6/19/2018

Expandable, Modular, Repairable - Component, HDMI video recorder

A mass produced brand name box is unlikely, S-Video was for SD and a product of its times. Component was early HD and replaced by HDMI. Capturing the SD or HD signal was entwined with a desire for timer or EPG driven schedulers, and multiple digitial Tuners or Cablecard slots.. dumbing that down to Component Input and timer or on-demand is one strategy but the cost and rarity of those connectors is being driven out by the HDMI single cable simplicity. Modern display devices more often have HDMI connectors so that has become the standard.

Up to HD 2K game capture latency has driven the demand for HDMI splitters, which sometimes didn't implement HDCP copy protection, so any HDMI card could be used to capture HD if caught off a splitter, but it wasn't by design and export controls actively work to find and drive those out of business as quickly as possible. Since HD 4K its my understanding that loop hole has been plugged.

Leaving Component input on legacy DVD recorders and some PC capture cards. The quality of legacy DVD recorders with Component input were poor and didn't really suit their purpose, SD video capture. It was overkill bandwdith wise and the cables costs more going from three cables for audio+composite to five cables for audio+component. Only Laser disc could really use the bandwidth and was a niche market.

And that comes to today and the home theatre pc market, also vanishing.. but more from archival apathy.. and an over abundance of trust in the cloud and belief that Copyright will be offset by Lifetime viewer rights.. tho.. if content owners could selectively erase human memories.. I think they would be overjoyed.

IMO.. magewell makes a very nice "tunerless" capture card that works with windows, linux and mac in PCIe and USB form.. it has a well defined built-in full frame TBC, DNR and Y/C comb filter with proc-amp.. full retail is around 300 usd.. ebay sometimes 100 usd. The drivers adopt the most popular api for each platform so it works with virtually any software. But being "tunerless" its not exactly on the typical home theater pc enthusiast radar.. its more "archivist" or content collector targeted.

What made the DVD recorder especially useful in my opinion was the remote, and simplicity of the task.. collect content, permit limited editing and burn to disc. Compressing and moving all those bits, even by ethernet was just too slow, and DVD-RAM never quite supplanted the write-once and done DVD-R backup.

Finding that simplicity on a pc is very difficult, unless you walk a fine line and don't try to complicate things.

The single simplest, familar interface on the pc for manipulating video content is Windows Media Center, deprecated in 2010 its increasingly hard to find.. so it is itself becoming "legacy". However, at least on Windows 7, until 2020 it is still under support and somewhat accessible.. Windows Media Center has a partner remote, and can record "live content" from a tunerless input card if it detects an RC6 WME ir blaster. these get cataloged into its library and can be added to a playlist and burned to DVD and in theory Blu-ray.

That's the theory anyway.. and I'm pursuing it as quickly as I can to confirm.

I really like DVD recorders.. some of the last ones are all linux based and have a lot of upgrade potential.. upgrading to HDMI or component input and blu-ray may be possible, someday, but their time post-burner phase has not yet come. They are too valuable as they are for the moment to the people spending a lot of money for them on the secondary market.

A lot of the lessons learned about VCRs with DNR, line TBC vs frame TBC and frame synchronizers, IRE, proc-amps and more are still applicable to an expandable, modular, repairable - Component or HDMI recorder.. doesn't help with the EPG or Tuner problems.. but for the archivist little is lost from a skills perspective.

ps. One thing to note about Component vs HDMI recording is that there is no known Copy Protection signal mitigation for false positives readily available. In the past Video Filters or something like a Grex could be used to silence the inaccurate signal degredation, whether on purpose or by accident or the result of noise.

" it is also beyond my knowledge to even know if the macrovision I, II signals that effect VBI effectively could be blocked because Components R,G,B is digitial and not analog.. however there are other levels of macrovision and CGMS flags as well now.. and Components digitizing chips recognize and honor these".

A popular method might be to use a Component to S-Video converter, that then runs through an S-Video Copy Protection mitigator, then back through an S-Video to Component converter.. but this reduces the value proposition of using Component by also causing picture quality degredation.

Also Component did not have a WSS or Wide Screen Signal "flag" procedural "standard" for advising the display device when a signal was being output in an anamorphic format (tall and skinny) that "should" be displayed on a widescreen display in an expanded pixel morphic aspect ratio. While one could be added "later" after capture, or with a specific "in-line" box for this purpose it was not "the norm" and complicated the use of Component out from sources or Set Top Boxes capable of outputing an anamorphic widescreen signal. In the beginning this wasn't much of an issue, but as DVD content became increasingly anamorphic and some cable channels would switch between 4:3 and 16:9 it has become an annoyance.

HDMI generally avoided most of the WSS problems by properly supporting it, and since the Copy Protection mitigation was a result of an oversight for lower resolution 2K signal splitter devices with low latency for game play and game recording.. temporarily at least.. HDMI has some advantages over Component recording.

6/07/2018

Waveform and Vectorscope, Bar signal on a Pedestal


I bought a Leader 5860c Waveform Monitor and 5850c Vectorscope from 1989 last weekend. Setting them up was a challenge.. this is that story.

The Waveform Monitor wasn't as much of a challenge.

Basically it has BNC composite inputs, and I had to get some adapters for my composite cables to convert them over and connect a VCR and a Time Base Corrector to its Input.

The Time Base Corrector could also serve up a 75% Color Bar signal.. which could produce the usual Stair Steps seen in so many old black and white photographs. That also let me find and recognize the dual side by side field 1 and field 2 "humps" and the full frontporch and backporch of each in the center. Along with the IRE (set-up) or Pedestal add signal that picks up the Black level in North American Video signals and "sets it" on a Pedestal just above the sync blanking level.

Even though I "sort of had guidance from a PDF manual" it was for the wrong vintage and kind of vague about terms and very short.

I learned I had to DC restore the signal to keep it from drifting up and down because the signal is by default coupled to the Input as an AC signal about a sync level used to represent the center point of the overall video signal.  The Monitor had a simple button for DC restore and focused the signal on its reference point.

I could then move the signal up and down and left and right with some alignment controls, and rotate the "horizontal level" of the scan using a small tool and a trimmer in the upper left hand of the monitors faceplate.

Scaling was automatic (or "Calibrated") or manual (or "Uncalibrated") when snapped into position the scale on the "graditule" represents the signal in terms of IRE units instead of voltages.

Good since most literature concerns itself with IRE units and not actual "voltage units".

I then put a SignVideo Proc-Amp into the signal path and played with its four controls

1. Black
2. Contrast
3. Saturation
4. Tint

The first two (Black and Contrast) allowed me to move the "floor" or blackest black level of the video signal relative to the "center point" sync refernce level. But that also had a slight effect on the top of the signal represented by the whitest "white" or brightest signal on the screen.

While the video signal on the monitor represents "Luma" or Brightness irrespective of Color.. each color bar has a declining "brightness" on purpose to create the stair steps. Left to right they fall off in perfect step with the bars on a normal video monitor.. but do not represent any color information.

This is exactly so that, the Black control only effects the overall video signal blackest black.

But after that adjusting the Contrast raises and lowers the top of the whitest or "brightest" color bar so that it could be set to IRE 100 .. or perhaps lower. IRE 75 is quoted as common, as are IRE 85 and 95 .. as a hedge against signal sources that may "overdrive" or "blow out" the perceived exposure.. loosing details in the "wash". This is called "clipping" and is to be avoided.

Clipping can also occur at the other end of the scale in the Blackest Black floor.. the goal is to keep tweaking to get most of the signal, most of the time to remain between these extremes.. which can depend upon the exact source used.. but the color bars serve as a first approximation and allow for some sand bagging of the range to protect against "clipping" at either extreme.

So the Waveform monitor is for calibrating or setting the "Black and the White" extremes of the signal using a proc-amp. And setting the Black also adjusts the height of the Pedestal for the Blackest black.. which in North America would be IRE 7.5 high (very important for the Vectorscope).

Next was the Vectorscope.

Its similarly easy to connect a video input signal, but it displays its results in a Polar or radial graph display. Magnitude is by Radius from the center, the other coordinate being an Angular value from a Color Burst reference signal.. not unlike the DC restore recovered "center sync" reference for the Waveform monitor..

And like that DC restoration.. the Vectorscope has to "recover" the Color Burst angle and decode the position of all colors from the signal arrayed in a circular fashion around the graditule or "scale" on the Vectorscope screen.

I made a mistake in seeking to set IRE to 0 for my video signal using a proc-amp to generate the Color bar signals. This caused the Vectorscope to "free wheel" or "spin" like a car drivers steering wheel.. or strobe like the struts on the wheels of a car. I couldn't get it to stop spinning, even using the phase angle adjustment control repreatedly.

Once I did try to switch IRE 7.5 (on) the bowtie pattern snapped on and stayed locked.

Also using a proc-amp as a color bar generator is not ideal.. in tiny fine print, it says you should also connect a video signal to the Input to the proc-amp composite input.. so that a "stable" color burst signal will be included with the color bars generated. This turned out to be true.

While acting as a bar generator the proc-amp cannot be used as a proc-amp, it locks all of its outputs to references.. presumably to act as a "standard" rather than a general purpose (much more expensive tool).

Radius of each bowtie, "spot" represents the relative "color saturation" for that color, as color video  has a familar pallete with the bar pattern, each bar creates one spot in the general vacinity of the graditule regions labled for their color. Angular offset from the color burst frequency determines their "color".

So a second proc-amp can manipulate radius by increasing or reducing "Saturation" and this effects the entire constellation and over all "size" of the Bowtie.

While the same second proc-amp can manipulate "angle" by increaing or reducing "Tint" and this "Turns" whe whole orientation of the Bowtie. The optimum goal being to adjust or tweak out common imperfections that lead to a "cast" or "overall" color problem that effects all colors equally.

Individual colors which require specific tweaks to Saturation or Tint requires the use of a "color generator" or "color corrector" but is independent of the video signal itself.. that would be a manipulation to grant false "color" enhancement and isn't strictly a result of the signal path or current video signal regenerator. That would be more akin to using a "paint brush" to touch up a moving picture as opposed to "fixing" a video signal to be within specifications for broadcast. And is usually something more common in film and telecine to add special effects, or add to a scene to enhance a particular emotion or psychological setting than a strictly physical situation.

So I did notice that the arrangement of the proc-amp controls from Left to Right were not arbitrary, as each from

1. Black
2. Contrast
3. Saturation
4. Tint

tends to progress from that a person would notice the "most" if left uncorrected to the one they would notice the "least".

So in this case Black offsets are noticed first, then Contrast problems, followed by Saturation problems and then Tint problems.

----+-- Black (bottom)
Y - | - ----- waveform monitor
----+-- Contrast (top)
      |
----+-- Saturation (radius)
C - | - ----- vectorscope
----+-- Tint (angle)