The reasoning behind aspect ratios?

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Steve Anderson, Dave Moll, Andrew Davie

The reasoning behind aspect ratios?

Postby ac7zl » Thu May 10, 2007 4:34 am

Folks,

I'm slowly gearing up to produce some Nipkow wheels to experiment with. I've gathered the materials, and collected/fabricated the tools I'll need to lay out the hole patterns.

As I prepare to drill holes, I can't help but wonder about the reasoning behind aspect ratios, both those chosen by Baird, and the standard chosen by the club. My thoughts are as follows:

The 30-line standard applied to a Nipkow wheel results in pie-shaped scanning segments with an apex angle of only 12 degrees. The 32 line standard results in even slimmer pie slices, with apex angles of 11.25 degrees. There isn't a whole lot of area within those triangular slices to which a rectangular frame aperature can be applied. The most efficient application of frame to the narrow triangular scanning segment is a rectangle that is comparatively long in one dimension (corresponding to the scanning segment's radial dimension) and short in the other dimension (corresponding to the scanning segment's chord dimension.)

If the requirement is to scan images from a position to the right of the wheel's axis (the 3 o'clock position), it follows that the best aspect ratio of the frame should be of the "landscape" variety, that is, wider than tall. This makes most efficient use of the limited area within the scanning "pie slice." Yet, for some reason, the Baird standard was the opposite...3(wide) to 7(tall). Even the club standard seems counter-intuitive at 2(wide) to 3(tall).

Stated another way, if the goal was to produce "portrait" images, it seems to me it would have been far better to place the frame aperature and optics at the 12 o'clock position on the Nipkow wheel.

Any ideas as to the reasoning/logic behind the existing standards? Am I misunderstanding/overlooking something here?

Pete
AC7ZL
User avatar
ac7zl
Mad Scientist
 
Posts: 62
Joined: Thu Feb 01, 2007 5:04 am
Location: Tucson, Arizona, U.S.A.

Aspect ratios and scanning directions.

Postby Stephen » Thu May 10, 2007 5:36 am

There are really two issues here, one being the best aspect ratio and the other scanning direction. With a very limited number of pixels available, it makes sense to use a "portrait" rather than a "landscape" picture orientation. With the portrait orientation the picture has rather good vertical resolution but a rather narrow view. However, simple panning of the camera makes up for the narrow view. It is natural to pan a camera from side to side to capture a scene, but not up and down! I like to think of the 30 by 70 pixel image as being one half of a 60 by 70 image, which is pretty respectable.

So the portrait orientation makes sense. Now as to moving the scanning area from the three o'clock to the 12 o'clock position, there would be two issues. The first would be that the scanning would be horizontal instead of vertical. This means that one would have to drill 70 accurately placed apertures instead of 30. The second is the desirability of horizontal scanning. I think that Klaas mentioned in another thread on the forum that vertical scanning has some advantages with respect to reducing visible flicker at low scan rates. I think that these two factors make the case for the placement of the scanning area at the three o'clock position.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Postby DrZarkov » Thu May 10, 2007 6:42 am

From historical view it is interesting, that countries like the UK and the Netherlands used the 30 line system much longer than for example Germany or the USA, where they very fast tried to improve the quality of the picture, even in mechanical TV. I think both, Germans and Americans thought about transmitting movies, another reason for landscape format. But in real live NBTV is not really suitable for that, you can better used special produced TV-programmes. Then 30 lines was enough, until a really good system was available. I've converted the same movie (Dr. Who, an old episode with Patrick Troughton) into 32 lines club standard vertical and landscape scanning. The portrait format is looking a lot better! Much more details, and even on non-portrait pictures (parts of a landscape, or groups of people) are looking better. O.k., 3:7 does not seem very useful to me, that's not portrait, that's person.
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1037
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Re: The reasoning behind aspect ratios?

Postby aa9dt » Sun May 20, 2007 4:28 am

<h3>Scanning the short dimension</h3>
ac7zl wrote:The most efficient application of frame to the narrow triangular scanning segment is a rectangle... long in one dimension ... and short in the other dimension...

...the best aspect ratio... should be of the "landscape" variety, that is, wider than tall. ...Yet, for some reason, the Baird standard was the opposite.
Pete
AC7ZL


<b>ADVANTAGES.</b> Pete, when I drilled my first scanning discs in 1977, I had the same idea. Since then, this system has produced vertically scanned, landscape pictures. Here are some advantages of scanning the short picture dimension...
<ul>
<li>Small disc size: In my case, 24-line pictures from just a nine-inch disc! (To scan the long dimension, of course the disc must be larger.)</li>
<li>Picture size: I take and reproduce pictures an inch and a half across.</li>
<li>The scanner is small.</li>
<li>The motor can be fairly small.</li>
<li>Compared to Baird's method, hole size can be larger in relation to the disc.</li>
<li>Reduced windage.</li>
<li>Reduced heat.</li>
</ul>

<b>DISADVANTAGES.</b> There is no free lunch. I can imagine disadvantages of scanning the short dimension...

<ul>
<li>Short-dimension scanning discs allow more line breaks than long-dimension discs allow. (True only for rectangular pictures.) This idea is hard to convey. The point is data loss. That is, the space between lines is lost data. The ideal is a picture without line breaks. Of course, Nipkow scanning makes such a picture impractical at best.</li>

<li>Small discs exaggerate the keystoning problem.</li>

<li>Keystoning reduces the definition of the picture's shorter side. Small discs tend to worsen the problem. </li>

<li>The short dimension has lower definition than the long dimension has. Let's consider 30-line, short-dimension scanning. In that case, 30 is the maximum number of lines in the long dimension. The short dimension has some 12.8 lines.</li>

<li>Speed inaccuracy (hunting) causes the picture to "dance." Traditionally, Britons thought that this problem was worse in the horizontal direction. (They scanned vertically.) Americans and Germans thought that this problem was worse in the vertical dimension. (They scanned horizontally.) The consensus: In any direction, speed inaccuracy is bad.</li>
</ul>

<b>SHORT DIMENSION SCANNING & NBTVA.</b> I've read about early standard meetings of the NBTVA. In those meetings, a few mechanical TV hobbyists proposed scanning the short dimension. The club chose instead to follow Baird more closely.

Jim AA9DT
aa9dt
Laboratory Assistant
 
Posts: 24
Joined: Wed May 09, 2007 10:09 am

Postby aa9dt » Sun May 20, 2007 5:19 am

<h3>TV Movies: </h3>
<h3>Not Reason for 4:3 Aspect Ratio</h3>
DrZarkov wrote:I think both, Germans and Americans thought about transmitting movies, another reason for landscape format. But in real live NBTV is not really suitable for that, you can better used special produced TV-programmes.


In design, the use of horizontal rectangles isn't a recent invention. This invention dates to Greek times. Television for entertainment purposes is the true reason for the 4:3 aspect ratio.

A portrait shot limits you to one person. Yet television entertainment is about relationships. A portrait makes a simple two-shot very difficult. Baird's 3:7 aspect radio was even more extreme. I don't know anyone who has a face <i>that narrow!</i>

Baird designed for the narrow bandwidth that he started with, medium wave radio. In this regard, American and German television were more versatile inventions. From the beginning, these systems used a broader aspect ratio. Here in the US, most TV pictures were either square or horizontal rectangles.

For example, stations W2XCR and W2XBS transmitted pictures 72 lines wide by 60 lines high. These proportions represent a six-by-five aspect ratio. Such wide pictures allow for the continuity of two-shots and even scenery.

Here in the US, TV soon jumped to the shortwave 160-meter band. This band allowed for up to 100 kHz-wide picture signals. Engineers soon increased the definition to 60 lines. A 60-line picture can easily resolve images of two people. I've even seen a recognizable 60-line picture of a baseball diamond. Think of the advantages for all sorts of programming!

Meanwhile, the Baird system became a local standard with no competitors. The standard locked in the unfortunate 3:7 aspect ratio and limited programming opportunities.

Jenkins seems to be a source of confusion about US television efforts. Sometimes Jenkins used film as his program source. The US was full of television inventors, though. Jenkins was a prominent one, but he was only one. Many others telecast live programs. Jenkins' own New York station broadcast much live fare, including boxing and fashion shows.

For example, Ernst Alexanderson telecast the first television play as a live event. I refer to "The Queen's Messenger." Alexanderson's GE system sent this play over the airwaves in 1928. This was two years before Baird's "The Man with a Flower in his Mouth." Western Television's "The Maker of Dreams" also preceded Baird's effort. This W9XAO program hit the Chicago airwaves in January, 1930. Station W9XAO followed U.A. Sanabria's television standard.

Despite its experimental TV license, Western Television intended to commercialize television. (At the time, the FRC only issued experimental TV licenses.) Chicago Western stations W9XAO and W9XAP broadcast on regular schedules. The Chicago Daily News newspaper operated W9XAP. The paper listed schedules for both W9XAO and W9XAP.

Jim AA9DT
Last edited by aa9dt on Fri Jul 25, 2008 8:16 am, edited 1 time in total.
aa9dt
Laboratory Assistant
 
Posts: 24
Joined: Wed May 09, 2007 10:09 am

Re: The reasoning behind aspect ratios?

Postby aa9dt » Sun May 20, 2007 5:34 am

ac7zl wrote:Stated another way, if the goal was to produce "portrait" images, it seems to me it would have been far better to place the frame aperature and optics at the 12 o'clock position on the Nipkow wheel.


<h3>Parker & Sanabria</h3>
<b>Long vs. Short-Dimension Scanning</b>

<b>BILL PARKER.</b> Once I mentioned to Bill Parker the idea of short-dimension scanning. Bill had been U.A. Sanabria's chief engineer. Parker agreed on the efficiency improvement of scanning the short dimension. Anyway, Sanabria's Western Television system scanned the long dimension.

By the way, Parker died about 10 years ago. What a huge loss to the television community. Up to the end, Bill Parker continued to design microwave equipment, his other love.

<b>SANABRIA.</b> Some of you know very little about Sanabria. Unfortunately, his publicity isn't as good as Baird's. Sanabria invented a mechanical television system at about the same time as Baird did. For Clem Wade's Western Television, Sanabria designed a 45-line, interlaced
television system. Sanabria earned a patent for this system.

Western Television technology produced excellent pictures. Some two dozen stations across the US, Canada and Mexico used this system. The Radio Manufacturers of America (today's EIA) adopted Western's system as the Midwest standard.

Many Sanabria stations concentrated on entertainment programming, just as Baird's stations did. Chicago had two such stations, W9XAO and W9XAP. Iowa City had one station, a pioneering educational broadcaster. During its six-year life, Iowa's W9XK studio telecast 389 programs.

Telecasting films had nothing to do with Western's four-by-three aspect ratio. For instance, the Iowa shows were all live. For more about Sanabria, see... http://www.hawestv.com/mtv_chicago/mtv_cgo.htm

Jim AA9DT
aa9dt
Laboratory Assistant
 
Posts: 24
Joined: Wed May 09, 2007 10:09 am


Return to Mechanical NBTV

Who is online

Users browsing this forum: Google [Bot] and 5 guests