Posted on

Output for print

Ready to print?

It should be simple, save the image and send it to a lab?  That’s fine for snapshots, but if you’ve spent time optimising and perfecting an image and need the print to reflect that, you’ll want the printer to receive the best data possible.

This article assumes you’ve checked the file for visible issues (refer to my other article, pixels to print ) and are ready for the conversion process which might involve re-sizing, sharpening, converting to jpeg or tiff, or changing colour space.

This article is especially intended for those of you who send me files for printing, but can be equally applied to images sent to a print lab (with some different limitations that will be noted).

ps1-image-size

 

Basic Workflow

Overall goal

Tto create an image file that can be printed without degrading quality from the original.

The basic steps are something like this:

Lightroom and Photoshop

The basic workflow is shown here and individual steps are shown below for Lightroom and Photoshop.  Lightroom somewhat simplifies this process within it’s export settings, so rather than complete each step individually as is required in Photoshop, Lightroom’s method is to ask what’s required and it takes care of the rest.

1. Check for sufficient resolution

Quick answer:

Minimum of approximately 3,000 pixels across the image (a guide for general printing).

Long answer….

It depends on viewing distance and other factors, 3,000 pixels is my interpretation based on assessing prints, judging what a ‘normal’ viewing distance is, and recognising the eye’s ability to separate detail down to about a 60th of a degree.  Lets say a 12 x 16 inch print is typically viewed no closer than an arms lengh, in this case around 180 ppi (pixels per inch) is sufficient resolution for ‘normal’ viewing.  That’s about 2,000 x 3,000 pixels or the equivalent of a 6 megapixel camera file at full size.  Doubling the resolution to 360 ppi (or 6,000 pixels across), will produce sharper results if the image holds detail, but a closer than ‘normal’ look may be required to notice the difference.

The larger the print, the greater the viewing distance.  When we view an image we would typically take in the whole image, we wouldn’t position ourselves so closely we’re inspecting details, so in the end I estimate that all images require around 3000 pixels across the wider dimension.   Higher resolutions will provide more detail when inspected closely, and in many cases that’s important (such as with large panorama’s where inspecting detail is part of the experience), but it isn’t necessary for every print.

If you’re far below the 3,000 pixel range, consider if the image is really suitable for a large print.

Today’s cameras are pretty good if you ask me, lets use a full size 24 megapixel image as an example.  A 24MP image would contain 4,000 x 6,000 pixels across each length, if this image is printed 40 inches (1m) across, it will contain 150 ppi which is enough for a great result at this size for most images, in fact, a small error in focus, stability or poor lens quality can easily degrade the image more so than the pixel count at this resolution.

More pixels will provide a more detailed print if captured well, but does that mean the emotional content is better conveyed?  Probably not, as much as I enjoy (and strive for) detailed prints, the emotional content of most images is founded just as much on personality as pixels.  In some images strong detail matters, we search these pictures for texture, detail and points of interest, but not all images fill this category where super detail is required to convey meaning.

This table shows print size vs resolution (click link)

 

201608_storm_creek_4726_2fs_copy

Consider this image, would print resolution be an important consideration?

_81D3925How about this one?

 2. Resize and Resample

Quick answer:

Resize to the ‘magic dpi’ number provided by your lab, as a random example, that may be 15 inches wide at 200 dpi.

It’s typically not 100% necessary to resize your final image because the print service can allow for this.  However print sharpness can be affected if  you don’t.

Don’t forget to crop/adjust your image to the intended print size if needed, some labs will chop off sections of the image automatically if the size ratio does not match, in my process I will never crop, the image is fitted within the print dimensions.

Longer answer….

The main reason to re-size is to ensure you can also sharpen the image at the final print resolution, so in this step the file is re-sized to match the printer (or lab’s) print resolution.

I print using Epson printers (eg Epson 4900 or 9070), for which the ‘magic dpi’ is 360dpi.  These printers have 360 nozzles spaced over 1 inch, this makes 360 ppi an ideal output resolution, or half of that at 180ppi is also acceptable for many images.

An example:  Lets say we have an image 3,000 pixels long, which is intended to print 15 inches wide, this will result in 200 pixels per inch.  The printer works best at 180 ppi or 360 ppi, so this image can be re-sampled down slightly to 180 or up to 360ppi at this point in the process, and both printed results will be very similar.

If the final image is lower than 180 ppi at the print size, re-sample up to 180 ppi, otherwise use 360 ppi to avoid losing data.

Labs may specify a different resolution, such as 200 ppi.

 3. Check for colour clipping

Photoshop and Lightroom both have methods of checking colour clipping.  Clipping can happen if your final image contains colours that are not printable, or if you convert from a larger colour space to a smaller one.  Typically photo’s are edited in a larger working space (eg pPro-Photo), and converted to a smaller space for output (eg sRGB)  If you find an image will clip, it’s often best to adjust the image manually until the problem colours fit within the target gamut.  The software (profiles) can take care of this for you, but perhaps not to your liking.

Choose output colour space

Quick answer:

It’s difficult to generalise here, sorry, the topic is very broad and this is a short tutorial, so please don’t throw things at me if you don’t like the answer.  If sending a file to the lab…refer to heading below.  If sending the file to me, my first recommendation is to save the file with the colour space you are already working in (no conversion).  If you are working from a raw converter like Lightroom, my general recommendation is to output to Pro Photo in 16 bits (tiff), simply to to guarantee there’s no colour clipping caused by the file export (ProPhoto is a large space, and 16 bits divides it into very fine tonal values).  When I print the image, it’s converted again into the printer’s colour space, which is different for each paper.  I check any suspicious (highly saturated) images for clipping issues before printing.

Longer answer….

There’s no single correct colour space.  I would argue that of the common spaces, the best choice is (often) the smallest space that doesn’t clip colour.

  • sRGB is the smallest space, this can be used with images of lower saturation (sunsets are typically out).  It’s the standard for the web, but limiting for prints.
  • adobeRGB98 is suitable for most images, but still doesn’t contain all the printable colours that a typical wide gamut printer can produce.  Even so, it’s a standard among pro print labs and most images will survive the transition to adobeRGB98.
  • Pro Photo is the largest space which reduces the risk of clipping printable colours, but it’s so large it comes with a risk of posterisation, the separation of colours that can be especially noticeable in smooth transitions.  Well that’s the theory anyway.

In practice, posterisation with Pro Photo doesn’t seem to be visible when saving files for print, at least not in my testing so far (so I stand to be corrected), but I do see banding in histograms when 8 bit files of any colour space are hit with large adjustments, or even just saved for printing, and Pro-Photo falls over a bit more quickly than sRGB or adobeRGB98 in this respect.  To eliminate the risk of posterisation, make sure you save any Pro Photo files in 16 bit mode (instead of 8 bits) which squeezes in more colour points.  If you’re saving as jpeg (8 bits only), consider adobeRGB98 or sRGB output (but check for clipping first).  jpg’s which are limited to 8bit’s will occasionally generate  strange colour shifts.

There are other colour spaces available such eciRGB_v2 (european pre-press standard) and the J-Holmes D-cam variants (paid), which help solve some of these issues.

Colour spaces for print labs

Short answer:

Use the colour space the lab recommends…..If they stare strangely when you mention ‘colour space’ they will use sRGB, if they use adobeRGB’98 it will probably be documented somewhere, or you can ask and they should know what you’re on about.

Long answer….

Most professional print labs operate in the adobeRGB’98 colour space, it’s become a standard among print labs. (refer to adobeRGB98 notes above).

‘Shopping center’ labs usually operate in the smaller sRGB space, because sRGB has become the default space to satisfy a world which is largely colour space ignorant.  These labs are much more focused on pumping out cheap prints by the thousand than perfecting colour quality, and results can be quite different at each lab.  They’re likely to recognise sRGB files only, if you try sending a file in the adobeRGB98 space (as I have) the colours can map incorrectly become and become even further displaced from where they should be.

 4. Output Sharpening

Output sharpening is compensation for ink hitting paper, it’s a small amount of sharpening applied at the final step before ink hits paper.  This is additional to any ‘creative’ sharpening you might apply.  It should be applied at the final print resolution.

5. File format

Save in JPEG or TIFF format

JPEG – Select max quality to avoid un-necessary compression artifacts.

TIFF – Use zip compression to save space (no data is lost) (better)

Lightroom and Photoshop

Steps in Lightroom

1. Check there’s sufficient resolution for a print.

In the develop module press i to toggle on the image dimensions.  If no information is displayed use View>Loupe Info in the menu to set it up.  Check you have at least 3000 pixels across and ideally 180-360 pixels for every inch of print size (may be less for very large prints).

2. Check for colour clipping

In Develop hit S for softproofing.

Choose the Profile: here sRGB is selected.

Select the paper icon in the top right to display areas where clipping will occur in the resulting conversion.

lr-soft-proof

Here the flame robin’s chest shows a clipping warning in red.  The resulting conversion to sRGB clips the brightest reds resulting in lower saturation. In some cases the resulting clipped appearance can become quite blotchy.  For the record adobeRGB’98 produced a small amount of clipping which was similar to the printer gamut, and ProPhoto showed no clipping.

lr-soft-proof-image

For this image it would be best to use adobeRGB or ProPhoto.

If  you must convert to sRGB (because your print lab only accepts sRGB files for instance), adjust the reds manually to bring them into gamut, before exporting.

3. Export

Right click the image > Export > Export…

In Lightroom copies are created by ‘exporting’.

The print dialog box appears.

lr1-export-box1

4. The obvious

….select a save location.

lr2-hard-drive

lr3-location

5. Consider re-naming the file with custom text

In this example ‘print’ is added to the file name so it can’t confused with the original file in 6 months time when I’ve moved onto other things.

lr4-rename

6. File Settings

Select preferred colour space (refer previous instructions)

Select JPEG or TIFF format.

JPEG settings

Make sure to set 100% quality when saving to JPEG to avoid unnecessary image quality loss.

lr-file-format-jpg

TIFF settings

For TIFF files it’s safer to use 16 bits, the file size will increase but there’s less risk of degradation.

Labs would typically specify 8 bits which is fine 99.9% of the time and reduces file size.

lr5-file-format2

I can accept images in any combination of the above.

 7. Resize

To resize, select 360 ppi and the print dimensions.

lr6-resize

8. Output Sharpening

Output sharpening is compensation for ink hitting paper, it’s a small amount of sharpening applied to bring that crispness back, this is an additional step to ‘creative’ sharpening.

lr7-sharpen

9.  Choose what to do after export

I like to to select Show in Explorer (PC) or Show in Finder (mac) to show the saved file.

lr9-show-in-explorer

10.  Save the settings as a preset

Keep these settings for next time by adding a preset.

lr10-save-preset

11. View the image

Check the output file looks OK by quickly checking it in your favourite viewer (this is why the ‘show in explorer/finder’ option is so useful), to make sure you haven’t made any glaring mistakes.

Steps in Photoshop

1. Duplicate the image

Duplicate the image at this point to avoid modifying the original.

ps2-duplicate

2. Flatten

To simplify the file

Layer > Flatten Image

3. Check resolution and Resize

To check the resolution, select the print’s width or height with Resample Image unchecked, the resulting resolution will be displayed, which is hopefully between 180-360 pixels/inch or higher for best results (very large prints can have less ppi).  An image with 3000 pixels across is usually sufficient for general printing (refer to the basic workflow section above for more information).

To re-size select 360 ppi, Resample Image and the print dimensions.

ps1-image-size

4. Check for colour clipping

This is only necessary if you are converting from a larger to a smaller colour space such as sRGB or adobeRGB’98.  If you’re sending files to a lab that requires sRGB or adobeRGB’98 this will be required.  If you are sending files to me it’s best to keep the file in your working space, as I’ll accept images in any space (so skip this step).

View > Proof setup > Custom > select colour space eg sRGB

View > Proof colours (ctrl + Y) to view resulting conversion

View > Gamut warning (shift + ctrl + Y) to view clipping

ps-proof-color

Using this example of a flame robin, the chest shows a clipping warning in grey.  The resulting conversion to sRGB clips the brightest reds, resulting in lower saturation, in some cases the resulting appearance can become quite blotchy.  For the record adobeRGB’98 produced a small amount of clipping which was similar to the printer gamut, and ProPhoto showed no clipping.

5. Change colour space

Again this is only necessary if you are outputting to a smaller colour space, otherwise skip this step.

Edit > Convert to profile.

Select colour space.

Since there’s no perceptual table for conversions to sRGB or adobeRGB’98 (without workarounds) you should first manually edit the image to bring any clipped areas into gamut the way you see fit before the image is converted, otherwise the conversion will clip out of gamut colours straight off, which isn’t always pretty.

5. Output Sharpening

Output sharpening is compensation for ink hitting paper, a small amount of sharpening is applied to the file as a last step before printing.

See quick output sharpening method here

6. Save as..

either JPEG or TIFF

File > Save As…

JPEG settings

ps3-save-jpg

Be sure to embed the ICC Profile.  Here I’m using a profile called DCam 3, you’ll likely see sRGB, adobeRGB or Pro Photo.  Without an embedded profile the colours can’t be read correctly and they’ll look weird.

ps3-save-jpg2

For JPEG’s, set the maximum quality to 12 (and don’t ask my why they used 12 steps, and not 10 or 100).

TIFF settings

Save without layers and embed the ICC profile.

ps2-save-tif

Use Zip compression to reduce the file size, not data is lost like JPEG compression.

Discard Layers to reduce the file size.

The other settings don’t seem to matter.

ps3-save-tif2

7. View the image

Re-open and check that file looks OK in your favourite software to make sure there aren’t any glaring mistakes.

Bonus

You can download the action below to automate the process of duplicating, re-sizing and sharpening the image.  When the image size dialog box appears enter your desired image size and print resolution.

Quick sharpening formula