Message-Id: <199706051613.JAA339002@dns.ccit.arizona.edu> Date: Thu, 5 Jun 1997 09:13:15 +0000 From: Bob Weaver <mailto:Bob.Weaver@NAU.EDU> Subject: Re: 4x6? 5x7? 8x10? To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU
Mr. Erickson (and IMAGELIBers): Easy answer to your question--experiment! It takes just a little time to create scans of small and large prints at various resolutions, then save them to JPEG format at various compressions, then compare the results. Only you know how "good" the results need to be, and you can tailor your procedures to get those results with minimum file size per image. Note that JPEG is "lossy," and the more compression you do, the worse your product will be--this means there's a tradeoff between scanning resolution and the JPEG compression ratio used for a given resulting "goodness." A "halving" of scan resolution (1/2 the pixels-per-inch in height and width) means 1/4 as many pixels and a quarter the file size, while quadrupling the compression of a high-res JPEG image instead may or may not result in similar image quality reduction. Again, experiment....and after experimenting, publish. IMAGELIB wants to know your results! Good luck, Bob Weaver Northern Arizona University
At 10:42 AM 6/5/97 +0600, you wrote: >A question for you experienced photographers who do scanning:
> I have just done some architectural shots, using Fujicolor Reala
>35mm. print film (ISO 100). I intend to have these processed by a local
>professional lab, and then to scan the prints for inclusion in our art
>department's web page. For a film such as this, shot with fairly good Canon
>equipment, and *all other things being equal*, will I get better scan
>results working with a smaller, *apparently* sharper print, or a larger
>print? I suppose part of the answer will turn on what one can do to a
>scanned image after scanning and compressing, using in this case Mac Image
>Assistant.
> The scanner I will be using is a Hewlett-Packard ScanJet 4-C. I
>will be scanning the images first in TIFF format, then JPEG compressing at a
>ratio of 1:25, at "80% quality" (Mac Image Assistant variable).
> Any thoughts on this?
> Thanks.
> J. Erickson
>
>
>Jeffrey S. Erickson phone: 704-892-2590
>Slide Collection, Visual Arts Ctr. fax: 704-892-2691
>Davidson College e-mail: mailto:jeericks@davidson.edu
>315 N. Main St.
>P. O. Box 1720From mailto:owner-imagelib@LISTSERV.ARIZONA.EDU Thu Jun 5 13:25:36 1997
Received: from dns.ccit.arizona.edu (dns.CCIT.Arizona.EDU [128.196.139.46]) by library.wustl.edu (8.8.5/8.8.5) with ESMTP id NAA13204; Thu, 5 Jun 1997 13:25:35 -0500 (CDT) Received: from dist (dns.CCIT.Arizona.EDU [128.196.139.46]) by dns.ccit.arizona.edu (AIX4.2/UCB 8.7/8.7) with ESMTP id L A236738; Thu, 5 Jun 1997 11:23:40 -0700 (MST) Message-Id: <mailto:199706051823.LAA236738@dns.ccit.arizona.edu> Date: Thu, 5 Jun 1997 13:00:38 -0500 Reply-To: IMAGELIB <mailto:IMAGELIB@LISTSERV.ARIZONA.EDU> Sender: IMAGELIB <mailto:IMAGELIB@LISTSERV.ARIZONA.EDU> From: Anna Martin <mailto:martin@MAIL.CEDARNET.ORG> Subject: image size/quality information To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDUI put some information comparing various compressions and formats (JPEG and GIF) on-line for students and others at:
http://www.cedarnet.org/provider/style/images.html This is a bit dated now, but cites (some one else's) experimentation with compression and format in relation to image quality, for various types of images. In general, the amount of gradation, and the types of color information involved, will interact with comporession to affect image quality.
The size of the original photograph may have a negligable effect, since the film speed > grain of the original will enter in. You are going to be limited by the resolution of the output device, of course. Web browsers will display a limited range of colors, at lower resolutions, so you will only increase load time with no gain in quality if you save too much information in your final image file.
Anna Martin Chair Electronic Media Interest Group, National Art Education Association
http://www.cedarnet.org/emig Anna Martin http://www.cedarnet.org/mar