{ "$schema": "/api/schemas/documents.schema.json", "apiVersion": "1.0.0", "generatedAt": "2026-07-09T21:16:49.031Z", "sourcePath": "src/main/data/docs", "docId": "10.5594-M00474", "document": { "abstract": "This paper explores how the temporal characteristics of image capture and image display devices interact with eye tracking. A display system with a large pixel count can be exploited only by having a wide viewing angle, such as the 30 degrees of HDTV Eye tracking – which is insignificant for conventional television – becomes significant in HDTV. A fast-moving element might take as little as two seconds to traverse the width of a screen. For 1920 samples per picture width, this corresponds to 16 pixels per field time at 60 Hz. Artifacts due to temporal effects at the camera, and artifacts due to temporal effects at the display, can be expected to surface in HDTV displays. Many emerging displays have pixels that emit a constant amount of light throughout a large fraction of the frame time – they have long duty cycles. We have little experience of motion portrayal on displays having long duty cycles. Computers cannot yet display full-screen motion, and although conventional television can display smooth motion, it is restricted to arrow viewing angles. But it's clear that wide-angle, long duty cycle displays will introduce substantial blur on objects that the eye is tracking, and thus will have poor motion portrayal. Many emerging display technologies, such as plasma display panels (PDPs) and deformable-mirror displays (DMDs) are intrinsically bilevel: At any instant in time, light is either emitted or not at any pixel. Apparent grayscale reproduction can be achieved by pulse-width modulation (PWM). The PWM technique works well when image content is static. But when PWM is combined with eye tracking in scenes with rapid motion, a new class of artifacts is introduced. Digital technologists have long speculated about displays where each pixel is updated independently. It is assumed that if updating is at least as frequent as the arrival of new frames, the display will be free of artifacts. I will prove using several very simple examples that artifacts will be introduced unless the updating process is spatially coherent.", "abstract$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "authors": [ { "affiliation": "Poynton Vector Corporation Toronto, CANADA", "name": "Charles A. Poynton" } ], "authors$meta": { "confidence": "high", "note": "Object form (schema 2.3.0) migrated from string form via migrateAuthorsToObjectForm.js; bio/affiliation backfilled from _source/SMPTE/HIGHWIRE/ NLM ++", "source": "parsed", "updated": "2026-06-23T00:04:25.562Z", "version": "authors-migrate-object-form@v1" }, "contentType": "orig-research", "contentType$meta": { "confidence": "high", "note": "Canonical-confirmed full paper (summary-abstract set is orig-research in the SMPTE canonical repos); HIGHWIRE-era abstract label was a delivery artifact", "originalValue": "summary-abstract", "source": "resolved", "updated": "2026-07-09T19:12:47.628Z", "version": "contenttype-fixes@v1" }, "copyright": { "holder": "Society of Motion Picture and Television Engineers, Inc.", "holder$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "year": "1996", "year$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" } }, "copyright$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "docId": "10.5594-M00474", "docId$meta": { "confidence": "high", "note": "Derived from DOI 10.5594/M00474", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "docLabel": "SMPTE Meetings and Conferences ( February 1996)", "docLabel$meta": { "confidence": "medium", "note": "Composed from venue title, volume/issue and date", "source": "inferred", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "docTitle": "Motion Portrayal, Eye Tracking, and Emerging Display Technology", "docTitle$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "docType": "Conference Paper", "docType$meta": { "confidence": "high", "note": "NLM
— SMPTE conference paper", "source": "inferred", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "doi": "10.5594/M00474", "doi$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "href": "https://doi.org/10.5594/M00474", "href$meta": { "confidence": "high", "note": "Constructed from DOI 10.5594/M00474", "source": "inferred", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "number": "14", "number$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "pages": "192–202", "pages$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "publicationDate": "1996-02-01", "publicationDate$meta": { "confidence": "high", "note": "Day absent in NLM source — padded to 01", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "publisher": "SMPTE", "publisher$meta": { "confidence": "high", "note": "Normalised to registry \"SMPTE\" convention from NLM publisher-name (The Society of Motion Picture and Television Engineers)", "source": "inferred", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "publisherLocation": { "city": "White Plains, NY", "city$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" } }, "publisherLocation$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" }, "resolvedHref": "https://journal.smpte.org/conferences/Proceedings:%20The%2030th%20SMPTE%20Advanced%20Motion%20Imaging%20Conference/20/", "resolvedHref$meta": { "confidence": "high", "note": "Backfilled from url.validate report #resolvedHref", "originalValue": null, "source": "resolved", "updated": "2026-06-19T20:59:07.439Z" }, "status": { "active": true, "active$meta": { "confidence": "medium", "note": "No explicit status in NLM source — conference papers default active", "source": "inferred", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" } }, "volume": "1996", "volume$meta": { "confidence": "high", "note": "Parsed from NLM article XML (_source/SMPTE/HIGHWIRE/Source Bak/smptem/10.5594_M00474.xml)", "source": "parsed", "updated": "2026-05-22T21:44:19.125Z", "version": "smpte-conference-paper-nlm@v1" } } }