...while using old CRT sets. I usually set the sharpness control to 0, meaning, no sharpness added to the picture. I tend to do this in order to get the most accurate and cleanest picture. I hate how sharpness 'smears up' the image. Way I see it, the sharpness control isn't really needed anymore since most people use satellite, digital cable, DVD, or OTA digital and the reception interference which the sharpness control is designed to hide is non-existent in many cases.

Well, the problem is this: Many sets I've used (everything from A to Z) show a white 'ghosting' type artifact between text when I turn the sharpness all the way off. Its amazing, how I can put the sharpness control to halfway and the artifact completely disappears. This happens on any source. Originally saw this on an '83 Trinitron, and thought it was the CRT's oddball design, but no. It happened on my '79 System 3, my CTC-97, and a bunch of others.

Its almost as if the fine tuning is off a notch, but when its adjusted it has no effect on this problem. Any ideas?

Don't get me wrong, its not a big deal. I'm just interested in what exactly causes this?

PS: I don't think its a 75-ohm/RF versus composite versus S-Video issue, because on my WEGA it used to happen regardless of the input.

I've seen that too and never gave it much thought but now you have me curious too. I usually set mine to mid point or just a little below. Different story with "Peak Picture" on a Zenith, those usually look best at about 90%, some at max, depending on how sharp the CRT gets and how well tuned the chassis is.

...while using old CRT sets. I usually set the sharpness control to 0, meaning, no sharpness added to the picture. I tend to do this in order to get the most accurate and cleanest picture. I hate how sharpness 'smears up' the image. Way I see it, the sharpness control isn't really needed anymore since most people use satellite, digital cable, DVD, or OTA digital and the reception interference which the sharpness control is designed to hide is non-existent in many cases.

Well, the problem is this: Many sets I've used (everything from A to Z) show a white 'ghosting' type artifact between text when I turn the sharpness all the way off. Its amazing, how I can put the sharpness control to halfway and the artifact completely disappears. This happens on any source. Originally saw this on an '83 Trinitron, and thought it was the CRT's oddball design, but no. It happened on my '79 System 3, my CTC-97, and a bunch of others.

Its almost as if the fine tuning is off a notch, but when its adjusted it has no effect on this problem. Any ideas?

Don't get me wrong, its not a big deal. I'm just interested in what exactly causes this?

PS: I don't think its a 75-ohm/RF versus composite versus S-Video issue, because on my WEGA it used to happen regardless of the input.

The way you worded this makes it really hard to tell what you mean - answer some questions please:

1) when you say set to zero, you mean full counterclockwise on a rotary control, correct?
2) minimum sharpness would usually be described as most smeary, but you say max is "smeared" - what exactly do you mean?
3) what kind of text has the white "ghosting"? Black text on white background or the other way around?

Black text on a white background, and I mean clockwise---on a rotary pot, all the way clockwise, so, less sharpness or all the way off.

On a rotary pot, clockwise is maximum sharpness or peaking, all the way ON.

Here are a couple of pages to explain it. Figure 13-127 shows an ideal "aperture compensation" circuit of the type used in studio cameras and not in receivers due to the need for delay lines. The idea is to increase the apparent sharpness of edges and the contrast of small details by adding a pre-shoot and a post-shoot to every vertical edge. Thsi is horizontal peaking, which can be done with delay lines or RLC circuits. Really expensive circuits in studio cameras also would have vertical peaking (aperture correction) for horizontal edges. The really expensive plumbicon studio cameras even had different aperture correction for diagonal edges.

Getting back to the figures: Most TV sets had a sharpness control consisting of a RC circuit with a variable R in series with a C, connected to the cathode (or emitter in a transistor set) of the video output. This circuit can vary the amount of high frequency gain, but it only affects the smear or overshoot following the edge (that is, to the right of the edge). However, the phase response of the video IF amplifier could produce a permanent pre-shoot (to the left of the edge). So, it ws common to design the IF to produce some medium amount of preshoot, and tehn use the sharpness control to vary the overshoot. This is not ideal, but could compensate for viewer taste, antenna tilts, etc. to some degree.

If you look at the frequency response (left side of Fig 13-127, you see that there is a steep slope on the high frequency cutoff when the peaking is at max. If this slope becomes too steep, the transient response will not only have a pre-shoot and overshoot, but the waveform will ring, which will look like two or more preshoots and two or more overshoots. To the eye, these appear as extra edges or ghosts, which is what you see when you turn the sharpness control clockwise, especially if you also have the fine tuning adjusted to emphasize high frequencies in the picture.

Now take a look at Fig 13-129. This circuit was invented in a current-mode design by my (now deceased) boss Howard Jirka at Zenith. I modified it into the voltage-sensing form you see here because it used fewer IC pins that way. It has the advantage of being the simplest and cheapest possible approximation to the ideal delay line circuit, and produces both preshoots and overshoots. My colleagues at Zenith and I combined this into an analog IC that did the peaking, contrast control, and edge gating all in one circuit. Edge gating was done by sensing the first derivative that appears across the R in the RLC. By gating, peaking only worked on edges and did not enhance noise in flat areas. I don't recall for sure, but I think some sets had a customer control to turn the edge gate on or off. If the signal wasn't weak, you could disable the gating, and the peaking effect would work on texture in flat areas also.

The nominal setting of peaking (at midrange on the control) was a matter of taste. Different brands of sets had different degrees of it, and it varied over the years at each manufacturer as different personnel got to decide what they liked and thought would impress potential buyers on the sales floor. All sets had more than a moderate amount, such that black letters always had exaggerated white edges - but this made the pictures look really "sharp" (with high contrast for small details) when viewed on the sales floor. In sets with the older overshoot-only adjustment, the customer had to live with whatever preshoot had been designed into the IF response.

The amount of peaking needed for a sharp looking picture is reduced if you have more real definition, as in a high defintion TV picture, but designers still put some peaking in, even in HDTV. Videophiles generally complain that the factory setting is too high and reduce it or set it to minimum.

000

That reminds me of one of the worst features to come to CRT TVs: scan velocity modulation. Maybe a very small amount ...

You got that right - and you probably never saw the worst of it. I forget who first started touting it, but once it was advertised, Zenith had to look into including it. When cranked up too high, it could turn a mouth full of white teeth into sharp white needle-fangs! So, the amount was limited to prevent that, but in some sets it was still excessive.

Tom, you have been referring to the CCW (minimum peaking) setting as "sharper". It's clear (no pun intended) that you find the picture better at this setting. Do you think "clearer" would be a better word to describe it? "Sharper" has been used forever to mean the case when the edge waveforms have a shorter rise time and the detail contrast is increased, and "less sharp" or "smeary" for the opposite case.

I realized this morning that I have examples of both the simple RC sharpness control (1967 Magnavox, so-called 25 inch, rectangular) and the symmetrical RLC circuit (1996 Zenith, model SY2788DT, 27 inch table model) at home. The Zenith has the edge-gating circuit, which I reminnded myself was called "video filter" in the user menus. I looked at some test patterns and my menu settings on the Zenith table model. I also played the patterns from a different DVD player into my HD CRT rear projo through the component inputs. Reveiwing the control sttings, I found the "sharpness" on both Zeniths was set to about 1/4 range, which sharpens the edges a bit, but does not "light up" the edges or spaces between letters. I had the video filter OFF.

It's hard to look at the Magnavox right now; it's behind the work tables for the CTC-5 restoration. However, IIRC, the sharpness control has an excessive range, and I usually set it just up from minimum. (It's also not worth trying to look at this morning, as sunlight is streaming into the room).

I'm tempted to take a slew of pictures to document this, but that will be a real chore, as you need to look at both test patterns to clearly indicate how much peaking is present, and some representative normal video to make a final judgement about what setting is best.

Right now, I'm going to have breakfast, and then figure out what I really should be doing today!