Forum Moderators: not2easy
[webmasterworld.com...]
For a long time we have had 800x600 as a common standard, and now 1024x768 is becoming more common. I want to ask a slightly different question, about pixel density rather than monitor dimensions.
Macs have historically had a pixel density of 72 pixels per inch on the screen, and Wintel machines have commonly been 96ppi. In planning image designs for the web, I'm wondering what the next level will be. I know many laptops have a higher standard density (~120ppi?), but since I mainly use desktops I'm not sure what the common numbers are.
I recognize that on a given monitor, if you increase the screen dimensions (change the operating system's setting from 800x600 to 1024x768, for example) you will increase the pixel density. But there is a point beyond which the monitor-software can't successfully go. Given that we often specify image dimensions in px, are there standard densities we should be thinking about now beyond 72 and 96ppi that will become common on the next generation of monitors?
But does that mean that creating high-resolution images is worthwhile for internet use? I don't think so. I still create all my web graphics at 72ppi, because the appearance is acceptable and the loading times are better than a high-res image would have.
Let me pick your brain again (or anyone else's here) about file formats for multiple use. I'm *definitely* not an expert on this, but I've read enough now to realize that most people who talk about it don't really know what they're talking about either. ;-)
I'm now taking most of my images for the web with a digital camera, and the original files are 3000x2000px. I can save them in-camera as JPG, TIFF, or raw Nikon NEF (which I can convert after downloading). The majority of my images will wind up on the web, but I have occasionally published a few in print magazines back when I was taking 35mm slides. I'd like to preserve the capability to do that as much as possible.
I know people *say* that the best master format is uncompressed TIFF, and yes, I can save every image as a 16-bit 32M uncompressed TIFF. Realistically, even the pictures that might go into print would probably never be reproduced at more that 4x6 inches, although I did get a full page shot in once.
So this is the real question: given the enormous savings in space and media cost, if I save the master copy of each image as a 3000x2000 JPG, will that truly rule out the possibility of good print publication of that image at a reasonable size? (4x6 say)
I know that each instance of re-saving a JPG will degrade quality, so if I were to capture the original in-camera image as a maximum JPG and them immediately convert it to TIFF format, would that be a wise procedure to insure I don't accidentally degrade the image further with any subsequent edits? (Like adding a (c) notice, etc.) I realize that the TIFF file will already be missing any information that was eliminated in the original in-camera creation of the JPG, but once in TIFF format I should be able to prevent further degradation, yes?
I'm just trying to decide the cost-time-quality tradeoffs in making the master image, and then I can adjust the creation of derivative images in format, dimensions, pixel density, etc., as needed.
Many thanks for any insights you can give!
DPI for displays is of course meaningless, as DPI varies with CRT size and bezel sizes at same resolution, or with resolution at same CRT size. Even LCDs have real and effective pixels when they run at a different resolution then their native one. As my monitor happily supports between 640x480 and 1280x1024 you have more then a factor of two in both dimensions (extreme resolutions are usefull for testing sites). In fact since I run two monitors in parallel with different resolutions and CRT sizes, my "DPI" changes simply while dragging a window from left to right. So DPI is meaningless unfortunately, and you can only go for pixel sizes and target screens at this point in time.
Regarding your large image question. I can only say that subjectively I've always preferred the quality of aminimally compressed hi-res JPG to that of an uncompressed but resolution reduced image. And therefore I store all my hi-res masters in JPG format with minimal compression.
You can easily test. Simply minimally compress that image into JPG, then reduce its size and save as a TIFF until it will be the same byte size. Then compare quality loss.
I believe JPG works best when it has many pixels to play with and even stands up to extreme magnifications such as full page glossy prints.
But this is my subjective opinion, so you best test for yourself.
SN
Personally, I'd take the images uncompressed, and then decide per-image how you will store them. That way, you can compress the snapshot-quality images as much smaller JPG files, and keep the pro-quality images in a lossless format.
