CompuServe Thread

#camera lense w/human FOV

5 messages in this thread
#18637From: KEITH TROMERJul 7, 1992 8:33 AM
A couple of months ago there were lots of messages talking about what camera lenses comes the closest to duplicating a human eye perspective. Most answers say it was impossible because a human eye constantly moves back and forth. However, someone gave an answer for what lense is most like what a person would see. Do you remember what it was?? Thanks in advance -KT
#18654From: Yost GroupJul 7, 1992 10:05 AM
Keith, I think that 50-55mm is about normal focal length. Note that the field of view is not as wide as a human's, however. – G
#18703From: MARTIN G FOSTERJul 7, 1992 9:33 PM
I don't know if anybody thought of this, but how about a 50mm lens combined with a wide aspect ratio such as 1.85:1 or more. This is one of the movie aspect ratios, here are some of the typical ones: 1.33:1 Academy Aperture (also 16mm and NTSC) 1.66:1 European wide screen 1.85:1 US wide screen 2.2:1 70mm 2.35:1 Anamorphic Try playing with the aspect ratios and using the safe frame feature in a camera viewport to see what you'll see. Since a monitor is physically limited to 1.33:1 you would need to create a "letter-box" image with black borders. Martin
#18745From: Mike MorrisonJul 8, 1992 11:30 AM
Sounds like you've been hanging out with the guys at Trilobyte…<g> -MM
#18704From: Ken Loss-CutlerJul 7, 1992 9:51 PM
Keith: Gary is right: 50-55 mm. . You may be referring to a dialogue we had 3-4 months back concerning an eyeview problem in a simulated driving situation. Though that project was executed in Topas, our solution may have some value here. We took data from the State Highway Department which stated that signals should be placed within a 40-degree left-right cone of vision to be properly visible at a given point on the road. Topas permits zooming to occur in an "aerial view"-mode which provides a prepresentation of the "angle of vision from the camera". We simply jacked with the angle, using a protractor on the monitor, until we got it to jive with the data. The resulting view was quite close to a flat-field representation of what would have been visible to the driver, and the mm value was (surprise!) 53 mm. . Since the eye does not "see" in flat views and is constantly altering depth of field and point of focus, this is a primitive representation, at best. Ken Loss-Cutler