I'm pretty sure Bob Mayson was "totally against" Digital Cinema two years ago... at least that's the impression I got every time he opened his mouth. But then again, company stratagies can change, nothing wrong with that. I'm simply pointing out the irony in that the two companies pushing Digital Cinema forward are the same two that were "considering" jumping out their windows every time the word "digital" was mentioned two years prior.

On a side note, John, your presence was missed this year at ShoWest... I could feel the "force" was not as strong this year. I hope your wife is doing better and that the force will be at full strength next year.

As I noted, Kodak (including Bob Mayson) will continue to ask hard questions regarding the technology and economics of Digital Cinema, as should everyone in the movie industry. That's the only way it will ever be better and more cost-effective than "Film Done Right".

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion

Please, don't think I'm dissing Kodak... we love you guys. Any chance you guys will be joining the rest of us in supporting the ETC? It's great to have your own research facility, but we got a good thing going at the ETC, and Kodak should be a part of it, don't you think?

The Kodak Imaging Technology Center (ITC) in Hollywood and the labs back here in Rochester are very busy with research and development efforts. As you know, early last year we renovated the 2000 seat Kodak "Theater on the Ridge" with a 25 x 63 foot screen and new projection room for this very purpose, well before the ETC even opened. Unlike the ETC, our large theatre already has 70mm film capability.

Kodak's participation and financial sponsorship of the ETC is a business decision that I am not privy to.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion

The sad thing is if you take the average person who does not understand such things, they will invariably say that a TV with the sharpness control turned up "looks better" or that an edge-enhanced image "looks better". I'm currently taking a digital image processing class and we have been studying image enhancement in both the spatial and frequency domains. A very common approach to edge enhancement for images is to use a convolution mask (various convolution masks can be calculated to approximate any given frequency based filter H(x,y)). The larger the convolution mask h(x,y), the better the approximation, but simple 3x3 or 5x5 masks can often be used to smooth or sharpen effectively. Another edge enhancement technique is to high-pass filter an image and add it back to the original image. Edge enhancement, while sometimes improving images for human perception, actually destroys the integrity of an image. Look at an edge-enhanced image or test pattern on a TV with the sharpness run up and between abrupt transitions of color, such as from gray to black, there will be a brighter gray area between the two solid colors that goes away when the sharpness control is adjusted properly (all the way down on some but not all sets). Be sure to turn off scanning velocity modulation, which produces the same effect!

I fear that digital cinema is going to have problems because edge enhancement is a cheap way to make an image look sharper and crisper to the average viewer, while distorting and corrupting true high-frequency detail in an image. The other problem is aliasing, which I have addressed in a previous thread. The nature of film grain prevents aliasing (except for temporal alaising due to the 24 fps frame rate), but a rigid array of sampling points will create alaising in the image, and if the projection resolution is different than the camera resolution, then additional artifacts will be added due to the resolution conversion.

I'd rather see extra film grain than spatial aliasing! For a demonsration of sampling a signal with a wide range of frequencies, refer to my previous thread "Film vs. Digital (trading noise for aliasing?". Such aliasing happens most frequently when near-vertical or near-horizontal straight objects appear in an image, such as blinds on windows in a background. Fine patterns on clothing or objects (car grilles, air vents, textures, etc) can cause aliasing as well, and can create false patterns in the image.

The way to minimize the impact of these problems with digital sampling using rectangular arrays (or any other regular array) is to use a high enough resolution so that most problems are not visible at even a relatively close viewing distance, and to do smart filtering (ideal filters which cut off everything above a given frequency produce an unpleasant ringing effect (which may be a counter-intuitive notion for many of you)). To prove it to yourself, take an image than has a single white dot an a black background. Let f be the desired cutoff frequency. Let D(x,y) be the distance from a point from the origin. Take the image's Fourier transform, than set every point (x,y) in the Fourier transform to 0 if D(x,y) is greater than f. Now take the inverse Fourier transform. You should have the original image, ideal filtered with cutoff f. You'll see a ringing effect around the original dot if you picked the right value for f. Play with that, varying the value for f.

Bottom line: use a high enough resolution to supply real detail and filter properly to minimize ringing and aliasing! If this is done properly by the digital projection equipment makers AND the digital movie camera makers, then digital cinema might be wonderful. However, I doubt that this will occur any time soon.

Evans A Criswell
http://home.hiwaay.net/~criswell/theatre/