| Author | Message | ||
     VRJUNKIE Rating: N/A |
I would have to strongly disagree with Christoph's recent review of these tapes. Classifying them as 'IMAX-3D Disease' is VERY undeserving. I have been waiting for YEARS for high quality 3D content! 2nd and 3rd generation VHS tapes PALE in comparison to these titles. I have paid 50.00 US for copies of these VHS tapes and these crystal clear DVDs with instant scene access is under 20.00!!!!! And as for complaints about things coming out of the screen... what do we watch 3D content for??? Who only wants the aquarium in-screen effect??? I love out of screen effects. The 2 best ones are Encounter and Haunted Castle - the rock video at the end of Haunted castle alone is worth the price! Just set as close to your TV as you can stand and dive in! If you have ANY interest in 3D video these DVDs are a MUST! You will only see better 3D in a professional 3D amusement park environment. And the sound if you have a home theatre set up is AWESOME! The Haunted Castle and Encounter movies also use good color and contrast selection to keep flicker and ghosting to a minimum - maybe Christoph was watching at 50 hz? To properly view these, set brightness and contrast both to a minimum, get as close to the set as tolerable - and ENJOY! I/O display has made a terrific package available. I think you can get the wireless glasses, video adapter box, and 4 videos for 100.00!!! This is fantastic! Their latest controller will operate wired or wireless glasses... if you have the old Wicked 3D glasses, or the wired internet surfer glasses - they're compatible! And there is a stereo reverse button on the front of the adapter! I was even able to get my Hitachi 53" to view these videos in progressive scan mode using component video connections! Yes that's right - I just fed another video signal into the box (from my VCR which was getting video from my DirecTv sat) to fool it to get it to tell the glasses to flash and it is in sync! I can repeatedly get them in sync - it appears everything syncs to the same frame rate? I also watched the videos on a standard 27" TV and was impressed there as well. If you have a standard rear projection TV I don't think it would be possible to be disappointed! I LOVE these videos - And the tech support by Ross at I/O display is great! We need to encourage high quality 3D content not slam it! -VR | ||
| Christoph Bungert (Admin) Rating: N/A |
>>> I would have to strongly disagree with Christoph's recent review of these tapes. First you should read it! Obviously you read little more than the headline. >>>Classifying them as 'IMAX-3D Disease' is VERY undeserving. The 'IMAX-3D Disease' addresses the excessive use of large negative parallax in all IMAX-3D movies, not these DVD's in particular. >>>... what do we watch 3D content for??? Not for headaches. >>>Who only wants the aquarium in-screen effect??? I love out of screen effects. Out-of-screen effects are great, but they must be used sparingly. >>>the rock video at the end of Haunted castle alone is worth the price! You must be kiddin'. >>> maybe Christoph was watching at 50 hz? That's technically impossible. >>> I was even able to get my Hitachi 53" to view these videos in progressive scan mode That's technically impossible. Progressive scan would mix the right and left eye view making it impossible to watch in 3D on a TV. Probably your TV applies some frequency-doubling and/or line-doubling which keeps the half-frames separated. Christoph | ||
| Larry Elie Rating: N/A |
Christoph is right about the progressive scan; I think your player or TV did some defaults. I'm not saying it couldn't sync, just that the half-frames (let's call them fields) would not carry all the 3D info. The TV is probably doing some line doubling. Having said that, Christoph, if you must compare resolution with a factor like you did, you don't do it with pixels, but with resolution. 71 times the number of pixels doesn't tell much anything except how much data the computer is processing. You wouldn't (or shouldn't) compare a 640x480 image to one that is 1280x960 and say it is 4 times as high, you say it is twice as high. 4096/720 or 5.7 is the factor of resolutions. BTW, 4096 is LOW; good prints from 70mm film may go that high or more. DVD isn't IMAX, but it isn't 1/71 of IMAX to an observer either. I have seen a DLP projection of a 720x480 (usually you count BOTH fields, and in this case it was progressive scan) on an IMAX screen in a demo. It wasn't quite full screen, but it was bigger than most 70mm screenings on IMAX that were the rage a few years ago. It was worse, but only about 4 times worse. It really helped me realize that resolution was not the most important issue. Personally, I think IMAX biggest problem is frame rate (I saw Beauty and the Beast 2D IMAX last night) and I think all IMAX 2D should be shot at 50+ Hz and 3D should be shot at 50Hz plus (100 fields/sec) and that may happen with a digital LF IMAX or IWERKS version eventually, but that's another discussion. Another point of discussion; although I would LOVE to see old Hollywood 3D films on DVD, there are OTHER IMAX titles in release that the producers choose NOT to print in 3D. For example; T-Rex. You can buy it at Amazon and probably even K-mart. I think I have seen others. These were NOT Slingshot. Slingshot deserves some credit here. Content is indeed lacking in these DVD's, but that isn't Slingshot's fault. Just some data on WHY these DVD's were released: Current Box Office numbers from Yahoo, these are 3D's still in distribution in the US. I don't think I missed any. Titles with a * weren't showing for all December. T-Rex: Back to the Cretaceous $38,078,959 Across the Sea of Time $15,803,102 Wings of Courage $14,986,752 Cirque du Soleil: Journey of Man $12,989,871 Galapagos $12,020,923 Cyberworld $10,462,583 Encounter in the Third Dimension $ 6,073,613* Alien Adventure $ 3,863,093* Haunted Castle $ 3,306,051* Mark Twain's America $ 2,136,630* Notice too that "Encounter in the Third Dimension" was also shown under the title "3-Mania". Other than Ultimate G's, NONE of the films are anything other than N WAVE pictures, and Ultimate G's wasn't even on the list of box office scores. The bottom line is that most of the real money makers in 3D aren't on DVD AT ALL (except T-Rex, in 2D). My prediction; Mark Twain's America didn't make much money (I think it only shows at one theater on the Mississippi) so it may go 3D DVD. Otherwise, don't hold your breath. Also note; since these are in release for YEARS, producers are going to try and repeat the 'success' of T-Rex. Pity. Ultimate G's (for example) was a better film and didn't even get 10% of the box office. For that matter Across the Sea of Time and Wings of Courage were much better too, but they were on the order of half of the box office (and over a longer period). Money is the bottom line. IMAX tickets are more expensive than most Hollywood features, so these numbers represent small numbers of viewer. Larry Elie | ||
| VRJUNKIE Rating: N/A |
Number 1 Here's my bottom line on these DVD's: Name a better consumer product for viewing 3D on your TV other than the the 2 DVDs I raved about .... I'm waiting... I thought so. Number 2 I may not know all the technical details of what's going on between my DVD player and TV - but here is my equipment and setup: I have a Pioneer 434 progressive DVD player. It has 3 modes for output - interlace, progressive, and auto. I have it set to progressive. I have 3 ways to send the video from my DVD player to the tv: Video out (rca plug), S-video, and component video output (using 3 cables just for video). The only connection I use is component video (where you have 3 video cables). Now from this setup it is impossible for the controller to get it's sync from the DVD - I'm not using the video connection and there is a physical switch on the back to go from video to component video - it's set to component video. And I checked the regular video connection is dead when set to component video. Since I have my DVD set to progressive I have to assume it is? I know from what I've read that its suposed to be frames at this point - I can't explain how I'm getting stereo - but I am! What's particulary interesting is that hhen I change the setting to interlace - it destroys the stereo! It only works when it's set to progressive. Now my TV does have line doubling... but get this - when I use the video out from my DVD to my big screen TV - I cannot get stereo regardless if I go thru the adapter box or not. Neither does it work with any setting of progressive, interlace, or auto on the output of the DVD palyer! I only get stereo when in progressive mode using the component video connections. I have researched Hitachi's faqs and my TV the (53FDX01B) is indeed using 60 hz when using component video. So you guys tell me what's going on if it's not working in progressive mode??? -VR | ||
| M.H. Rating: N/A |
Just some notice of watching the DVD. I was watching all of this DVD by 3 mrethods: 1) On TV in the 30/60 Hz mode. H+ Real magic decoder in computer used to get the analog signal .. 2) In VFX3D HMD. S-VHS of the H+ Real Magic mpeg2 PC decoder used to get the signal 3) On 22' monitor in 60/120 Hz mode. mpeg2 streem reprocesed to above/below coded DIVIX4 before playback. Filtered OpenGL HW accelerated zooming used (P III at 1G and GeForce 3), experimental playback software I have enjoyed the method 2) as the best. Reasons: The HMD optic compenzate partialy the out of screen negative effct. A lot of white color in the movies produce too much ghosting for method 3). Hi flickering ocuure in method 1). I am realy looking fororward to some next generation HiRes HMD with stereoscopic support compensating the low-res of VFX3D. | ||
| Christoph Bungert (Admin) Rating: N/A |
Larry, thanks for the box office numbers and insights. The commerce vs. art/quality problem adheres to normal movies as well. Same for the frame rate. I would love to see more frames in any movie theatre. At 24p horizontal pans are dreadful. >>> you don't do it with pixels, but with resolution. >> You wouldn't (or shouldn't) compare a 640x480 image to one that is 1280x960 and say it is 4 times as high, you say it is twice as high.<<< Some people do and they're flat wrong, mathematically and optically. 1280x960 is 4 times as high as 640x480 in every respect. Christoph | ||
| lelie Rating: N/A |
Thanks for the nice comments. I appreciate them. Let's talk about resolution for just a moment. I don't like to give up the original meaning of the word. Perhaps I didn't read you right, so let's try again. 4 times Christoph? The data is 4 times higher. That's all. What about resolution for film? If I use a Kodak film that is 80 lines per millimeter and go to one that is 160 lines per millimeter (as rated by any film manufacturer), the resolution (that is LINEAR, i.e. 1 axis or 1 dimension) is TWICE as high. Look at lens charts; same thing. To use your 4 times logic, you would say my film is 4 times higher resolution. It has 4 times as much DATA, but the RESOLUTION is only twice as high. I know people get sloppy with the term resolution, but it is literally axial. Resolution was originally measured as I stated above, in lines/mm, (made it easier to compare resolution across film sizes) and was only truncated when the number of mm was fixed. Resolved in english originally meant to discriminate between two objects. The highest resolution media I ever used were Kodak 4"x5" spectroscopic plates; they were rated at 2000 lines/mm. They were of course B&W. They are meant to discriminate between spectroscopic lines, so linear was the only meaning. Lithographic films also only have linear meaning. It was never (lines squared)/(mm squared). Had they been, the data density would have been 5.16 E 10 pixels on each plate. Larry Elie | ||
| lelie Rating: N/A |
Almost forgot. Anyone who has seen 'Encounter' on the DVD, PLEASE SEE IT ON IMAX. It is also alias 3-D Mania. The comparison is something to behold. It is more of a stage show; the characters appear life size on stage. It isn't so much the 3-D is better, or the resolution, it is just that this is what the director intended, and you will never see it that way at home. Second, VR-- I think your TV has a sync-scan processor; it can sync on several types of things including blanking intervals. I have one on my Sony and it can actually mistake the black at the bottom of the screen on a LBX movie for a blanking interval. These puppy just love to sync, and the do a credible job on without the right inputs. I still suspect that you have what amounts to as progressive scans of only part the the 3D info on each eye. For example, progressive scans with perhaps 3/4 left eye with 1/4 right eye building into the left side then 3/4 right eye 1/4 left eye building to the right side. A frame averaged nightmare. Your brain finds enough 3D info even there to build a usable image. That's a guess of course, but that's the only way I can see progressive scan working at all. What happens when you pause? For that matter, what happens to any of you out there when you pause? I go flat instantly. Larry Elie | ||
| VRJUNKIE Rating: N/A |
Larry, Pause definitely destroys ALL the 3D... it has to since you are reviewing a single frame over and over. Slow motion gets varying degrees of 3D left in place for me. And on further inspection you may be right (or very close) about what's going on in progressive mode. There are brief times when it will get a small part of the frame wrong or flat (usually in the top half) - sometimes it goes back and forth depending on the scene - but then next time I watch it it may play the entite sequence in stereo correctly! It has done a perfect job one on Haunted Castle all the way thru a few times and it was fantastic. Viewing the 53" big screen from about 6 feet in stereo is awesome! Closest I'll get to I-max at home anyway! -VR | ||
| lelie Rating: N/A |
VR, A single 'frame' should consist of a right 'field' and a left 'field'. They can be shown repeatedly and be fieldwise correct. In progressive mode, you would always go flat. The problem is, I DON'T go progressive mode, and I still go flat. That's the part I don't understand. Larry Elie | ||
| VRJUNKIE Rating: N/A |
Larry, These DVD videos are not interlaced stereo - they are frame sequential stereo (this is in spite of the fact that the TV itself is interlaced). Each eye gets a full resolution image. The adapter just flashes the glasses at 30 hz 180 degrees out of phase. Since the frame rate is 60 hz this gives each eye a 30 hz image. For example every even frame would be for the left eye and every odd frame for the right eye. This is why there is no stereo when you pause because there is only one frame - and hence viewing angle to see. -VR | ||
| lelie Rating: N/A |
Thanks. To be 'consistent' still would black out one eye; not marketable, just consistent. That explains it. Larry Elie | ||
| Christoph Bungert (Admin) Rating: N/A |
>>> What about resolution for film? If I use a Kodak film that is 80 lines per millimeter and go to one that is 160 lines per millimeter (as rated by any film manufacturer), the resolution (that is LINEAR, i.e. 1 axis or 1 dimension) O.K. you got me there. You're right, in the past resolution was measured in lines. What bothers me about this is the single dimension. How to compare formats which have different aspect ratios? What about non-square pixels? Nowadays if you check the specs of a digital camera the 'resolution' is given in total pixels or in horizontal x vertical pixels. I should should use the term pixels from now on and leave 'resolution' alone. Christoph | ||
| Christoph Bungert (Admin) Rating: N/A |
>>>A single 'frame' should consist of a right 'field' and a left 'field'. Correct. A proper freeze-frame, like on a standard VHS-VCR, would consist of both fields, thus producing a still stereo-image. >>>The problem is, I DON'T go progressive mode, and I still go flat. That's the part I don't understand. Probably your DVD player only uses one field in freeze frame for interlaced material to avoid wipe-effect. My DVD-player for expample uses only one field when I play NTSC-interlaced material in PAL-output format and doubles the lines. The image only has half-resoltion and looks blocky. The 3D-DVD's turn into 2D. >>>These DVD videos are not interlaced stereo I think they are interlaced. I even believe that they use the special MPEG2 interlace compression mode. >>>they are frame sequential stereo (this is in spite of the fact that the TV itself is interlaced). Television is frame/field-sequential and line-sequential at the same time. Each field contains only either the even or odd lines. Actually that's what 'interlace' means. >>>Each eye gets a full resolution image. No, I don't think so. It looks like half-res per eye to me. Having 60 full resolution images instead of half-res would be against the DVD-Video and TV-system-standards. It wouldn't work on standard TV's and DVD-players. Christoph | ||
| VRJUNKIE Rating: N/A |
>I think they are interlaced. I even believe that >they use the special MPEG2 interlace compression >mode. But the "stereo" information is strictly frame sequential as it relates to pause. There is only one viewing angle in each frame. For anyone who may still be confused on why a paused view goes flat: Look at any stereo VHS tape or DVD without your glasses on... you will see a double image. This is the result of two frames (as perceived by your mind) as being on the screen at the same time. Now pause any stereo VHS tape or DVD... you will see a single image. Only one frame is being displayed - hence the term "freeze frame". A VHS tape player or DVD player displaying a single frame when paused is not the exception - it is the norm. >No, I don't think so. It looks like half-res per >eye to me. Having 60 full resolution images >instead of half-res would be against the DVD- >Video and TV-system-standards. It wouldn't work >on standard TV's and DVD-players. I may have misled/mispoke - what I meant to point out say was the freeze frame image is as "full" as any other tape or DVD.... it's not like looking at sync doubled display (window blind effect) on a computer. Or in other words, pausing one of these stereo format DVD frames will not yield a result of you seeing every other line as black on your screen - I was thinking Larry may have been wondering why he did not see every other line missing if he was only viewing one field. You are right though, the perceived resolution or clarity of the picture does indeed suffer. However, if you have a newer TV with any form of scan doubling technology - this is greatly compensated for. While it cannot generate missing information, it can smooth out the 'blockiness'. -VR | ||
| Christoph Bungert (Admin) Rating: N/A |
>>>For anyone who may still be confused on why a paused view goes flat I'm pretty shure you're not entirely right, but it's a long story to explain why. >>>There is only one viewing angle in each frame. Not exactly! One full-frame contains left and right eye view! In television each frame consists of 2 half-frames (or fields). The first half-frame contains the odd lines of a frame and the second half-frame contains the even lines. A second of NTSC-video consists of 60 half-frames. The head-drum of a NTSC-VCR rotates at 30 revs per second. It has two heads for the two half-frames. Now if you take a real old VCR and press pause the tape stops, but the two heads rotate and deliver both half-frames. Both half frames should be visible on the TV-screen and if it's a 3D-VHS tape you should see a 3D-still-image. If a movie is filmed with a video camera at 60 Hz and there is fast movement there's a time gap between both half-frames. If you did a freeze-frame there you saw a picture which 'jumped'. A 3D-tape would also produce a jumpy freeze frame. NOW IT COMES!!!!!!!!!!!! The trick is that the VCR-manufacturers (and later the DVD-player-manufacturers) at one point decided to throw away one of the half-frames in freeze-frame-mode to achieve a rock-steady still image. They only use the first half-frame and double the lines or at least keep the v-frequency stable (which would result in black lines) to get the full number of required lines to stay within the TV-standard specifications. (The doubling isn't neccessarily done by a digital buffer, there are also analogue tricks to do this.) That's why you can't see stereo. One of the first VCR's which did the trick was a Betamax machine by Toshiba I think. The reviewer wondered why the freeze frame doesn't 'jump' back then. I never forgot this. Half of the effective resolution is sacrificed in the process of course. I have a 3D-VHS test tape where the left-eye and right-eye view are marked as such. In freeze-frame I always only see the left-eye view no matter what I do (i.e. frame by frame advance), because they don't use the second half-frame which contains the right-eye view. The frames in television always come in pairs and the TV always needs a whole pair or it wouldn't work. The two halfs of a pair have a different timing, there's a one-line time-gap. These pairs are used for television-stereoscopy and that's the reason why the stereo-orientation almost never flips. You usually don't need the reverse button on your TV-shutterglasses-controller. On 3D-VHS and 3D-DVD each eye always only sees it's own half-frame, and therefore only half of the lines. This is definetely half-res-per-eye stuff! Christoph | ||
| GregK Rating: N/A |
To follow up on what Christoph said, when a VCR is paused, it typically only shows one of the two video fields. This ensures any movement between one field and another is not shown as a "jerky" freeze frame. A VCR field-freeze (yep, it's often called "freeze frame" by VCR manufacturers) will not have this potential problem with moving objects. DVD is another can O' worms, because the MPEG-2 bitstream can be flagged a few different ways. MPEG-2 can be flagged to show if a video was originally shot at 24fps, 25fps, 30fps, or as a 50 / 60 fields interlaced. Without getting knee -deep into all the hows and whys, this can effect how your DVD shows a still image. My Sony stand alone player only shows one field when I freeze my 3-D DVDs. But my Sony Playstation 2 DVD player has a unique frame setting: "Frame" or "Auto". "Auto" freezes the image however the MPEG flags dictate. But the “frame” mode shows both fields regardless of the MPEG flag instructions. So of course using the “frame” setting, the PS2 produced excellent 3-D stills from the Slingshot 3-D DVDs. As for the 3-D DVDs being half the resolution, I’ll take the middle ground. When each image is viewed by itself, yes, you’re only seeing 262 ½ lines ala NTSC. But we do watch 3-D DVDs with both eyes after all, so I think the reduced resolution is minimalized. Would a true 480p 60 frames per second NTSC video look better? Sure. ..But I wouldn’t think it would be anywhere near a 100% improvement for the reason I stated above. | ||
| M.H. Rating: N/A |
GregK: I thing it is 95% improvement. I have tested it in praxis with 800x1200 movies (800x600 per eye) in comaprison to 800x600 movie (800x300) per eye. Unfortunately the nessesery bitrates and computer decompression power + data storage space required are to much for existing HW (even GeForce3 and P III at 1G). | ||
| Craig Taylor Rating: N/A |
Christoph & GregK, I think VRJunkie is confusing the terms "frame" and "field", but your explanation as to why a 3D-VHS tape goes flat when paused is the best I've heard yet! | ||
| lelie Rating: N/A |
M.H. wrote; Just some notice of watching the DVD. I was watching all of this DVD by 3 mrethods: 1) On TV in the 30/60 Hz mode. H+ Real magic decoder in computer used to get the analog signal .. 2) In VFX3D HMD. S-VHS of the H+ Real Magic mpeg2 PC decoder used to get the signal 3) On 22' monitor in 60/120 Hz mode. mpeg2 streem reprocesed to above/below coded DIVIX4 before playback. Filtered OpenGL HW accelerated zooming used (P III at 1G and GeForce 3), experimental playback software I have enjoyed the method 2) as the best. Reasons: The HMD optic compenzate partialy the out of screen negative effct. A lot of white color in the movies produce too much ghosting for method 3). Hi flickering ocuure in method 1). I am realy looking fororward to some next generation HiRes HMD with stereoscopic support compensating the low-res of VFX3D. " How did you do 2? I have a Reel Magic, and I tried. I didn't expect a sync, and didn't get one. Larry Elie | ||
| M.H. Rating: N/A |
Larry: did you connect the S-VHS output of the Real Magic decoder with the S-VHS input of the VFX3D control box ? Did you use NTSC output setting in Real Magic ? Did you activated manualy the stereo mode of VFX3D ? It works perfectly for me ... | ||
| Christoph Bungert (Admin) Rating: N/A |
Just found a press release for a chip for 100Hz TV which is supposed to be compatible with 3D: http://www.micronas.com/press/pressreleases/pressrelease.php?s=1&ID=171 Christoph | ||
| M.H. Rating: N/A |
Christoph: Yes !!! Finaly !!! Another small wish: fast phosphors for such TV | ||
| Larry Elie Rating: N/A |
Christoph, fast phosphors USED to be easier than slow ones; balancing the phosphors persistance and brightness was the hard part. | ||
| ronnie stuart Rating: N/A |
is anyone attending the SPIE Photonics West Stereoscopic conferance in San Jose CALIF this week? if so please post info on new stereoscopic ,hardware, software etc at this show. thanks or e-mail me at d2d3dron@webtv.net | ||
| Anonymous Rating: N/A |
IMAX For $79 http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=1516607375 | ||
| Anonymous Rating: N/A |
Or Buy Directly With Paypal for $79 iMAX 3D KIT: USA Shipping From California http://www.nvidia3d.net/imax/ | ||
| dragoncarver Rating: N/A |
VR I have been looking for a way to trick my new sony 11HT into playing 3D. If I understand you right ; Set the player unit to progressive/ (not sure I can. ... but the projector can) then feed my little black box off of anouther video source. ( everything is triggered from the input power signal of 60 cycles....) and with luck The glasses will sync ? Can't hurt to try !!!!! | ||
| rainer emmerich Rating: N/A |
Hallo iam rainer from germany , my problem is I have a plasma tv and a 3D system with shutterglasses , but it works not so good. the glasses are from razor.is there anybody who have an idea . what can i to | ||
| Puppet Kite Kid Rating: N/A |
"Anaglyph" to many people is still a dirty word, but I am able to make anaglyphs that are comparable to any other type of 3D, and although it is an "art form", it is not a mysterious science. Anyone could do it. That was the good news... now the bad news: You will have to confirm the following, because I have only confirmed it with LCD TV's, not plasma. I am excited to hear if it applies to plasma, so please respond, if possible. I can't think of any reason why it won't work perfectly with plasma: You have to process the anaglyph on-the-fly on a computer and output the results via VGA (types) or DVI. Don't try other connectors, because they will produce much less than perfect or very poor imagery. There is an easy way, using freeware on a PC, but it will be at the expensive of some (not much) retinal rivalry and poor color choices. Still, the 3D integrity should be perfect. Open the interlaced 3D movie in Stereoscopic Player or StereoMovie Maker and play it as "half color" red/cyan anaglyph (my personal choice). You should be able to play the movie directly from a DVD with Stereoscopic Player. Here are the steps, using an MPEG file: Stereoscopic Player: http://www.puppetkites.net/misc/SP_howto.htm StereoMovie Player: http://www.puppetkites.net/misc/SMP_howto.htm The more difficult way is also the best way, and with some work can produce virtually perfect anaglyphic imagery. I won't go through the details, here, but can answer any specific questions, if desired. You basically have to analyze a movie scene by scene and make appropriate changes. There are also a few different format choices. This is obviously time consuming, depending on the degree of perfection/artistry, and is geared more for people wanting to create anaglyphic movie content and not convert complete existing commercial 3D movies. Of course, however, it can be done: VirtualDub3D: http://www.puppetkites.net/virtualdub3d.htm P. K. Kid Non-commercial stereoscopic 3D video: (All G-Rated) http://www.puppetkites.net | ||
| Puppet Kite Kid Rating: N/A |
Oh, whoops :-) Those instructions are for my parallel videos on my web page. Hey, you can try those, too :-) But, you have to use "Interleaved" or "Interlaced" as the input file type for field sequential DVD's. Also, if you use the VirtualDub3D instructions, those are for parallel videos, too. An interlaced source is actually easier. You just _ignore_ these two steps: 2) Add the "Deinterlace" filter and "Fold Side-By-Side Fields Together". -and- 5) Add the "Resize" filter and change it to the size you want. (half the height) PKK | ||
| Peter Wimmer Rating: N/A |
Hi Rainer, I don't have a plasma TV myself, but as far as I know, most plasmas TV don't work well in 3D. I believe the reason is that plasma displays are progressive displays which deinterlace the video before displaying it (using some kind of adaptive deinterlacer). Deinterlacing blends the left and right fields together and therefore destroys the 3D effect. 3D-TV works well on 50 Hz CRT-TVs only, 100 Hz TVs also apply deinterlacing and can't be used for 3D (except some very old ones). On 50 Hz TVs, you'll perceive a 25 Hz per eye flicker which causes headache after a few minutes. The best way to watch 3D videos is using shutterglasses on a CRT-PC-Monitor running at 100 Hz or higher. If your shutterglasses are not suitable for PC usage or if you own a LCD, use anaglyphs instead. Peter | ||
| Puppet Kite Kid Rating: N/A |
I was looking at plasma TV's again yesterday, and VGA inputs (or "PC input") seem to be very common... almost standard. This is great news, and I'm hoping that Stereoscopic Player or StereoMovie Player will work perfectly this way with anaglyphic imagery. Again, I recommend using the "half color" views for testing to reduce retinal rivalry. Please, when someone tries this, let us know, so we can forever add this to our stereoscopic viewing option repertoire :-) It might just totally blow our socks off ;-) I've got 16:9 segments on my web site... just remember they are low resolution, made with cheap digital camcorders, but the 3D effect should still work perfectly for testing. It would also be nice to know if the "DVI input" works, too, for this purpose, because I think I've seen a couple of plasma TV's with a DVI input but no actual "PC input". Since my video card has a DVI output, I bet it works, too. I use it for my LCD TV with perfect results, but I use a DVI to VGA converter (cheap). You'll have to match the DVI connectors, too, as there are three different types. Obviously, you might be able to use a VGA to DVI converter, too (the opposite of what I have). Again, I wouldn't even bother trying any other connectors. In some situations, some things might work to a certain degree, but I do know _for sure_ that VGA (types) and DVI _is_ capable of 100% perfect 3D... What I don't know for sure is if that is true for plasma TV's. I can't think of any reason why not :-) PKK http://www.puppetkites.net | ||
| Anonymous Rating: N/A |
Do you loose any resolution by connecting direct to a plama? Is it as clear as a LCD or CRT PC Monitor? Can I read e-mail? | ||
| Puppet Kite Kid Rating: N/A |
I have seen a PC hooked up to a plasma monitor, but I haven't seen an anaglyph. To me, it just looked like *any* PC monitor, but simply bigger than usual. It also had very impressive clarity. Plasma TV's look amazingly great to me. In this case, it was used for displaying a certain brand of computers in a shopping mall. You have to consider that a large plasma monitor would typically be used for viewing from a distance, so even a loss of some resolution may not matter when viewed from a typical viewing distance, i.e., you are not going to be looking at it from two feet away. Also, aren't some (most?) plasma screens 1280x768? That would be the "widescreeen equivalent" of the typical 1024x768 display of a PC. (That's what I use, these days) Other than the horizontal stretch from 4:3 to 16:9 (widescreen), you aren't losing any resolution. Also, BTW, MPEG's stretch very well, and this is exactly how "anamorphic DVD's" work. The native file is in 4:3 format, but is stretched horizontally to 16:9. Yes, you will be able to read email while sitting in your easy chair, 15 feet away from the TV :-) PKK | ||
| M.H. Rating: N/A |
I am afraid plasma dispaly scan+digitalize the input VGA signal and modify it to the internal fixed frequency on witch they work. This means that page flipping stereoscopy can not work. It needs testing in praxis. Anybody have a plasma display at home ? | ||
| Alatar Rating: N/A |
>Anybody have a plasma display at home ? Not plasma, but large screen LCD TV, yes (a Sharp Aquos 37"). It does have a DVI-in connector and works basically like a 1366 x 768 LCD monitor when driven by a PC. I haven't tried it for any kind of stereo yet, mostly because I am not very optimistic about the results: - anaglyph will look as lousy as always - the LCD will be too slow for alternate-eye - polarized with a full-screen polarizing panel (I have a Nuvision 21" one) won't work as the LCD output is prepolarized. | ||
| M.H. Rating: N/A |
LCD can not work. It is to slow and in this case it is 100% sure it rescan the input frequency. Plasma is on the other hand eqauivalently fast as CRT. I had heard some rumors that plasma works, but it is not easy (direct sync of LCD glasses to plasma monitor base don phototransistors) ... | ||
| Alatar Rating: N/A |
> The 'IMAX-3D Disease' addresses the excessive use of large negative parallax in all IMAX-3D movies, not these DVD's in particular. Being new here, can someone explain this to me in more detail? Is this referring to IMAX theaters, or to IMAX 3D films which are -- through one route or another -- viewed in a home 3D system? Are there links or posts here which discuss this in more detail (as I don't seem to have a working search function)? I am in a position to know quite a lot about IMAX 3D and can assure you that they do not allow this to happen on purpose. | ||
| M.H. Rating: N/A |
Alatar: Description of the problem is not trivial and it require deep knowildge of human eye physiology and larg screen stereoscopic projection geometry. It is a very big diference whatever you watch IMAX movie in IMAX or at home at PC monitor. They key geometrical factos are: user distance from the screen, screen size, maximal left and right eye separation of object comming out of screen (producing so coled positive divergence). Positiv divergence is something what can not happen in real world. Left eye is heading not forward but a bit to left, rigth eye to the right: \ /. Normaly the maximal divergence is zero ||. Average maximal physilogicaly acceptable positive eye divergence is about 1.5 degree. From the standard cosine equation you can easy calculate what is the maximal object separation (producing out of screen effect) in dependece on user distance from the screen. A lot of older IMAX movies exceed this limit. The main reason is mechanical trouble of IAMX recording devices - the can not put for some scenes the camera objectives close enougth to reduce the separation. But with the new recording camera setups (Para cam stereo-camera montage, complety CG generated movies) this problem is slowly disapering ... Let me know if you need more technical details, equations .... | ||
| Alatar Rating: N/A |
Well, you may be surprised to hear this, but I do know all this, as I have been designing 3D viewing systems for more than a decade. My question is *where* people have seen Imax content exhibiting what you call positive divergence (and is somewhat loosely called hyper-stereo around here, although that is admittedly a misnomer). Imax is very careful to control the exact allowable angles of this divergence in their films and theaters -- although for theaters with two-projector systems nothing prevents the projectionist from misadjusting it -- and its use is very limited to scenes that are deliberately intended to have an exagerated sense of depth. Even then, it is kept to a strict maximum of about 1.5 degrees. The problem arises, as I imagine you know, when content is shown on a screen of unforeknown size (i.e. a home theater system), and only with objects intended to be rendered behind the screen. For example, something calculated to appear at infinity (and thus with an image shift of 55m) on a 33" television screen would have a shift of 220mm on a 132" screen, resulting in the viewer having to go wall-eyed to converge their eyes on it. This difficulty is one of the major issues with creating generalized content for 3D viewing systems. Imax designers sidestep this by having carefully controlled projetion and viewing geometry in their theaters, but they have less control over how their content is viewed in home settings. | ||
| Alatar Rating: N/A |
PS, by the way, I forgot to thank for your kind response to my initial question. I am glad to have found a board on which some of these questions can be discussed less formally than at various conferences etc. You will have guessed that "image shift of 55m" should have read "image shift of 55mm" in my post. | ||
| M.H. Rating: N/A |
Alatar: The 1.5 degree deivergence was totaly exceeded several times e.g. during screening following titles: "Into the deep", "Tyranosaurus Rex 3D" as screened in IMAX theater in Prague Czech Republic. I do not have exact nuber but I queees it was 5degree in the middle of the teatre. I will take with me a thodeolit for measuring the angles exactly next time :-) | ||
| Alatar Rating: N/A |
M.H. Thanks for that info, I will pass it on to the people at Imax who worry about such things. "Into the Deep" was a very early Imax 3D film and has a number of stereo errors and problems. All the shooting was done with two discrete cameras on a special mount, and that makes possible mis-alignment a real concern. The issue here is that you only discover the problems when it's too late to go back and reshoot. I can recall at least one incident (on another film) in which *all* the material from some very expensive location shooting had to be thrown away due to misalignment in that rig. T-Rex is much newer and shouldn't have any such errors. I believe it was all shot with 3D cameras and it should be physically impossible for them to be mis-aligned that way (the two lenses are locked together with parallel axes, I believe). Still, it's conceivable that they had a few shots with accidental divergence which they let through, but I know the filmmakers were well aware of this issue at the time it was made. I'll try to track down whether this is on the negative, is a print problem, or is specific to that theatre. Call you recall any specific scenes? Alatar | ||
| M.H. Rating: N/A |
Alatar: Into the Deep - all secenes shoted from very short distance had non-acceptable parallaxes. The reason was probably non-adjutable camera distance . Accrding T-Rex - have a look on the movie yourself, I did not make notice about the specific scenes ... I can confirm that all other Imax movies I have seen vere looking O.K. (Space Station 3D, Cyberworld 3D, Santa versus Snow Man) .... | ||
| Christoph Bungert (Admin) Rating: N/A |
It's not O.K. Since the invention of 3D-photography and film there is a dispute about where to place the 'virtual window', the plane where the parallax is zero. If the 'virtual window' is placed far in the back most of the objects in the scene are in front of the screen and vice versa. However real stereo experts with decades of experience in shooting and projecting 3D footage place the virtual window in the front with most objects inside the screen and very limited out-of-screen effects. Objects which come out of the screen should not be allowed to touch the border of the screen. This fact usually limits the use of out-of-screen effects to smaller objects which stick out just in the middle. Scenes with no or little out-of-screen effects cause much less eyestrain and usually less ghosting. For example, each year the best european stereo slides of the year are presented on the german "Raumbildtag". You can watch excellent stereo slides all day long without getting headaches, opposed to IMAX theatres. There is usually very little use of out-of-screen, but the stereo-effect is natural, amazing and beautyful. As I said many of the guys have decades of experience. The people who do IMAX-3D movies usually have less experience with 3D-shooting. They are in for the money and try to achieve the maximum short term effect on the audience. Putting everything in front of the screen just isn't a good idea. Christoph | ||
| M.H. Rating: N/A |
Christoph: Both Cyberword and Sanat Claus ver done in CG by computrer specialist witch now something about correct stereoscopic scene geometry. This movies looks better than the normal camera shooted IMAX stuff ... According the out of screen effect - the problem is not out of screen (interfering with acoomodation problems) but exceeding the maximal acceptable acceptable positive angluar parallax as already mentioned ... Fo IMAX and projection thigns do not work identicaly as for PC ... I did not beleve this before I did not make a lot of practical testing ... | ||
| Puppet Kite Kid Rating: N/A |
Am I missing something here? If you set the stereo window to the closest object (as you probably should with stereoscopes, slide viewers and with small monitors to avoid "window violations") and you project this imagery this way, as soon as the distant objects reach eye distance apart, your eyes have to diverge to merge the left and right images. With projection (like at an Imax theater), this can happen very quickly (as soon as they are 63mm apart!!!). Now, with natural vision, we always cross our eyes, and never diverge our eyes. Think about it... this means that during projection, the stereo window has to _always_ be set to infinity (or the farthest object) to avoid eye divergence. And we wonder why we get a headache? So much for window violations... those are the least of our worries... PKK | ||
| Puppet Kite Kid Rating: N/A |
In the photo-3d Yahoo group, "DrT" just replied with the magic formula. Imax fans, we have a problem :-) George says: "Here is one way to calculate things... The recommended amount of on-film deviation for 35mm film is 1.2mm (can be as large as 2mm). In a typical stereo club setup using 72" screen where only about 50" height is used (to accommodate 7p mounts) then the magnification is roughly 50x (1" inch of height - 23mm to be exact, the height of unmasked 35mm film is 24mm - is projected to a 50" size). 50 x 1.2mm = 60mm hence we are safe. 50 x 1.5mm = 75mm, should be OK too. Any larger projection size or on-film deviation will lead to divergence of the eyes." And my response: Thanks. This is what I was looking for :-) How big is the Imax screen? (ahem ;-) The deviation math? ...anyone? What percentage of the Imax imagery _has to_ come out in front of the screen to avoid divergence (or let's say "excessive divergence", whatever that is ;-)? Comin' at ya ;-) PKK | ||
| Alatar Rating: N/A |
I think that each of the above explanations is lacking something in its approach to explaining how Imax goes about this -- or it maybe that some are explaining it properly, and I'm just not understanding the post correctly. You will have to take my word for it (or not, if you prefer) that I understand it. You have to remember that Imax content, especially digital content, is specially created for their theaters. There is no relation to 35mm film, except insomuch as it there may be a printdown later (which Imax always discourages). There also is no "stereo plane" per se. All one starts with is the Imax screen and the viewer's two eyes. Four lines, each one originating at the right eye and extending through one corner of the screen, form an infinitely large pyramid, the base of which is perpendicular and at infinite distance. A similar pyramid is drawn for the left eye. The intersection of these two pyramids is where elements of the scene "work" in stereo: anything in only one pyramid is clipped in one eye. Now consider Imax 3D filmmaker -- if he wants an object (let's use a point to keep it simple) to appear at the same depth as the screen, he renders that point so both eyes display the point at the same location on the screen. If he wants his point appear at infinity, he renders so that the point in the left eye shows 55mm to the *left* of where it does in the right eye, and if he wants it to appear closer, it is rendered so that the left eye point shows to the *right* of the right eye. He can bring it closer by moving the left point to the right (and vice versa, up until the point leaves the intersection of the two pyramids above, which is another way of saying "until the point leaves one frame or the other". This all works because Imax is completely "master of their own house" -- their content is all shown in theaters with the nearly the same viewing angles and roughly the same distances. The filmmaker can take this as a given and therefore the rendering geometries are both simple and constant. It is worth noting that from the filmmaker's point of view, one can assume that the projector and lense geometry is irrelevant, all that matters is that projection system orthagonally places the film frame onto the screen so that the frame entirely fills the screen -- it could be done thru back projection and mirrors, it just doesn't matter. Disclaimer: the above explanation is full and correct as far as it goes. There are a couple of additional elements that would make it clearer to explain, but I am not certain these elements are in the public domain. I will check, and clarify on Monday if I am able. | ||
| Alatar Rating: N/A |
I think the point I am trying to make is this: There is a well-understood mapping which tells where a point should be rendered in each eye, given the three-space location you want it to appear at in the theater. This mapping is constant, as all theaters are the same. Essentially, three-D filmmakers design their shots in three-space, run it through these mappings and voila, you get two rendered frames ready for the film recorder. Simple! No need to worry about stereo planes and such, those are all things that are only needed by people who are trying to display content at some *other* display geometry than it was rendered for. | ||
| Alatar Rating: N/A |
Quote ----- The people who do IMAX-3D movies usually have less experience with 3D-shooting. They are in for the money and try to achieve the maximum short term effect on the audience. Putting everything in front of the screen just isn't a good idea. ----- In general, I think it is fair to say that things are exactly where the director / stereographer wanted at time it was shot. He may, of course, have changed his mind when he first saw it in the theater. Somewhat sadly for technophiles and purists (like us), the content created by for-profit organizations like Imax is neccessarily going to be determined by what sells. The phenomena of directors using ludicrously flashy, almost painfully overdone effects and shot design is not limited to Imax. Just look at any 3D action movie. | ||
| Puppet Kite Kid Rating: N/A |
Check this out: International Stereoscopic Union, The Projection Rule: http://www.stereoscopy.org.uk/Pages/GoldenRules/GoldenRules3.html "No person viewing a projected stereo image should be forced to diverge his eyes." What I conclude from this is that the left and right images can never be spaced farther apart than 63mm. i.e., the distant of your eyes. What Dr. T was trying to explain is how this calculates with 35mm film and a (small) 72 inch screen. He says that this imposes close to the limit, and a larger screen could easily "break the projection rule" if you are not careful to keep the separation less than 63mm. If you think about it, with a huge screen like an Imax screen, this pretty much forces the images to be "mounted" near a point that's almost at infinity. I am leaving for Japan for two weeks, but I will leave with this following demonstration. Of course, I'm just guessing, and I could be wrong... or at least wrong about some aspect or another :-) http://www.puppetkites.net/projectiontemp/image1.jpg This is a red/cyan anaglyph which I have "mounted" and set the stereo window to what I think is a typical placement for monitor viewing. Use your mouse cursor to check where the stereo window is. You cursor should appear to be at the same depth as your monitor, and the nearest point in the photo is a ways beyond that point, behind the stereo window. Look at the distant vertical objects in the photo on the far upper left of the photo... those poles sticking up in the air. Notice the separation of the red (left) and cyan (right) images. If I were to project this photo onto a large screen, this separation might (at some point) exceed 63mm and cause people to have to diverge their eyes, thus breaking the "projection rule". So, depending on the size of the screen, I might have to do something similar to this: http://www.puppetkites.net/projectiontemp/image2.jpg For this demonstration, I have simply moved those two distant poles completely together, to avoid any possible divergence of the eyes of the audience looking at in on the large screen. Of course, that is the extreme, and they wouldn't have to be closer than 63mm. (At this point, I don't know the actual size of the screen.) I realize, in practice, you don't change the mounting for the projection, but I'm just trying to prove that the same thing results... I now have changed the stereo window, correct? Look at your cursor, again. It is now is at the same depth as infinity (or far object) in the photo, and everything in the photo has "popped out in front" of the stereo window, correct? This is identical to shifting the lenses of the projectors, correct? I cropped of the left and right sides of the second view, after shifting the images, otherwise you'd see a red strip on one side of only one eye's view, or a cyan strip on the other side of the other eye's view. So, using this image as an example, what is the difference between shifting images on a computer screen and shifting projector lenses? Both should change the placement of the stereo window! Again, my original point was that with a large screen like an Imax screen, you would have to shift the projector lenses to line up the point near infinity to avoid eye divergence, and this will change the stereo window accordingly, thus making *most* objects in the scene "pop out in front of the stereo window". PKK | ||
| Alatar Rating: N/A |
Ok, I think I understand how you are looking at this differently than I do -- I want to say in advance that both paradigms are valid and I do not disagree with what you have said. It has been my experience that once people start looking at it in the way I am trying to explain, it becomes much simpler and easier to visualize. It is important to understand that the following is *not* relevant to people here who are trying to playback existing content; it is about how to create 3D content. The idea that some people carry over from 2D imaging that it is possible to scale up 3D images the same way as you can in 2D is simply erroneous. You cannot "blow up" -- even from a 17" to a 21" monitor -- a 3D pair without (sometimes only subtly) changing the stereo relationships between various elements in the picture. You would have to also scale the viewer's eye displacement and his distance from the screen as well. You are correct that *if* someone were to take images that were created for 35mm projection or a computer monitor and simply blow them up to Imax size the result would be unviewable; but no-one would ever do that. Anything done digitally for an Imax screen is carefully calculated so that the maximum divergence of any object on the *Imax screen* does not exceded 55mm (or 62, if you prefer). Actuality filming is a different kettle of fish; as I said in my previous post, I have to check what I can say about this before explaining in more detail. You have to remember that they are creating the film for Imax theaters. You really must realize that the "stereo window" you refer to is a red herring and limits the discussion. Whether an object appears in front of or behind the physical screen is more-or-less incidental and is simply a question -- as mentioned above by M.H. -- of whether the particular item is diplayed with negative or positive divergence. If I want an object to appear 5 meters away, and my screen is 10 meters away, I'm going to have to display it with negative divergence, but if the screen is 2 meters away, I will need positive divergence. Of course, the exact number of pixels of divergence depends on the render resolution and the screen size as well. In other words, what you call the stereo window is simply that plane at which objects get rendered with left- and right-eye elements in the same horizontal position. The point is that the divergence is the output of the function; not the input. What I have explained above is from the point of view of digital content, and actuallity filming is a different kettle of fish. However, the mathematics are the same. I hope to go into more detail about this tomorrow. Have a good trip to Japan. :-) | ||
| Alatar Rating: N/A |
I now have a bit more time, so I will try to go through your answer point by point: >"No person viewing a projected stereo image should be forced to diverge his eyes." Agreed, although up to about 1-2 degrees tends to be acceptable to most people for brief periods of time, and is used for dramatic effect on occasion, usually when one wants the "back to fall out". One example is the final shot of "Paint Misbehavin'", when there is a cut to outer space. But this is a rare exception to the rule. > What I conclude from this is that the left and right images can never be spaced farther apart than 63mm. i.e., the distant of your eyes. Agreed. > What Dr. T was trying to explain is how this calculates with 35mm film and a (small) 72 inch screen. He says that this imposes close to the limit, and a larger screen could easily "break the projection rule" if you are not careful to keep the separation less than 63mm. Agreed, but I hope we both agree that it is mathematically and geometrically wrong to simply jump to a larger screen size without recalculating, or at least understanding what is going to happen with, the stereo. In any case this isn't an issue with Imax content, as it was never in a smaller format to be enlarged from, it was made from the start for the large screen. > If you think about it, with a huge screen like an Imax screen, this pretty much forces the images to be "mounted" near a point that's almost at infinity. Now here is the first place that we differ :-) Not because you are wrong, but because adopting the paradigm of "mounting" images adds a bunch of baggage that just confuses things. The two images should just reflect the ray-traced reality of what is being displayed. > I'm just guessing, and I could be wrong... or at least wrong about some aspect or another :-) this is a red/cyan anaglyph which I have "mounted" and set the stereo window to what I think is a typical placement for monitor viewing. Use your mouse cursor to check where the stereo window is. You cursor should appear to be at the same depth as your monitor, and the nearest point in the photo is a ways beyond that point, behind the stereo window. Look at the distant vertical objects in the photo on the far upper left of the photo... those poles sticking up in the air. Notice the separation of the red (left) and cyan (right) images. If I were to project this photo onto a large screen, this separation might (at some point) exceed 63mm and cause people to have to diverge their eyes, thus breaking the "projection rule". So, depending on the size of the screen, I might have to do something similar to this: http://www.puppetkites.net/projectiontemp/image2.jpg For this demonstration, I have simply moved those two distant poles completely together, to avoid any possible divergence of the eyes of the audience looking at in on the large screen. Of course, that is the extreme, and they wouldn't have to be closer than 63mm. (At this point, I don't know the actual size of the screen.) I realize, in practice, you don't change the mounting for the projection, but I'm just trying to prove that the same thing results... I now have changed the stereo window, correct? Look at your cursor, again. It is now is at the same depth as infinity (or far object) in the photo, and everything in the photo has "popped out in front" of the stereo window, correct? This is identical to shifting the lenses of the projectors, correct? I cropped of the left and right sides of the second view, after shifting the images, otherwise you'd see a red strip on one side of only one eye's view, or a cyan strip on the other side of the other eye's view. Hmmm, its hard to know how to divide this into separate points, so I will try to answer it as a whole. I understand the demonstration, and as I said above it is one (valid) way of looking at the issue, but it is overly complicated and does not easily lend itself to mathematical analysis. In your first image, the poles are about 3mm apart on my 21" monitor (at 1600x1200), and about 10mm if I enlarge it to fill the screen. It is, of course, trivially obvious that they will be further and further apart on increasing screen sizes, and that at some point will pass 62mm. This is a simple demonstration of why one cannot enlarge this type of image. In fact, the only way one *can* guarantee that this kind of image can be enlarged is to keep everything at the screen plane or in front, i.e. keep all the convergences positive. But here you see the limitation of this kind of approach: a stereo system that can't show objects at any depth it chooses is too limited to be very useful. Better to devise a system that calculates the correct divergence for objects at infinity *already knowing* what the screen size and audience position is going to be. > So, using this image as an example, what is the difference between shifting images on a computer screen and shifting projector lenses? Both should change the placement of the stereo window! There isn't any difference, but that's not the point. The point is that you don't need to do it. If the Imax screen was a giant LCD panel, the displacements would all be calculated (correctly, one hopes) to show objects where they were supposed to be. This is the way the system is set up, and nobody needs to worry or think about toe-in or toe-out at all. But the key point is that it works this way because it is not intended to be shown at any other size later. > Again, my original point was that with a large screen like an Imax screen, you would have to shift the projector lenses to line up the point near infinity to avoid eye divergence, Well, in fact they can be pointing any direction one likes, as long as the filmmaker knows what it is, and every theater is the same. But I do follow the point. > and this will change the stereo window accordingly, It would, but change from what? Please don't take this the wrong way, but I have the impression that you are thinking in terms of showing content that was done with some other geometry for some previous purpose. It is right for an Imax screen because it was shot that way. It would look lousy on a 35mm print, and even worse on a home theater system without going through stereo correction. But unless one rerenders, stereo correction can only be mediocre at best, and the framing and other aspects are all wrong anyway, so you can't really compare it terms of quality. > thus making *most* objects in the scene "pop out in front of the stereo window". Well, again, I think you are confusing the issue unneccessarily by using the term "stereo window". Obviously, if you are watching a 3D movie in Imax, more of the action is going to happen in front of the screen rather than behind; this is because the screen is much further away than, for example, a computer monitor. One of the goals of all Imax presentations -- both 2D and 3D -- is to diminish the presence of the screen to the maximum extent possible. It isn't supposed to matter whether it's 10m away, 100 away, or at infinity, because you're not supposed to be aware of it. But the action is supposed to stay at the same place in the theater (mostly, this is a whole topic in itself). But in all cases, objects intended to appear at infinity have to be presented on that screen exactly 55mm apart, and that is exactly what the system is designed to do. | ||
| Peter Wimmer Rating: N/A |
IMHO, the stereo window IS crucial for a good viewing experiance. If any object in front of the stereo window intersects with the image border, the sterescopic effect gets disturbed and the image is unpleasent to view. However, due to the large viewing distance in IMAX theaters, objects have to appear in front of the stereo window to get an reasonable extent of depth. Personally I like IMAX very much and think IMAX films pleasent to view. The reason that window violations don't disturb the stereo image is that the screen is so large that you don't see the border! Sometimes, IMAX movies show exaggerated depth - but is this really a problem? I think no, because you don't watch IMAX movies a whole day long. 2D movies also contain exaggerated low frequency effects in the soundtrack and nobody complains about it. Why? Because people like the sensation and don't care that it might cause headache after a few hours. When viewing IMAX movies at home, the stereo window must be moved much closer to the audience (relative to the virtual sceen), so that some objects appearing in front of the stereo window in IMAX theaters appear behind it on your monitor or home cinema. By the way, I disagree that you cannot simply blow up a stereo image from a 17" to a 21" monitor. Even when scaling from 17" to 3 meters, you just have to adjust parallax to avoid divergence. No need to change recording settings or something else. IMAX movies not only look great in IMAX theaters, but on home cinema systems too. So no need to make stereoscopy more complicated than it is! PS: Since this thread originally was about Slingshot DVDs: I've heard Slingshot ran out of money, can anybody confirm this? | ||
| Alatar Rating: N/A |
Yes, it's true that the stereo window does play a role in that any clipping that happens in front of the stereo window is unsettling to say the least. I guess since most of my stereo viewing is done in real or simulated Imax theaters I don't appreciate how important it is in other viewing setups. It has been a goal since day one in Imax to try to get the edge of the screen past the edge of peripheral vision. The screen edge is a very strong reference point for viewers and as soon as it is removed, the sense of reality is significantly enhanced. You are right about about a 17" vs. a 21" monitor from a subjective point of view, obviously it works just fine. But from an objective point of view, all the numbers change (sightly), and this makes everything move in 3-space. To understand this, consider two points, A (at infinity) and B (60cm away), viewed on a 17" monitor 60cm away. In this case, the divergence for A is 62mm and for B is 0mm. Now we blow it up onto a 21" monitor. The divergence for A is 62mm * 21"/17" = 77mm and for B is still 0. So (using the image shift approach that I am trying to keep away from) we shift the images 15mm to get A back to 62mm -- now B is at -15mm, or well in front of the screen. :-( The above is only the case if you put the 21" monitor at the same viewing distance as the 17" one was. However, if you move it away from you, a similar set of difficulties causes similar distortions, *until* you back it up to the point where the viewing angle is the same as it was before -- at which point, you are looking at it in the way I have been promoting all along :-) | ||
| Alatar Rating: N/A |
I guess I should add a note here to say that I do not work for Imax, and certainly do not speak for them. It would be fair to say that I have been associated with them in various ways for a long time. |
