My problem is not with future recordings.
These videos are recordings made with a Panasonic VZ9 camcorder.
My first camcorder also meant my first recordings, so the imperfection is present in those recordings.
However, these videos also document a very dear past.
Thanks for the tips. The grey card wasn’t there but I can fall back on the simple things as you rightly say.
I’m so used to watching tutorials with colour enhancements on high quality files that the results I got on my videos seemed poor. I’ll keep trying
1 Like
Wow, that was a morale booster. Maybe I’ll get lucky on the next attempt tomorrow.
Thank you very much for all the comments.
Once the videos are recorded at a certain quality, you are pretty much limited to how well it will look. You can make it look better, but not as good as those tutorials where they pre-plan the white balance and exposure. Some are just going to look like crap, but at least better than the original.
2 Likes
Given a VHS-C source, if the colors are dull or muted, try adding some Saturation once the White Balance is figured out. Or turn Saturation down if the colors look like giant patches of vibrant solid colors with no tonality or gradients. On the vectorscope, Saturation changes how far the spikes will extend from the center of the circle.
1 Like
I made a montage for visual comparison with various settings.
Above left is the original video.
To its right, I placed the following filters in order:
White Balance
Old film: Technocolor
Noise reduction: Wavelet
Sharpening.
Below left I tried a change of order and placed first the Technocolor filter and then the white balance (it doesn’t work any better).
Bottom right I did not use the Old Film Technocolor filter, instead, I used the Saturation filter.
Once I had reviewed the footage and compared the different scenes, I wanted to take some stills.
For this I disabled proxy and preview scaling.
Then I noticed that the video with the wavelet noise reduction filter worked differently between the old Technocolor film setting and the video with the saturation filter.
You can see how the inside of the cabinet has noise (Technocolor) and yet the wall and scale models have a better sharpness in that setting.
In motion, this is almost not visible, but I found it curious and here I comment on it.
Either way, the colors look more natural than in the original video. Thanks for the tips. 
1 Like
The NoiseReduction:Wavelet filter is a strange and wonderful tool.
Not only is the interaction with Sharpen a complex one, the interaction with a following Contrast filter is also complex.
I found that there are multiple “sweet spots” for the Contrast, each spot only 0.1% wide, between settings that yield horrible big blotches.
It must be an unavoidable effect of the algorithms and the digitization, so I am not complaining.
I simply point it out so that others will know that when using this combination, fine-tuning upwards past the “That looks terrible!” point can sometimes yield good results.
1 Like
Re: Skin Tones
If you look at the vectorscope, it has targets for the colours seen in a set of 75% and 100% colour bars (white and black would appear at the centre of the scope). The scope is achromatic at the centre and increases in saturation the further out you go.
There is then another, fainter, line with no targets on it. That is meant to match the average skin tone pigment under reference lighting (there are slight variations and it changes slightly with temperature).
If you have a close up of a person, there should be a spike along the line, if not, you can tell if the colour is too red or too blue. You can also use it to match saturation between shots either side of a cut.
1 Like
Thank you so much for sharing this information. I will definitely take a closer look at this.
I put this project on hold for a week because I wanted to also start a series of short videos on the practical use (with some examples) of Shotcut audio filters.
I have a lot of videos to improve and I don’t want to rush into doing a quick and bad job, so it may take me longer to do this than I had anticipated.
There seem to be many variables to understand this and it doesn’t seem like a simple thing.
1 Like
I’ve set up TV cameras hundreds of times over the years. We, and the high-end colorists I know, would never use a vectorscope to set flesh tones. There are too many variations in flesh tones for a vectorscope to be of much use.
If you have an achromatic object in the shot such as a neutral white or gray object, white balance on that using ShotCut’s white balance tool. Failing that, try adding a little blue and eyeball it. It looks like the camera was balanced for a bluer light source than the light the scene was shot with. First make sure your monitor is properly calibrated.
When my proverbial ship comes on, next on my list after a high-end Nvidia card that supports four or more monitors is an old used Conrac CRT monitor off of eBay, the calibratable standard for TV color for decades.
Sure, it maxes out at 640x480; I’ll mostly use it for TTY Terminal amber-on-black, but when I want to see my video in calibrated color, it will be there waiting.
Do you have a good color checker for your monitor? I use the XRite i1.
An old Conrac CRT monitor? Really? I don’t know what it could do that a modern monitor couldn’t. Will you have an analog signal to feed it?
It’s what I am familiar with, from back in the day, when we built the 48-channel PP1 “Mickey Mouse” console at Harrison Systems that replaced the “Fantasia” console at Disney Studios, and I wrote the software for the Input, Group and Equalizer modules. The automation and VU/PP meters were displayed on Conracs, because that is what was spec’d. We had no problem driving the Conracs with the controller card that we used, driven by a Zilog computer using digital interfaces I designed.
It’s been so long, I had to research what the inputs were on those old Conracs. However, whichever input system those used, Conrac did put an SVGA interface on some models, and those are what are on my eBay “Watch” list.
I have used X-Rite to set the color temperature of monitors at work, but I would hate to try to set gamma with it. The benefit of using a computer monitor for this purpose is that the computer can read the values returned by the sensor and adjust parameters such as gamma. What’s amazing is that for a couple of hundred bucks you can have a setup for calibrating monitors that would have been prohibitively expensive in days gone by.
I can understand the desire for a Conrac because it’s what you’re used to from back in the day, but times and the technology have changed and you may as well get with the modern stuff.
Personally, I use an LCD TV set rather than a monitor designed as a computer monitor. It has a brightness control which works like a brightness control is supposed to, so I have finer control over the black levels. I created a custom test pattern which has a proper pluge so I can set black levels according to SMPTE like we used to do on CRT monitors. If I go through the menu settings I can select the RF input and watch over-the-air TV signals.
https://www.bhphotovideo.com/c/product/798930-REG/X_Rite_EODIS3_i1Display_Pro.html?sts=pi&pim=Y
One data point you should know is that it’s a de-facto standard to set the white patch to 100 nits (100 cd/m2).
Thanks for the tip.
It’s been about two decades now since I was keeping up with video monitor calibration professional literature; at that time, any LCD was scoffed at as “uncalibratable, really” because the red, blue and green are not the same nanometer points on the photon spectrum as the “standard” CRT phosphors. “It is like comparing pineapples to durians” they would moan. “Just because you have six of each doesn’t mean you have the same fruit basket.”
LOL
Nice.
Back in the day, the recalibration sensor unit “knew where it was” on the screen by time-synchronizing the photo-diode spikes with the video horizontal traces. An expensive and finicky piece of equipment.
But out of the box, the Conrac was good enough for most work; it was factory calibrated, and so was within all but the tightest specs when you never broke the seals on the potentiometers.
ebay currently has a new, never used, one for $700…
Yes.
People are making YouTube videos that look great on an iPhone, and look like crap when you watch them on the big flatscreen in the comfort of your livingroom via FireStick.
Nice to know what it will look like on a real TV before publishing it.
Thanks for bringing me up to date with all this info, @chris319.
(My son is showing me how to do animations with a stack of index cards by flipping the corner of the stack. Why back in the day, we…
…and the audio over the string between the papercups…)
The problem is usually the other way around. iPhones are stunningly accurate to the DCI-P3 and sRGB standards. Proper transforms from BT.709 will look equally good. YouTube videos look awful on consumer televisions because it’s the uncalibrated TVs that have terrible color. That’s why every TV on display at Walmart has a different color picture even though they’re all playing the same video. But every iPhone in Walmart looks identical. They’re just that good. A properly calibrated studio monitor or decent home TV will of course not have this problem.
iPhone accuracy is plotted here if curious:
DisplayMate Reference Colors and iPhone 11 Pro Max Absolute Color Accuracy Plots
Also, the RGB color primaries were changed yet again for BT.2020, so I’m not sure the Conrac or its phosphors are a great long-term investment… 
Good to know.
And yet the on-air signal on all those Walmart TVs looks good enough to sell a TV, which means that video engineers at the big networks know what color balance to use to accommodate all of those out-of-calibration TVs reasonably well.
When I was working in the music industry in Nashville, every studio had very expensive perfect sound reproduction speakers by Klipsh or Harrison or a few others. But the most successful sound engineers and producers would bring their own speakers, typically four-inch RadioShacks mounted in coffee cans. Most top-40 listening was done in cars using the speakers Ford or Chevy put in stock, so if you wanted to go gold, your music had to sound good on those speakers, as well as on the top of the line.
Ah, the color primaries war. The old defenders must all be dead now; they never would have let that happen.
I’ve put my custom test pattern on YouTube. The colors have been turned to mincemeat. I don’t know if the problem lies with YouTube or my browser. I’d have to download the file and check it.
Having been one of those video engineers, we set up the video to look good on what else? A Conrac CRT monitor! Then Ikegami and Sony monitors came in, all CRT. The Ike’s had a probe to set them up.
That said, I strongly discourage you from using a Conrac CRT for any kind of critical video work in 2021. It’s not worth pouring money down that rathole. Get yourself an LCD TV set and a proper calibration program and sensor.
That said, I own an RCA 77-DX, hardly state of the art. It sounds good but for me it’s more an objet d’art. We all have our indulgences.
Here is what I have. It allows me to set proper black levels.
https://www.amazon.com/gp/product/B01N2Z17MS/ref=ppx_yo_dt_b_search_asin_image?ie=UTF8&psc=1