JD750 wrote:
See Burkphoto’s post 3rd from bottom on p4
When I worked for a yearbook and portrait company, I ran the pre-press prep area of the elementary school memory book department. We made extensive tests of JPEGs vs TIFFs. There was absolutely no discernible quality difference when making CMYK separations from 8-bit JPEGs vs 8-bit TIFFs, both of which were in sRGB. There were only minor differences when working from 16-bit TIFFs in Adobe RGB, vs 8-bit files in sRGB. HOWEVER:
> It took roughly ten times the network bandwidth and server storage space to move 8-bit TIFFs around the image setting department instead of JPEGs. In 1998, that was a HUGE deal!
> It took roughly three times longer to use 16-bit Tiffs than it did to use 8-bit TIFFs. THAT was an even bigger deal.
Here is the skinny on WHY publishers have traditionally demanded TIFFs and "300 dpi scanned" files:
> In the early days of digital photography, few photographers calibrated their monitors to ICC Specifications. So when they adjusted images on uncalibrated monitors, they were most often too dark, too light, had inappropriate contrast, or had some off-color tint to them.
> Having 16-bit TIFF files in Adobe RGB provided enough latitude for the printers' technicians to adjust the customer files on a calibrated monitor and output 8-bit files that were color-correct and tonally adjusted with soft proofing to look good on the press, with the inks and paper that would be used for publication.
> Editors and publishers once had a workflow using VERY slow computers. So they would request scans of prints and transparencies in two resolutions: 300 dpi, 16-bit TIFFs, and 72dpi JPEGs. The 72 dpi (small, low resolution) images were used as "for position only" placeholders when creating page layouts on Macs. When the final document was ready to send to image setting, the 72dpi FPO images were swapped for the 300dpi (fully adjusted JPEG) images. This created high resolution halftones or color separations at 133 to 200 lines per inch, without slowing down the pre-press layout and proofing operations.
> There is little actual need for 300dpi scans (or 300 PPI images made to reproduction size). A 200 dpi scan or 200 PPI image made to reproduction size is sufficient for a 150 line halftone or separation. HOWEVER:
> Editors are notorious for changing their layouts at the last minute to accommodate changes. Having a 300 dpi scan or 300 PPI image made to the (initially intended) reproduction size allows 50% enlargement without quality loss. So requesting 300 PPI is simply a FUDGE FACTOR, just like telling photographers they have to submit 16-bit TIFFS in Adobe RGB color gamut.
Page editors are notorious control freaks. If they could get you to submit raw files, some of them would, so they could do your job for you, to their specifications...
In our operation, we provided 200 PPI files as 8-bit JPEGs in sRGB, with no significant quality loss compared with 300 PPI files submitted as 16-bit TIFFs in Adobe RGB. But we could get away with that because we were using a color-managed workflow, and we had trained technicians adjusting color and brightness on all files. We also made all files to ultimate reproduction size specifications.
This saved us tons of time and money.