SDTV Lens on HDTV Camera: To Be or Not to Be?

The essential distinction between SDTV and HDTV is bound up in the core issue of the term “high definition”. It implies that HDTV is distinguished from traditional video primarily by more “definition” – thus producing much higher picture sharpness. Thus, any discourse on mixing SDTV products with HDTV products needs to be closely examined from the viewpoint of its impact on the “high definition” aspect of the final imagery. Unlike the digital cameras and recorders that make up contemporary digital imaging systems, the lens is a totally analog technology. It is a very physical technology, in the fullest sense of the word. The lens is also dynamic— in terms of the substantial degree of control it can exercise over the object image that it presents to the camera image sensors.

Those variations in light level, focus, and focal range offered by the lens come with some technical penalties. As this paper is intended to examine “definition” the focus will be on the behavior of lens resolution. It is useful to establish some metrics in optical terms, for picture “definition”. Audio and video systems (such as a television camera) are described by considerations of bandwidth and the specific system responses over the frequency ranges encompassed within their respective bandwidths. A similar approach can describe the resolution performance of a lens.

OPTICAL BANDWIDTH
Lens Contrast and Resolution are inextricably intertwined. A series of closely spaced alternating black and white lines are visually distinguished by their relative contrast to each other. As their thickness and spacing are progressively reduced our human visual system is tasked to distinguish between these alternating lines. At some point we fail to do so and they blur into a gray patch. The same thing occurs as the test chart object scene passes through a lens. As the alternating lines increase in spatial frequency their optical representation by the lens will exhibit a progressive roll-off as simplistically illustrated in

Figure 1. In other words, the contrast reproduction capability of the lens is modulated as a function of the fineness of detail of the alternating black and white lines. This particular representation of the lens output is technically termed the Modulation Transfer Function – or MTF. The horizontal axis represents the spatial frequency (increasingly fine detail from left to right) in Line-pairs per millimeter (Lp/mm). The vertical axis is the contrast (amplitude of black to white) of the optical image output of the lens.

Download pdf SDTV Lens on HDTV Camera: To Be or Not to Be?

Experiments with Delivery of HDTV over IP Networks

The conversion of broadcast television from the legacy analog PAL and NTSC standards to digital format has many exciting implications. These include the possible convergence of television distribution and computer network infrastructures, allowing interactive applications, and the increase in quality possible with high definition digital formats.

To date, the different aspects of this convergence have been studied in isolation: there has been much work on the transport of compressed standard definition TV over IP, and much work defining protocols and standards for high definition TV (HDTV), but few have studied the transport of HDTV over IP. In this paper we present our initial experiments with a system to deliver production quality uncompressed HDTV over IP networks.

Why do we chose to deliver uncompressed HDTV? Several reasons, primarily to maintain image quality and reduce latency. This is most useful in a production facility, where image degradation due to repeated compression cycles is undesirable, but may also be appropriate for very high quality telepresence applications. Delivery of compressed HDTV, using existing MPEG-2 over IP standards, may be more appropriate for other applications.

The outline of this paper is as follows: section 2 covers background in HDTV technology, protocols for transport of video over IP networks and network performance. This is followed, in section 3 with a discussion of the options for protocol development, with our design being outlined in section 4. Section 5 provides preliminary performance analysis of our system, demonstrating transmission of HDTV over a wide-area IP network, with section 6 outlining directions for further development. Finally, we summarize related work in section 7, and provide conclusions.

Get pdf download Experiments with Delivery of HDTV over IP Networks

Tips for Computer Scientists on Standard ML (Revised)

This note is inspired by a brilliant piece of writing, entitled Tips for Danes on Punctuation in English, by John Dienhart, Department of English, Odense University (1980). In a mere 11 pages, Dienhart’s lucid writing gives the reader the impression that punctuation in English is pretty easy and that any Dane can get it right in an afternoon or so.

In the same spirit, this note is written for colleagues and mature students who would like to get to know Standard ML without spending too much time on it. It is intended to be a relaxed stroll through the structure of Standard ML, with plenty of small examples, without falling into the trap of being just a phrase book.I present enough of the grammar that the reader can start programming in Standard ML, should the urge arise.

The full grammar and a formal definition of the semantics can be found in the 1997 language definition[2]. Some of the existing textbooks also contain a BNF for the language, e.g., [3]. I have tried to use the same terminology and notation as the language definition, for ease of reference.

Download pdf Tips for Computer Scientists on Standard ML (Revised)