Hi all, I'm a complete beginner to having prints made from digital images.
So my question is this, how do I get the sharpest possible print,
eg, do I just send the full res converted jpg (3888 x 2592 pixels converted at 350 dpi), or should I take into account the printers native dpi and print size I want and do a conversion using those values?
So, for example, this image is 800 x 533 at 350 dpi. If I view the same image at full res (3888 x 2592 at 350 dpi) on my monitor, resized to fit, it looks fuzzy and lower quality. This same thing won't happen during printing right?
You have asked a simple question that some folks have written entire books about trying to answer.
Some quick rules of thumb:
- You want to keep sharpening haloes in the final print in the range of 1/100th to 1/50th of an inch. That means if you are printing at 300 ppi, a sharpening halo of approx. 3 pixels would be a good starting point. Unsharp Mask can be used for the output sharpening.
- It is very difficult to judge final print sharpening from a monitor. The most reliable way is to run a print sample.
My recommendation:
- Get a copy of Real World Image Sharpening by Bruce Fraser and Jeff Schewe. The second edition was just published recently.
In the meantime, Google the following:
- output sharpening
- three-pass sharpening
- Bruce Fraser & sharpening (there are book excerpts and articles of his online)
- Jeff Schewe & sharpening (same thing)
The simple answer is make some prints both ways and let your eyes tell you which is best. Whenever an image is resized up or down there will be some loss of original detail, but the human eye is easily fooled and unsharp masking can be used in varying degrees to compensate for the loss.
The more important perceptual variables are print size and viewing distance. Beyond reading distance the eye loses the ability to discern fine detail and our perception of sharpness begins to be based more and more on contrast. What USM does is enhance contrast along tonal boundaries, which perceptually creates the illusion of more sharpness. Don't lose sight of the fact the entire photographic process is based on creating the illusion of 3D with contrast patterns. On a pixel peeping level the sharpened image will not look as good, but perceptually it will look more natural.
The advantage of resizing based on printer native resolution (i.e. printhead density per inch) is that you control the resizing, not the printer driver, eliminating some uncontrollable variables. Also you are able to apply USM after resizing. For example when getting prints made at Costco I follow the recommendation in the tutorial at http://www.drycreekphoto.com which is the company which creates the profiles for the Costco printers and I apply USM as the last step before saving..
For very large prints the file sizes can get unmanageable if the image is resized in Photoshop to printer resolution and at some point its better to send the native camera file to the print (cropped perhaps but not resampled) and let the printer manage the resampling. USM will still be need in most cases, but with the printer handling the resampling you would need to apply it to the camera file.
There are just too many variables, not the least of which are your quality standards, to suggest one over the other definitively which is way I suggest taking a few of your favorite images and trying it both ways at various sizes with various amounts of USM.
The weakest link in the chain is human perception. it can spot minor difference in side-by-side comparison, but absent a baseline for comparison will adapt to whatever you are viewing. You'll learn more about the process by printing one photo 20 different ways and comparing side-by-side than basing assumptions on 20 different images.
I recommend buying PK Sharpener Plug-in. It has 3 different kinds of sharpening in the total sharpening process. Capture sharpening, creative sharpening and Output sharpening. The plug in is around $100, but is well worth the investment and will make a big difference in your final results.
