts in Vdub.
While this thread is still open, I gotta ask if anyone has experience in color restoration? If hue AND saturation is lost in greyscale, how do they make those colorized films look so natural? Adding hue to a region is easy, but saturation is so diverse that it never looks right to pick a fixed value. Anyone know what they do? -
-
I didn't even know about that. This will make things a lot easier but it's still annoying having to edit the script, erase numbers, retype new ones, save script and press F2 when Vdub has filters with sliders and preview windows where I have more direct, convenient control. The animate feature in avisynth eases that up a bit though, but then you have to figure out in which order the settings of a filter come.
Similar Threads
-
Strange Hue problem
By dan59 in forum Capturing and VCRReplies: 26Last Post: 16th Nov 2012, 02:41 -
How to adjust Hue of vob file ?
By vidyarathne in forum Newbie / General discussionsReplies: 3Last Post: 11th Sep 2012, 14:07 -
Hauppauge PVR-150 Hue and Saturation adjustment
By Spektre in forum Capturing and VCRReplies: 3Last Post: 15th Jun 2011, 10:09 -
Pink hue on TV
By uniks in forum Newbie / General discussionsReplies: 4Last Post: 5th Apr 2011, 06:00 -
fluctuation of saturation and hue levels
By mark23 in forum RestorationReplies: 0Last Post: 1st Feb 2011, 17:12