Identifying Good & Bad 2D to 3D Conversion

Metadata

Publisher: SMPTE — White Plains, NY
Doc Type: Conference Paper
Content Type: Original Research
Volume: 2010, No. 10, pp. 1–9
Abstract: 2D to 3D conversions vary greatly in terms of quality. Understanding what causes bad conversion and how to identify those issues is an essential skill to critique 3D and manage a conversion project. A bad conversion is usually the result of a cost cutting measure. Usually 1 or more of the 3 main artistic phases (depth, roto/matte, and paint) is shortchanged. Inaccurate or conflicting depth cues and the “card-board cutout” effect arises when the depth process is shortchanged. Poor or low detail models, used to drive the depth of the scene, are the usual culprits. The “rubber sheet” effect, “watery, stretched, or messy” edges, or lack of transparency is the result of shortchanging the paint process. To avoid painting occluded surfaces, the “rubber sheet” approach creates a sense of depth with no distinct separation between elements. Automated paint process often lead to “Watery, stretched of messy edges”. A transparency, such as smoke, is often allowed to play against a wall instead of in mid-air because that could require paint to remove and replace the smoke. A “composited” looking shot can arise when the roto/matte process is shortchanged. Flyaway hair may be removed or matte edges may be overtly apparent because they are too hard or soft. This makes the shot look composited, even if it was captured practically. Uncomfortable or fatiguing conversion arises if there is poor QC or stereography. Vertical alignment, too much positive parallax, and poor stereo continuity are likely causes of uncomfortable of fatiguing conversion.
Publication Date: 2010-10-01
DOI: 10.5594/M001386
Link: https://doi.org/10.5594/M001386
Author(s): Matthew DeJohnIn-Three Inc., 4580 E. Thousand Oaks blvd., Westlake Village, CA 91361
Keyword(s): Stereoscopic
Copyright: © 2010 Society of Motion Picture and Television Engineers, Inc.

Source Data (JSON)

Full registry record with provenance metadata. Open directly: /api/doc/10.5594-M001386.json

Reference this Doc

Plain text (ISO 690 compliant)

Preview:

Matthew DeJohn; Identifying Good & Bad 2D to 3D Conversion, SMPTE Meetings and Conferences ( October 2010); SMPTE, 2010. Available at https://doi.org/10.5594/M001386

Snippet:

Matthew DeJohn; Identifying Good &amp; Bad 2D to 3D Conversion, SMPTE Meetings and Conferences ( October 2010); SMPTE, 2010. Available at https://doi.org/10.5594/M001386

HTML (ISO 690 compliant)

Preview:

Matthew DeJohn; Identifying Good & Bad 2D to 3D Conversion, SMPTE Meetings and Conferences ( October 2010); SMPTE, 2010. Available at https://doi.org/10.5594/M001386

Snippet:

<span class="citation">Matthew DeJohn; <cite>Identifying Good &amp; Bad 2D to 3D Conversion</cite>, SMPTE Meetings and Conferences ( October 2010); SMPTE, 2010. Available at <a href="https://doi.org/10.5594/M001386" target="_blank" rel="noopener">https://doi.org/10.5594/M001386</a></span>

SMPTE's HTML Pub

Preview:

Matthew DeJohn; Identifying Good & Bad 2D to 3D Conversion, SMPTE Meetings and Conferences ( October 2010); SMPTE, 2010
doi: 10.5594/M001386
url: https://doi.org/10.5594/M001386

Snippet:

<li>
Matthew DeJohn; <cite id="bib-10-5594-m001386">Identifying Good &amp; Bad 2D to 3D Conversion</cite>, SMPTE Meetings and Conferences ( October 2010); SMPTE, 2010
<span class="doi">10.5594/M001386</span>
</li>