Zoetrope Bench — credits, sources and differences https://zoetrope.skillsafe.ai/ WHAT THIS IS ============ An independent reimplementation. The zoetrope is Victorian, and the people who built it are long out of copyright, so there is no rights holder to credit. What is credited here is the work that was read, and every source is named together with exactly what was taken from it and whether it could be read at all. No code, data, art, text or asset was taken from any other simulation, game, textbook or teaching package. The sampler, the recovery, the renderer, the user interface and every assertion in the test harnesses were written for this bundle. The page ships no third-party library, loads nothing from a content delivery network, and calls no programming interface. THE PEOPLE WHO BUILT THE MACHINE ================================ 1833 Joseph Plateau and Simon Stampfer, independently, publish the stroboscopic disc (phenakistiscope). Stampfer's July 1833 pamphlet, "Die stroboscopischen Scheiben", suggests the sequence could equally go on a cylinder or a looped strip. 1834 William George Horner publishes the Daedaleum, January 1834, in The London and Edinburgh Philosophical Magazine and Journal of Science, volume 4 number 21, page 36. His slits are BETWEEN the pictures, not above them. 1865 William Ensign Lincoln invents the definitive zoetrope at about eighteen, at Brown University, Providence, Rhode Island. His change is to put the slits ABOVE the pictures, which allows easily replaceable paper strips. 1867 US Patent 64,117, granted 23 April 1867, assigned to Milton Bradley and Co. Lincoln coined the name "zoetrope". James Clerk Maxwell later built a variant using concave lenses of focal length equal to the drum diameter instead of slits, giving a sharper and steadier virtual image. This history is taken from the English Wikipedia articles "Zoetrope" and "Wagon-wheel effect", read in full through the MediaWiki REST interface. It is used for history and for Schouten's published band figures, and for nothing else. No physical constant on this page comes from a tertiary source. SOURCES READ IN FULL, AND HOW ============================= S1 VanRullen R, Reddy L, Koch C (2006). "The Continuous Wagon Wheel Illusion Is Associated with Changes in Electroencephalogram Power at ~13 Hz." Journal of Neuroscience 26(2):502-507. doi:10.1523/JNEUROSCI.4654-05.2006 Read as a born-digital PDF from the CaltechAUTHORS repository (record ha3yz-b3q40, file 526.pdf), extracted with pdftotext. The file carries embedded Type-1C fonts: it is not a scan and nothing here comes from optical character recognition. TAKEN: the discrete-sampling side of the dispute and its EEG evidence; the 10-15 Hz estimate of the postulated snapshot rate; the stimulus description (16 sinusoidal spokes, 100 per cent contrast, 12.5 degrees of visual angle, 160 Hz monitor); the statement that the mechanisms are "heavily debated". S2 Kline KA, Eagleman DM (2008). "Evidence against the temporal subsampling account of illusory motion reversal." Journal of Vision 8(4):13, 1-5. doi:10.1167/8.4.13 Read as the PubMed Central author manuscript PMC2856842 (NIHMS-181651), real text. TAKEN: the correspondence-matching rule as this paper states it — a feature travelling more than half a period between snapshots is matched with its predecessor; the rivalry/motion-aftereffect side of the dispute; the two objections that overlapping orthogonal motions reverse separately and that reversal occurs with non-uniform and non-periodic stimuli. S3 Tan ES (2018). "A psychology of the film." Palgrave Communications / Humanities and Social Sciences Communications 4:82. doi:10.1057/s41599-018-0111-y Read as gold open-access full text. TAKEN: the history of the persistence-of-vision explanation, including Roget's 1824 report to the Royal Society; the Andersons' argument against it as Tan reports it; the cinema frame-rate figures (at least 24 frames per second, 72 flashes per second better, citing Anderson 1996 pages 54-59); and — importantly — the statements that "the riddle of apparent motion has not yet been solved up to now" and that of the proposed mechanisms "none of the proposals have been accepted as the final solution". S4 Ehrenstein WH (2003). "Basics of seeing motion." Arquivos Brasileiros de Oftalmologia 66(5 suppl):44-52. Read as a born-digital PDF from SciELO. TAKEN: the accepted account of stroboscopic motion as "discrete, successive changes of stimulus positions which lead to the perception of continuous motion... the basis for perceived motion in movies"; Wertheimer's 1912 sequence from simultaneity through partial movement and pure motion (phi) to optimal motion and then succession; the beta/phi distinction; the bilocal motion detector tuned to a span over a delay. S5 English Wikipedia, "Zoetrope" and "Wagon-wheel effect" (MediaWiki REST HTML). TAKEN: the history above; Schouten's three bands; the fact that discharge lamps on alternating current flicker at twice the line frequency. SOURCES REPORTED UNREAD ======================= Each of these is cited for its EXISTENCE and for the claim that a source read in full attributes to it. None is quoted from an abstract, and none is quoted from an optical-character-recognition dump. Purves D, Paydarfar JA, Andrews TJ (1996). "The wagon wheel illusion in movies and reality." PNAS 93(8):3693-3697. WHY UNREAD: PubMed Central holds this article only as scanned page images (pnas01515-0543.pdf) and pnas.org returns HTTP 403. The only text obtainable was an OCR dump. Its claim is taken from S1 and S2, which both state it and were both read in full. Kline KA, Holcombe AO, Eagleman DM (2004). Vision Research 44:2653-2658. WHY UNREAD: sciencedirect.com returns HTTP 403. Claim taken from S2 (same first author). Andrews TJ and Purves D (2005), Trends in Cognitive Sciences 9:261-263, and Holcombe AO, Clifford CWG, Eagleman DM and Pakarian P (2005), same journal 9:559-560. WHY UNREAD: closed access, no repository copy found. Named as the published exchange. Anderson J and Fisher B (1978), Journal of the University Film Association 30(4):3-8, and Anderson J and Anderson B (1993), Journal of Film and Video 45(1):3-12. WHY UNREAD: JSTOR, not reachable. Their argument is taken from S3, which summarises it explicitly and was read in full. Schouten JF (1967), in "Models for the Perception of Speech and Visual Form". WHY UNREAD: print only. His band figures (alpha 8-12, beta 30-35, gamma 40-100 cycles per second) are taken from S5 and are flagged on the page as adjacent-source evidence. Horner WG (1834), Philosophical Magazine 4(21):36-41. WHY UNREAD: scans only. The bibliographic record is from S5. WHAT THIS BENCH MEASURED FOR ITSELF =================================== Everything numeric on the page comes out of tools-study.js, running the sampler in js/zoetrope.js, and is regenerated into js/findings.js by tools-findings.js. Nothing is typed in by hand. The sampler is never given the aliasing law; the law is what its cross-correlation turns out to produce. - 576 grid points over spoke counts 8, 16, 24 and sampling rates 10, 12.5 and 24 per second: the recovered apparent rate and the closed form differ by at most 2.13e-14 per second. - Standstills recovered exactly; reversal thresholds recovered to exactly minus one half-bin of the correlation at every resolution from 16 to 512 bins per period, the error halving at each step, with an independent crest tracker at 8192 bins per period giving 5.0006189 against a closed form of 5. - The drum rate cancels: six drums measured across a 30x range of drum speed moved by at most 7.56e-16 cells per flash. - Contrast against smear is not monotone; it reaches exactly zero at whole numbers of pattern periods and returns contrast-reversed where the transfer function is negative, at 13 settings and at none of the 27 where it is positive. - A mark of 0.3 per cent on one spoke abolishes the reversal at every speed. - The persistence-of-vision model is time-symmetric to 2.64 units in the last place. WHAT DIFFERS FROM A REAL ZOETROPE ================================= 1. THE SLIT IS A PURE TEMPORAL GATE. The strip is treated as advancing one Nth of a revolution per flash and the gate as open for a fixed fraction of the interval. That is the standard idealisation. The exact ray geometry of an eye at a finite distance D from the axis of a drum of radius R adds an aperture gain of (D+R)/(D-R), which slides the visible WINDOW across the far wall during the opening. That crops what is seen; it does not blur it. The blur is the strip's own travel during the opening, and that is what is integrated. The gain is computed and exported so the size of the omission is stated. 2. THE PATTERN IS A LUMINANCE RING, not a drawn animation. The bouncing figures in the drum view are decoration painted from each cell's phase; the measurement is made on the ring, not on the figures. 3. NO EYE IS MODELLED. There is no retina, no receptive field, no adaptation and no aftereffect here. Where the question is what a visual system does, the page quotes people who measured it and says so. 4. ONE PERCEPTUAL ASSUMPTION, STATED. Each flash is matched to the nearest correspondence in the previous one. That is an assumption. The aliasing law is not: it is what the assumption produces. 5. NO FLICKER MODEL. The critical fusion frequency is quoted from S3 and not derived. A drum too slow to fuse is out of this bench's reach and the page says so. 6. NO COLOUR, no three dimensions, no sound, no physical drum inertia, and no air. THE NAME ======== The displayed name is "Zoetrope Bench". The slug is `zoetrope`. A TMview search was run while building this (POST https://www.tmdn.org/tmview/api/search/ results, criteria "C" = contains, paged to exhaustion, filtered locally to Nice classes 9 and 41). It returned 42 records containing "zoetrope", of which 5 are LIVE (Registered or Filed) in Nice class 9 (computer software) and/or Nice class 41 (entertainment and education services): zoetrope vinyls | TR | Registered | ISMAIL NECAT PAKYUZ | 2025-040726 | Nice [9,10,12,16,25,35] Zoetrope Vinyls | GB | Registered | Zoetrope Vinyls Ltd | UK00004256808 | Nice [9,42] ZOETROPE STUDIOS | US | Registered | AZX, LLC | 76263937 | Nice [35,38,40,41,42] AMERICAN ZOETROPE| US | Registered | AMERICAN ZOETROPE | 74344688 | Nice [41] AMERICAN ZOETROPE| US | Registered | AMERICAN ZOETROPE | 74345506 | Nice [9] The BARE WORD MARK "ZOETROPE" is live in NEITHER class 9 nor class 41 anywhere in TMview. Every bare-word record is either in an unrelated class or is Ended or Expired: ZOETROPE | US | Ended | Coppola, Francis | 74345508 | Nice [9] <- exact string, class 9 ZOETROPE | CN | Registered | Shanghai trading co. | 21657289 | Nice [25] ZOETROPE | US | Registered | GMYL, L.P. | 86630485 | Nice [33] ZOETROPE | GR | Registered | UNI-PHARMA A.B.E.E. | N292344 | Nice [5] Zoetrope (JP composite) | JP | Expired | Nippon Telegraph and Telephone | 1995124209 | Nice [9] Zoetrope (JP composite) | JP | Expired | Nippon Telegraph and Telephone | 1995124216 | Nice [41] ZOETROPE MUSIC | FR | Expired | DURAND Frederic | 1402935 | Nice [9,41] So the name was NOT forced to retitle by a live mark. It retitled anyway, for two reasons that are visible in the rows above: (a) there is an EXACT-STRING prior application for the bare word in exactly class 9 — US 74345508, filed 1993-01-04, status Ended. An exact string in the right class, even Ended, was treated as reason enough to pick a different title; (b) "AMERICAN ZOETROPE" is live in BOTH class 9 and class 41 and "ZOETROPE STUDIOS" is live in class 41, so a bare "Zoetrope" title would sit next to film-industry marks living in the same two classes. Candidates probed, all on criteria "C", all offices, all classes, all statuses: "zoetrope bench" -> 0 records worldwide <- CHOSEN "zoetrope lab" -> 0 records "zoetrope sampler" -> 0 records "zoetrope stroboscope" -> 0 records "sampling bench" -> 0 records "strobe bench" -> 0 records "phenakistoscope" -> 0 records "wagon wheel effect" -> 0 records "slit sampler" -> 1 record, none live in 9 or 41 "praxinoscope" -> 3 records, none live in 9 or 41 "zoetrope studio" -> 5 records, ONE LIVE in Nice 41 (ZOETROPE STUDIOS, US 76263937, AZX LLC) -> REJECTED "Zoetrope Bench" was chosen over the other zero-record candidates because "Bench" reads as an instrument bench rather than as a collection, and a visitor still recognises the object in the first word. NO CLEARANCE IS CLAIMED. A database search is not a legal opinion. TMview aggregates participating offices only and does not reach unregistered or common-law rights. The full search output is reproduced verbatim in tmview-findings.txt alongside this bundle's source. LICENCE ======= MIT. See LICENSE.txt.