pkr@media01.UUCP (Peter Kriens) (07/13/90)
-------- [ OPI is a defintion from ALDUS [Pagemaker] that allows [ the manipulation of high res pictures in postscript while [ the end user uses a low res version. The spec can be [ obtained from Aldus: Developers Desk, Aldus Corporation, [ 411 First Avenue South, Seattle, Washington 98104 (206)-628-6593 -------- While writing software for using the Aldus OPI defintion for high def pictures in postscript, we have found something peculiar. Is there anybody who can verify those facts for us? In OPI you specify the original size of the picture in pixels. Now it seems to me that pixels are a nasty way of measuring sizes. If I make a low res version from my picture and give it to the end user, he will use quite a different resolution than what I store on the main system. I do not want to include the low res in a high res tiff file because then I would still need to send all the data to the user. So what now is the meaning of the pixels? Are they specified in the low res or in the high res. And if they are in the high res, how does the user know [or its application] what measuring system to use. I think the best solution would be if the image size and cropping rectangle could be specified in 1/72 of in inch, recalculation of these values to the pixels is quite easy if you know the dpi. And the dpi is a mandatory field in a scanned TIFF picture. Anybody any suggestions or comments on this problem? A A A A A
glenn@heaven.woodside.ca.us (Glenn Reid) (07/14/90)
In article <1289@media01.UUCP> pkr@media01.UUCP (Peter Kriens) writes: >-------- >[ OPI is a defintion from ALDUS [Pagemaker] that allows >[ the manipulation of high res pictures in postscript while >[ the end user uses a low res version. The spec can be >[ obtained from Aldus: Developers Desk, Aldus Corporation, >[ 411 First Avenue South, Seattle, Washington 98104 (206)-628-6593 >While writing software for using the Aldus >OPI defintion for high def pictures in postscript, >we have found something peculiar. Is there anybody >who can verify those facts for us? >In OPI you specify the original size of the picture >in pixels. Now it seems to me that pixels are a nasty >way of measuring sizes. If I make a low res version >from my picture and give it to the end user, he will >use quite a different resolution than what I store >on the main system. I do not want to include the low >res in a high res tiff file because then I would still >need to send all the data to the user. I had a look at the specification for "OPI" a while ago, when Aldus was just bringing it out. The problem you point out is very real, and to my mind, makes the format more or less useless. As I recall, you can't scale it, and it won't work at a target resolution that's different from the one you originally specified. In short, it's an expedient hack, not really a solution to the problem. The problem, however, is quite real. With high-resolution sampled images (especially in color), there are a whale of a lot of bits, and it's no fun even to store them on the disk, let alone ship them around over the network. OPI is intended for real Pre-Press systems, so you can scan a photograph on a Scitex machine and leave it on the Scitex. You can then design your page with a low- resolution version of the file on your personal computer, and when you're ready to really print, you can figure out what transformations and cropping to apply to the original file that should be still on the Scitex disk, unless somebody deleted it because they needed the space. Part of the issue is whether or not you really have a PostScript interpreter in the final pass (e.g. on the Scitex machine). The PostScript code that does not represent the image itself must get interpreted and rasterized eventually, and merged back together with the image. I remember saying to the engineer at the time (and having him agree with me), "...so if you have a PostScript interpreter at the end, this convention doesn't help you, and if you don't have an interpreter, it doesn't work [when you scale something]." My recollection of OPI is admittedly a little fuzzy and it may have been fixed since I last saw it, but the fact that you raise the same issue leads me to believe that it's essentially still device-dependent and really only a partial solution.... /Glenn -- Glenn Reid PostScript/NeXT consultant glenn@heaven.woodside.ca.us Independent Software Developer ..{adobe,next}!heaven!glenn 415-851-1785