Recent Comments:
US looking to deploy long-endurance hybrid airship over Afghanistan {Engadget}
Sep 23rd 2009 6:40AM I thought the Hindenberg could no longer float because it's outer skin could no longer contain the gasses within. i.e. if it was filled with helium then a great big hole in it would still have caused it to crash.
UK film critic doesn't understand using HDTV to watch old movies, why kids are on his lawn {Engadget}
Jul 17th 2009 5:50AM I wonder what he would say when I currently have some scripts scanning 3 strip Technicolor on a 6K RGB line array scanner in greater than 10 logarithmic bits... outputting only 4K though. perhaps we're wasting our time, I must go up and tell them to stop it and switch instead to photographing it with a old home video camera....
3D: is this the resurgence that counts? {Engadget HD}
Jan 27th 2009 8:11AM To clarify the 75% loss is what happens after you boost the light output of the projector and you use a high gain silver screen... simply using two polarisers isn't that good :-)
3D: is this the resurgence that counts? {Engadget HD}
Jan 27th 2009 8:06AM To the comment about light loss almost all of the current state of the art Cinema systems for 3D loose about the same amount of light which is about a 75% loss - think two polarisers at about 50% loss each, although similar results occur for the other methods e.g. Infitec.
The really big win with today's systems is that the digital projection systems can use a single projector using frame sequential images at high frame rates, this means the two eyes see both images registered correctly and the projected image is pretty stable and potentially flicker free. For me they need more than 2K resolution which they currently use.
Personally if the images suffer any kind of problem I usually start getting headaches. This includes shifting focus depths across cuts. Given this bias I still think that 3D will have a wider appeal this time, although I believe that it may only be 10-20% of movies. For me the real success will be other 'events' (live or otherwise) e.g. sports, concerts, etc that will really work.
Eyes-on with Hitachi Super Resolution TV {Engadget HD}
Jan 13th 2009 11:51AM Though not stated, implied in the name of super resolution is sampling not only in the spacial domain, i.e. in the image being displayed, but also temporal i.e.previous and possibly subsequent frames, done correctly its quite surprising just how much you can get from other images of similar content.
For instance consider a Bayer pattern array, you could consider this to be 4 different signals (slightly offset and selecting varying parts of the visible spectrum) which if you know how can be recombined to give a better picture than their native resolution would suggest. Alternatively consider what stochastic multi-sampling in games does to the quality of the images.
Admittedly in the case of streaming media, you'd probably only be able to implement a small number of temporal samples without adding a noticeable amount of lag to the signal, but if you delay the audio too then all should be good.
Another example would be SIGGRAPH last year had a paper or two on the concept of 'upresing' SD content using a single high resolution image of the scene (and a lot of image processing).
The Dark Knight Blu-ray disc leaking out, complaints already rolling in {Engadget HD}
Nov 25th 2008 5:43AM The key to determine where the 'artifacts' come from is knowing how the film was made, a large portion of the film was made from the film negative directly, with no 'DI'. Obviously the VFX sequences were scanned either from IMAX, or 35mm, and recorded back onto both 35mm and IMAX through different routes. These were send for printing and went to the theater.
If you find the effect on 'straight shots', then it could come from the telecine/scanning, or in the mastering of the HD master (noise reducing/edge enhancement/colour correction/etc) or in the compression, or in any number of places post the disc (like the decoder/player/tv!)
Or for a really big spanner, it could come from the original film stocks which have edge enhancing technology inbuilt as part of their designs!
Does the film look good?
Do you like the film?
then buy it, of not...
Ericsson: 20 megapixel cellphones shooting Full HD video in 4 years {Engadget HD}
Nov 7th 2008 10:58AM squiggleslash, actually I wasn't very clear on that, I was more thinking of data rate than capacity, 20 mega pixels at say 30FPS, even if your only doing 8 bit bayer type sensor, that is about 4.8 Gbits/s off the sensor into RAM prior to compressing... AVCHD (max of about 25Mbits/s) being what it is is fine for use on SDHC cards but for larger formats you'd need either some pretty remarkable compression to be invented, or to use the oversized heat sink as your RF antenna :-)
Optically, they could do something along the lines of a micro lens array/light field capture and do some computational photography (on camera) - thus doing away with a focusing system - I guess that will be in the next marketing powerpoint...
Ericsson: 20 megapixel cellphones shooting Full HD video in 4 years {Engadget HD}
Nov 7th 2008 8:45AM 12-20 megapixels...
It is not just optics that needs to be better the size of the sensor would need to be increased, else you'd seriously drown in noise. Also you'd need some interesting storage system for all that data, even assuming you had hardware compression for the video.
Packing all that technology into a mobile sized device might prove to be the downfall of all of this :-)
Baraka: first ever 8K HD restoration Blu-ray Disc gets ship date, reviewed {Engadget HD}
Oct 27th 2008 7:27AM From the publicity still from the shoot it is possible to see the Panavision 65/70 cameras this was shot on:
Camera aperture: 2.066 by 0.906 in (52.48 by 23.01 mm)
(Source Wikipedia)
Compared to Full aperture 35mm 0.980" x 0.735" or Academy 0.866" x 0.630 camera apertures. Take the *width* as the height is ratio dependant and you find 65mm is > 2 times the width. So given a fixed film stock and thus grain size and resolution limit, you need to sample that many more pixels for the same resolution capture (excluding lens MTF differences for the 2 formats and scanners).
Normal process for 35mm is 4K scan oversampled to 2K, so for 65mm you'd want to do 8K to 4K for a similar quality scanning process. You would want to do this if you felt that you restoration should work on the 4K digital cinema standard or if you want to be prepared for future video standards.
Alternatively if your scanner is noisy, you can trade the resolution for increased signal to noise by down sampling.
Now if you were to then try sell your disk as "8K UltraDigital HD" one would have to include a big tutorial on this process to explain why Blu-Ray is not 8K. Otherwise they would be miss selling the product:
• Available for the first time ever on Blu-Ray in 8k UltraDigital HD
• Includes over 80 minutes of all new bonus features – "Baraka: A Closer Look", "Baraka: Restoration"
not that I expect most people to even play the extra content never mind actually understand the process or even for it to include a suitable description of the process anyway :-)
Film vs Digital, told through the story of Mr Pixel and Mrs Grain {Engadget HD}
Oct 23rd 2008 5:17AM Since when do either of video or film attempt to be real? They both have a 'look' and are never going to look real. Real is when your own eyes see the world and are constantly adapting to things like the intensity of your point of interest and its colour. When shooting with a camera the camera does not do this it is fixed for any particular image. These images are then reproduced usually at a much lower intensity and your eyes adapt differently to this situation. This is before you even consider the addition of noise sources like CCD/CMOS/Film, or the temporal sampling imposed by the capture choices, device gamut/metamerism, etc.
Its an artistic choice, for me they missed out on one obvious in joke... the colour of her dresses isn't quite right across the 3 shorts :-)
