Talk:Edward Raymond Turner

Online source for the British patent

Is anyone aware of an online source for the patent relevant to this article? British Patent (B.P.) no. 6,202 (1899), 22 March 1899, Means for taking and exhibiting cinematographic pictures Gareth Jones (talk) 15:22, 13 September 2012 (UTC)

This would be really useful, shall have a look and see if anything turns up. Da5nsy (talk) 16:33, 8 April 2013 (UTC)

Found it: http://www.brianpritchard.com/GB189906202A.pdf though don't know how stable it is, might be worth re-hosting/finding alternative if possible. Da5nsy (talk) 08:24, 9 April 2013 (UTC)

Photograph

I just wanted to say... the picture in this article is of the wrong Edward Raymond Turner. This is a photograph of Edward Raymond Turner, the inventor: http://nationalmediamuseumblog.files.wordpress.com/2012/09/edwardturner.jpg

The man whose photograph is depicted in this article is, in fact, a deceased University of Michigan faculty member with the same name as the inventor: http://um2017.org/faculty-history/faculty/edward-raymond-turner

What is recommended to correct this error? Stolengood (talk) 23:43, 19 October 2012 (UTC)

Never mind; have now corrected it myself. Stolengood (talk) 00:11, 20 October 2012 (UTC)

Needs Work

Following restoration of film by BFI/NMM and news coverage by BBC et al. this article requires work.

See following:

http://nationalmediamuseumblog.wordpress.com/2012/09/12/edward-raymond-turner-discovery-re-writes-history-of-early-film/ http://nationalmediamuseumblog.wordpress.com/2012/09/27/stories-about-edward-and-edith-turner/ http://www.nationalmediamuseum.org.uk/PlanAVisit/Exhibitions/LeeAndTurner.aspx http://www.nationalmediamuseum.org.uk/~/media/Files/NMeM/PDF/PlanAVisit/LeeTurnerTimeline.pdf

http://www.bbc.co.uk/news/uk-england-19423951 http://www.bbc.co.uk/news/uk-england-19557914

http://www.youtube.com/watch?v=KkqW6Vu3GUUDa5nsy (talk) 12:23, 8 April 2013 (UTC)

Adding links

usefull link to add to this page as it provides details of the reconstruction http://www.nationalmediamuseum.org.uk/PlanAVisit/Exhibitions/LeeAndTurner.aspx — Preceding unsigned comment added by 82.47.140.47 (talk) 04:39, 14 September 2012 (UTC)

Have added but still needs more adding Da5nsy (talk) 16:32, 8 April 2013 (UTC)

Misinformation gone viral

Cit quality, not quantity, is what needs improvement. The many popular press reports and institutional blog entries are very duplicative and most of them manage to include at least one misleading or erroneous statement. Especially ubiquitous and annoying is the treatment of the films as a recent discovery. Perhaps the BBC (not their finest hour), the New York Daily News (a tabloid and hardly a high-quality cit) and even the director of the National Media Museum have only lately "discovered" them, but mention of them may be found in several standard books on the history of motion picture technology. They were included in a 44-page treatise published in 1969, an illustration from which, showing RGB frame sets from three different sequences, has been floating around the internet since the late 1990s, along with a color composite of the frame set showing the two children and the swing. Newspaper reporters and bloggers can hardly be blamed for not being experts on this arcane subject, but the most that can be hoped from them is that any direct quotes of actual experts have been accurately transcribed. I would gladly trade the lot for this one authoritative account by Brian Pritchard, one of the restorers:

http://www.brianpritchard.com/Lee_Turner_Project.htm

I've just added it as a cit to support, among other editing, a fix of two egregious errors: attribution of projection defects to ordinary parallax, rather than to what Pritchard terms "time-parallax error", a very different thing; and the 48 fps myth, a longstanding error which Pritchard debunks. I've fleshed out the description of the process somewhat but am wary of overwhelming the reader with too much detail about Turner's unique and interesting bucket-brigade projection method. Although I've left it undisturbed for the present, I question the characterization of Turner's system as merely "not perfect"—George Albert Smith, writing in 1908, called it "unworkable" and described the jittery registration of the projected images as "almost unbearable". Conversely, to describe Kinemacolor as merely "semi-successful" seems like a gratuitous slap: despite its obvious shortcomings, Kinemacolor enjoyed considerable international success for several years until a patent suit and World War I shut it down. AVarchaeologist (talk) 12:06, 20 April 2013 (UTC)

What I got from the American Widescreen Museum is that the film was indeed lost for a century and all that we had was an improper print of maybe 6 or 9 b/w frames from it in a 1925 book (which is where we got the boy-and-girl-by/on-a-swing composite frame from), up until the 2012 discovery of the original. It sounds to me like that 1925 book itself wasn't aware that those frames came from Turner, or was openly lying in appropriating them to Urban and Smith instead, and only with Turner's original we found that he was the genius behind it all.

The American Widescreen Museum says nothing of the kind [1] and that page is unchanged since 2008 according to the copyright notice.

For the record: The films and projector were donated to the Science Museum by Charles Urban in 1937 as part of an archive of Kinemacolor-related materials. The Science Museum certainly knew what it had in 1969, when it published The First Colour Motion Pictures, a slim volume apparently available at its bookshop for many years. It includes an illustration showing three sets of three frames, correctly attributed to Lee and Turner, which has been online here [2] since 2008 and elsewhere for a decade longer. The archive was transferred from the Science Museum to the National Media Museum in 2009. There, a curator who was evidently unaware of the films "discovered" them among the transferred items. If the contents of the Louvre are ever transferred to the National Media Museum, perhaps the curators will "discover" an astonishing 500-year-old painting of a bemused woman with no eyebrows by some fellow named Vinci and announce their sensational "find" to a breathless press. But all's well that ends well—the original nitrates have been backed up on safety film and color-composited and now at last we can all see moving pictures of 1902 in full natural color.

As to Turner's "genius", no disrespect to the man, but he did not invent three-color photography (the stock in trade of Ives' Photochromoscope Syndicate, where he was working in 1898) and he was not the first to patent the idea of shooting movie film through a rotating red, green and blue color filter wheel. But it appears that he was the first to actually do it, and his bucket-brigade projection method was a very clever and apparently novel idea, a way of avoiding color bombardment without cranking at an impossibly high frame rate. Nonetheless, even setting aside the inevitable color fringing around moving objects, it proved to be unworkable. It required an impractically high degree of mechanical precision throughout the system. The fact that we can produce nicely registered color composites of the films now does not alter the fact that it was impossible to do so in 1902-1903 and Turner's system was therefore a failure, not an unjustly forgotten success.

On a secondary note, I find the telecine of his work that has been disclosed much more bearable than Kinemacolor. You can tell that they haven't stabilized the invididual frames (or "synchronized", as a few blogs have called it) beyond what the perforation gave them, so this is the original registration Turner's three-color camera gave us.

Nope. The color-compositing of each trio of frames was computer-driven and evidently locked on the images rather than the original large round sprocket holes, which are only partially visible and not physically present on the new standard 35 mm safety prints that were used. We are not seeing the ruinous overall registration problems that plagued the system 110 years ago. The restoration is summarized in Brian Pritchard's one-page authoritative account (the first link in this section, but don't click there if guessing is more fun than knowing). AVarchaeologist (talk) 22:54, 30 May 2013 (UTC)

Anyway, how come that the only time I see a good telecine in motion anywhere on the web, it has heavy watermarking oll over it, whereas all moving versions without those watermarks look like a very poor off-the-wall transfer with hotspot, heavy brightness flicker, very poor black level that resembles a light pinkish-yellowish grey (a black level indeed rivaling the worst ORWO reversal stock I've seen before), and burned-out highlights? I mean, it's not like you can even project the original, right? Wait, didn't the BFI people say they had it copied to regular 35mm? That would explain the off-the-wall transfers. --88.128.70.186 (talk) 15:28, 21 May 2013 (UTC)

In response to AVarchaeologist who counter to WP tradition just crammed his responses into my own post above which makes this discussion confusing to look at for anybody else: What I mean and what you linked above from the American Widescreen Museum makes it sound like The First Colour Motion Pictures from 1969 only reproduced an improper book print from A Million And One Nights: a history of the motion picture through 1925 by Terry Ramsaye from 1925. So it seems like they *DID* discover in 2012 what nobody had their hands on in 1969.

And believe me, I've been working in the trade of stabilizing old footage for a few years, and if they did any kind of digital stabilizing on Turner's footage, they did a pretty poor job. Even Apple Motion could do it better. Rather, what we have now looks like they used a very simple stabilization-by-perforation method, which can be found on some modern telecine units where a laser locates the perforation holes during the scanning process. What you get that way is simply the very same registration you have during projection for as long as your perforation ain't damaged. Anyway, what does the fact that the color-compositing was done digitally have to do with where the sprocket holes are? You don't stabilize footage by adding a particular layer blending mode.

As for Smith's claims of poor registration, when you look at the BBC documentary The Race for Color, you see not only Turner's own footage, but also test footage from the same telecine but shot by Smith using Turner's sequential three-color system after Turner's death, which would be at least a year after Turner took his footage. Smith's later test footage does indeed look much poorer in registration than Turner's. My guess is that the former hypnotist and showman Smith was using experimental gear that hadn't been serviced for some time, and as far as I know, Turner died at work in his workshop, so we can assume that he was still tinkering with his color system, still fiddling with his apparatusses, which could explain the contraption's poor registration that occurred when Smith got hold of it, a lot poorer than with Turner's original footage. --37.80.60.7 (talk) 15:25, 30 June 2013 (UTC)

Interleaving correctly indented responses between the paragraphs of someone else's post is a well-established and accepted means of locating the responses closer to the points being addressed. It is not "counter to WP tradition" and does not confuse an attentive reader, who can see from the level of indentation which of the undersigned "voices" is "speaking". In consideration of your evident horror at finding responses "crammed" between your paragraphs, on this occasion the whole reply is simply appended.

My purpose here is to debunk the misinformation that accompanied the unveiling of the restorations, so as to prevent it from being reported as fact in the WP article and consequently acquiring still more web-wide presence as seemingly authoritative and confirmed truth. My basis is not my own guesses or arbitrary opinions or thought-experiments, or a TV program, but the verifiable historical record of pre-2009 print and high-quality online sources, reinforced in some particulars by the hard evidence in Brian Pritchard's high-resolution scans of parts of the original black-and-white negatives and prints.

Lest anyone think that mine is just one cranky Wikipedia editor's eccentric view of the matter, please see this online article, which I came across only after adding my earlier comments above. Its author is even more blunt about the downright bizarre and inexplicable statements that emanated from the National Media Museum. He also characterizes the BBC's The Race for Colour program, which I have not seen, as "unreliable" for reasons which he details. The relevant material begins about halfway through the article.

No argument about Ramsaye's A Million and One Nights. He thoroughly muddies and mangles the prehistory of Kinemacolor, enlivening his account with suspiciously picturesque anecdotes and obviously confusing one of Turner's 1902 three-color tests with Smith's later two-color tests. Despite (or perhaps because of) being largely based on his own correspondence with the pioneers themselves, Ramsaye's book (published in 1926, BTW) is an unreliable "pop" history rather than an objective scholarly work. It is not the source of the three RGB frame groups reproduced in The First Colour Motion Pictures, but I can see how the Widescreen Museum page could easily mislead someone about that, as the source of the frames used to produce the color composite shown is not actually specified. For the purposes of this discussion, A Million and One Nights can safely be ignored.

A source that cannot be ignored is G.A. Smith's 1908 account of the work he did with Turner and Urban in 1902. Widely quoted in snippets, it first appeared in the December 11, 1908 issue of the Journal of the Royal Society of Arts (vol. 57, whole number 2,925, pps. 70-76, see p. 73) as part of "Animated Photographs in Natural Colours", a paper Smith read to the Society to introduce a demonstration of some Kinemacolor test films. A Google search for the title will soon link you up with the Google Books scan of the original. Smith recalled that "in good sunlight we did succeed in taking a few negatives in which the three colours were duly recorded. It was when we came to superimpose the pictures on the sheet [i.e., projection screen] through three coloured glasses that we found the process unworkable."

So much for the NMM's statement that in 2009, by scanning three frames from one of the films and compositing them as color channels with Photoshop, they made the astounding discovery that Turner had actually succeeded in recording the three color components. That is very old news to anyone familiar with the subject. As the last-cited online article points out, the NMM was using an RGB color composite of the boy-girl-swing scene in 2008, the year before the Turner films were transferred to them, which makes their statement about this particular aspect of their 2009 "discovery" even more absurd.

What the NMM did not, could not and never can "discover" is that the films could be successfully shown with Turner's projector, because the historical accounts and the surviving physical evidence all indicate that in practice it was impossible to achieve satisfactory registration of the three images.

If 37.80.60.7 has been "working in the trade of stabilizing old footage for a few years" (a noble enterprise IMO, and in this era of annoying digital compression artifacts not just a luxury but an urgently needed preliminary treatment—kudos!) he or she should be well aware that software can be used to select stabilization anchor points from among various features in the telecine scan: the actual physical perforations, non-coincident imaged perforations left over from a previous film generation, the edges of the frame, stable objects within the scene, etc. In making color composites, software can be used to rubber sheet mismatching elements, such as three-strip Technicolor negatives with differential shrinkage. In the case of color elements from sequential-exposure processes such as Turner's and Kinemacolor, if the objective is an optimized image regardless of historical accuracy, even so-called time parallax errors (subject motion between exposures) can be cleaned up, within limits, by existing algorithms. No such advanced meddling appears to have been used in the Turner film restorations. Pritchard states that the telecines (Arriscan, 4K) were done from the hand-registered Kodak B&H perf 35mm safety copies, which barely show the original large round perforations if the test strip he illustrates is typical. It is therefore very unlikely that the original perforations were later used as anchor points. He reports that Smoke software was used to create the color composites from the telecine scans. Perhaps 37.80.60.7 is familiar with the features and versatility, or lack thereof, of that particular software package.

In any case, image stabilization should not be equated with color compositing. Logically, they ought to go together—first stabilize, then make the composite—but in practice the stabilization can be optional. For example, with the Turner test film format it is practical to omit stabilizing the film as a whole and use each succeeding frame as the anchor for color-compositing with the preceding two frames, with the result that the three color elements stick together well in each frame of the composite but the image as a whole still weaves and jitters. As odd and irrational as that may be, the visible results strongly suggest that something of that sort was in fact done.

This dogged insistence on regarding Turner's system as a forgotten or suppressed success runs contrary not only to the historical record and physical evidence, but also to logic and common sense: after Turner died, Urban owned the rights to his patent free and clear and could do with it as he pleased. If Turner's system was actually workable, what possible reason would Urban have had to mothball it and substitute a two-component system that yielded a limited range of color and, at 32 fps instead of 16, expensively burned through twice as much film? AVarchaeologist (talk) 07:59, 14 July 2013 (UTC)

Again, whether the frames are tinted via an analogue or digital process has nothing to do with registration. As for Thomas's 1969 book, again, according to The American Widescreen Museum (which had that information on display since at least 2003), Thomas had only copied those frames and all other images from Ramsaye's 1926 book, so apparently in 1969 nobody knew where the original footage was. It's also where the NMM got the one composite frame for its 2008 information sheet from, they all originate from Ramsaye's 1926 book as the only source up until 2012.

Why should we "ignore" that? Just to call out some mysterious conspiracy theory on the NMM? Ramsaye's improper 1926 composite has been online at The Widescreen Museum since at least 2003, and that's exactly the same poor reproduction found on the 2008 information sheet.

And the only time Ruffles ever speaks of "registration", it's not in relation to jumping images, but simply to the very same color fringes that were just as present in Kinemascope, simply due to the chronologically additive nature of both processes.

As for re-writing history, nobody knew about Turner until 2012. Ramsaye 1926 had improperly reproduced those stills of the two kids on the swing as coming from Urban & Smith, not Turner. Due to Ramsaye's improper reproduction, neither did anybody know about the actual quality of Turner's three-color process until 2012. The papers Ruffles talks about identifying Turner were also unknown and undiscovered until 2012. Smith mentioned none of all that in 1908. However, Ruffles confirms my argument that part of the three-color footage was shot by Smith, not Turner, which is obviously the shakey footage, i. e. that taken by Smith.

The rest of the "innacuracies" Ruffles talks about concerns whether Smith was a crook for not sending Ms. Turner any money for her husband's invention, but that's of no concern for our dispute here so far.

And again, it's obvious that no stabilizing options were used beyond stabilizing by sprocket holes, if any, and even that is questionable, as we get more fringing than we would if registration was perfect (as that's what you keep claiming, that it would be perfect because it's been "digitally meddled with"). It's also obvious that no reduction of color fringes has been attempted so far, as those are still annoyingly visible, unlike in Claude Friese-Greene's The Open Road where the fringes have recently been digitally removed (and besides, eliminating temporal color fringes *STILL* doesn't have anything to do with lack of registration as in jumpy images).

That's why Smith's footage of Knightsbridge and Brighton pier (Ruffles even identifies at least the latter as shot by Smith!) is much shakier with poorer registration than Turner's footage. The registration on Turner's footage is far from perfect but still within usual tolerances for analogue filmstrips (one can tell that it's been far less played than most other footage from the era, which is why the registration is better than most from the time), whereas Smith's footage of Knightsbridge and Brighton pier was obviously shot with faulty equipment after Turner's death.

What confuses me though is where Ruffles got the information from that Turner's projector required "three different lenses" *ON TOP OF* a color wheel. That sounds like total overkill where only a color wheel woulda been necessary, and the photos I've seen of Turner's projector only show a single lens with a color wheel. It's true that an orthochromatic stock woulda posed a problem, but that woulda easily been solved either by using filters of varying thickness, or even better by simply using panchromatic stock that was around by 1903.

Also, you're wrong in assuming that Kinemacolor had a higher framerate than Turner's three-color process. It's the other way around because three rather than two frames have to merge in the viewer's eye (Ruffles even mentions some exposure problems of not enough light because Turner's three-color process had such a high framerate, although Edison had been using something like 60fps for some time by then). So Turner's process required more stock due to its higher framerate and was thus more expensive. Oh, and the brightness flicker we've come to know from "old films" was not removed by "digital tweaking", because that flicker we associate with "old films" today is a result of pointing an unsynchronized video camera to a projection screen. To the naked eye, such flicker is not visible on the projection screen. You can go down to 16fps (which for instance was the standard framerate for Regular8) and not see any flicker on the screen.

Now, this may be original research, but I suspect that the *REAL* reason to strip Turner's three-color system down into Kinemacolor was to not having to pay any royalties or interest to Turner's widow by calling Turner's process a failure. Urban didn't "fully own" the process or patent to Turner's three-color process, he only had a right to two-thirds of the profits, whereas the rights to the remaining third were held by Turner (and, by law, were passed on to his widow after his death, although she never saw a single penny), plus an option for Urban to buy the whole three-color process, which never happened. But that foul deed came back to haunt Urban & Smith in 1914 when William Friese-Greene used it to try and steal Kinemacolor from them by pointing out that Turner's original patent that they were still using was a three-color rather than a two-color process, but all he managed was having Kinemacolor accidentally end up in the public domain. --2.240.84.167 (talk) 03:09, 9 December 2013 (UTC)

The fuel sustaining this slow-motion exchange has now been identified. You have finally made it clear that you don't understand what Turner's invention was. Given the amount of misinformation out there, much of it coming from seemingly reliable sources, perhaps that should not surprise me, but I foolishly supposed that anyone would have studied the subject more carefully and in greater depth before presuming to pontificate about it here.

For the sake of saving time, space and effort, I will fast-forward past your repeated misrepresentations of what the WideScreen Museum has to say, of what Ruffles says, and of my own statements; what is and is not actually in Ramsaye's 1926 book; the blind eye you turn to the fact that the 1969 book correctly attributes the frames to Turner when you claim yet again that "nobody knew about Turner until 2012"; your evident confusion about the several different kinds and causes of flicker; the seemingly obvious difference between the inevitable registration breakups caused by moving objects in a sequential-exposure color system and a problem with the overall registration of the three elements; indeed, what "registration" means in the context of a color system using discrete elements, and the difference between that and the ordinary stabilization of "jumpy images"; etc., etc., coming to a complete stop only at the heart of the matter: Turner's projector. Hermann Isensee had already patented the basic idea of filming and projecting through a rotating RGB color filter wheel in 1897, so Turner's unusual projection method was the only major element of his patent that could have withstood the acid test of novelty in a patent dispute. For that and other far more immediately relevant reasons, understanding it is of crucial importance.

If you think that three different lenses in addition to a color wheel "sounds like total overkill where only a color wheel woulda been necessary", then clearly you are still in the dark about what Turner's projection method was and what its theoretical advantages were. You write that "the photos I've seen of Turner's projector only show a single lens with a color wheel", so evidently you have not even bothered to look at the photo of the projector in the Wikipedia article. It plainly shows the triple projection lens: a vertical stack of three lenses with their adjacent edges ground flat, providing the largest apertures possible within the available space. As is explained in the patent itself and on Pritchard's page (secondary sources usually manage to get some part of it wrong), even though the film frames were shot one after the other in a continuous sequence, the projector always showed three of them at a time (superimposed on the screen, of course), but in overlapping groups: frames A, B and C were shown, then frames B, C and D, then frames C, D and E, and so forth. Each frame was projected three times, once through each lens, but always with light of the correct color thanks to the rotating wheel of radially staggered RGB color filter sectors. The purpose was to eliminate color bombardment (one of those several different kinds of flicker alluded to above) without having to shoot and project the film at a high frame rate. The ordinary rate of only 16 fps could be used, and according to Pritchard 16 fps, not 48, is indeed the approximately correct speed for the test films, as demonstrated by the digital video composites.

The Smith-Urban Kinemacolor system did not use any form of Turner's simultaneous projection method, so even if Turner's widow had had any patent rights during the Kinemacolor era (in fact, according to several sources, Urban bought out Turner's share of the rights in 1903), Urban would have been under no obligation, legal, moral or otherwise, to pay her a penny for something he was not using. AVarchaeologist (talk) 11:27, 28 December 2013 (UTC)

Reading your reply (skipping all your unsourced claims you deem irrelevant anyway, and where, among other things, you obviously can't tell the difference between registration, flicker, and color fringes, which I only did my best to properly diferentiate from each other) and then looking at the image in the article again, I see no color blue. How is that supposed to work? Also, changing the color sequence of every composite frame three times sounds awkwardly crude and over-the-top when the desired effect could be achieved much simpler with a bladed shutter. And most of all, what the digital composites evidence is only the recording speed, not the projection speed. A proof for projection speed is much more water-proof when it relates to the actual projector than to digital composites. --87.169.94.205 (talk) 14:09, 27 March 2015 (UTC)

Happily, I did not go to the trouble of making the extensive replies above just to enlighten one tenaciously misinformed anonymous talk page poster. That, it is now clear, was a hopeless cause. Prophylaxis was my primary purpose, because some general Wikipedia users, if interested enough to dig for further details, are apt to comb through an article's talk page, too.

At this point the facts of the matter should be clear enough to anyone who wades through the above and is not so blindly wrapped up in their own self-described guesses, rash assumptions, ill-founded thought-experiments and a conspiracy theory which occasionally descends into the lurid (FYI "foul deed" was a quaint Victorian euphemism for murder) that they are immune to the corrective effects of statements by an actual expert (Pritchard), hard physical evidence, and the verifiable publication dates and contents of several books issued from 1969 on up to the time of the recent so-called "discovery".

The only worthwhile new piece of business seems to be the matter of the projector's blue color filters. You do not see any blue because, unlike the still-vivid stained glass cathedral windows that survive from centuries past, photographic color filters were seldom made of integrally colored glass. They usually consisted of a layer of gelatin or collodion stained with an aniline dye and sandwiched between pieces of colorless glass. The dyes are to varying degrees subject to fading and color shift, sometimes severe, as the result of prolonged exposure to light or gradual chemical deterioration due to aging. If you really have some significant experience with telecine transfers, as you indicated above, perhaps you have encountered a reel or two of old Eastmancolor which was shifting to a nasty overall magenta-red hue. That is a classic example of differential dye fading. The dye used for the projector's blue filter sectors was evidently very fugitive and has faded almost completely, while the red, possibly a blend of magenta and yellow dyes, has faded less but shifted to a magenta pink. When they were new, the projector's filters were unambiguously red, green and blue (or "blue-violet", as deep blue was then often called), according to the patent and early descriptions in print.

Your other comments make it plain that you still do not understand Turner's invention. The filtration alternative would have been to dye the successive frames red, green and blue, a cruder approach later used by others, but not, AFAIK, for simultaneous projection of overlapping sets of frames, which is the peculiar and distinctive feature of Turner's method. The projection speed? The mechanism is hand-cranked, so obviously it could be as slow as the potential overheating and combustion of the arc-light-lit frames in the gate would safely permit or as fast as the operator's arm, the mechanism, and the film itself would bear. There is no reason to suppose that the film was not meant to be run through the projector at about the same rate it was run through the camera. AVarchaeologist (talk) 19:51, 26 April 2015 (UTC)

Multiple sources of registration problems

OR—a rare talk page indulgence (it's legal here): Studying available snippets in which the two perforations ("sprocket holes" seems a misnomer) are visible makes it clear that some of the registration problems originated in the camera. Nearby stationary elements in the scene weave about in relation to the holes, sometimes very substantially, as in this animated gif. Some of the original prints show that the perforations of the negative were varyingly misregistered with those of the print film when the printing was done. That is not especially surprising, because by measuring the distances between pairs of perforations, which the high-resolution scans on Mr. Pritchard's site make possible with a high degree of precision, it is clear that the perforating machine produced inequally spaced perforations; irregular film shrinkage is not an adequate explanation. Add to all that the particular registration problems introduced by the oddball projector, and there seems no reason to doubt George Albert Smith's statement that the ceaseless jittering and weaving of the three color elements relative to each other was "almost unbearable". AVarchaeologist (talk) 23:15, 30 May 2013 (UTC)

I maintain my stance that the claims of poor registration in fact refer to Smith's footage, not Turner's (see above). When looking at The Race for Color, you can clearly see the difference in registration quality between Turner's and Smith's footage both from the same telecine. --37.80.60.7 (talk) 15:46, 30 June 2013 (UTC)

My point is that even if Turner's projector had worked to absolute theoretical perfection, there would still have been serious overall image registration problems due to the demonstrable deficiencies in other parts of the system.

Is a TV program about the films, or what a pair of video color composites, their components registered on an unspecified basis, seem to show—here again, please note that the telecine does not produce the color composite, just raw monochrome video, the frames of which are subsequently sorted out, registered in threes on one of several possible bases (original sprocket holes, new sprocket holes, frame edges, internal anchor points, manual registration by the operator), then composited as RGB color channels—really better evidence than what can be seen and measured in high-resolution scans of the original films themselves? I think not.

The bottom line was not what was theoretically possible, but what was mechanically and commercially practical in the early years of the 20th century. AVarchaeologist (talk) 09:20, 14 July 2013 (UTC)

Well, so far, the only reputable source we have for how the telecine was done is the BBC documentary, and it clearly states that they had to design a unique telecine gate from scratch exactly because they were using the non-standard sized original, *NOT* a 35mm safety print. And again, we see a notable difference in registration between the Turner-shot footage and the Smith-shot footage. If they'd used a 35mm print or had otherwise stabilized the footage, there would not be such a gigantic difference which, as said, makes exactly the Smith footage almost unwatchable, if Turner's design itself was at fault at the degree seen in Smith's footage. Again, it's further unlikely that they've used any modern method at all to enhance registration exactly because, as I've also said before, not even Turner's footage has perfect registration, although it's a lot better than that of Kinemacolor and especially Smith's three-color footage. --87.169.94.205 (talk) 14:26, 27 March 2015 (UTC)

To put it bluntly: the more you say, the clearer you make it to anyone who does their homework that, on this subject at least, you don't know what you're talking about. Don't you find that embarrassing, anonymous though you may be? What do you expect to accomplish by reiterating the same pet theories about which films were shot by the martyred genius Turner and which by the evil and incompetent Smith? Is there a citable source for that stuff, or is it just the product of your own brain under the influence of television, commonly and perhaps too politely classified around here as Original Research?

No, the "only reputable source" is not the script writer for a general-interest TV program, which is in error if it really says what you claim it does, although given your history of seriously misreporting sources there is ample cause to suspect that the error is actually yours. The highly reputable source of my information on this specialized technical subject is Brian Pritchard, the long-established and world-renowned expert on early motion picture technology who, with David Cleveland, built and used that custom film gate you mention, not to do a telecine transfer, but to laboriously copy the original odd-format nitrate films frame by frame onto 35 mm safety film, which was then transferred to video and ultimately color-composited in the digital realm. Is it really possible that after all this time you still have not carefully read or adequately comprehended the brief authoritative account of the matter [3] on Pritchard's site? If only you would take a few minutes to do so, the scales might finally fall from your eyes, but you might find that the foundation under your pet theories was gone.

Wikipedia is text-based and favors verifiable print and high-quality online text sources, while generally deprecating TV programs and online videos as sources. Unlike a good book or a scholarly article, a video provides no citations that would allow the accuracy of its reporting to be checked. A video is inherently less amenable to scholarly study and close comparison with other sources. It appears that your priorities are exactly contrary to Wikipedia's: you accept a BBC TV show (or rather, your understanding of it) as gospel and treat a text source that contradicts it, no matter how expert the author may be, as just a bit of heretical noise to be ignored. An irrational and foolish preference, it seems to me. AVarchaeologist (talk) 23:05, 26 April 2015 (UTC)

Birthplace

He was actually born in Bedminster, Somerset. Now part of Bristol. https://www.familysearch.org/ark:/61903/1:1:3RXB-LMM?cid=fs_copyhttps://www.familysearch.org/ark:/61903/1:1:3RXB-LMM?cid=fs_copy — Preceding unsigned comment added by 2A00:23C4:B416:9D01:2147:D2ED:BEEA:87C1 (talk) 16:23, 24 October 2021 (UTC)