One of the things I like about using Ghost as a content creation platform is that the writing and story composition tools are well thought out and easy to use, without a gazillion choices to make to do the simplest of things.
Even so, every story deserves good images as well as catchy titles.
Choosing Images
When choosing the image that goes across the top of the post page and is displayed on the home page and in the story stream, I follow advice given by Seth Godin over a decade ago about another subject: Powerpoint. In an epic rant about how Powerpoint is abused, Godin suggests that the way to create presentations with impact is to select images with emotional weight that help you tell the story, rather than reading the text on a slide (after slide after slide after slide ... ).
So that is what I do. My preferred image source is the site Unsplash (see Resources), which has a collection of tens of thousands of royalty-free high-resolution images. I think about a concept that relates to the story I want to tell and words associated with the concept. I search Unsplash using those keywords until I find something that resonates with me.
Often, as you can see in the Original Image below, while the photo is dramatic on its own, it lacks a certain je ne sais quois de vivre. The aspect ratio might not be dynamic enough, or it might appear hazy because the contrast is flat – any one of a number of things that call out to be fixed.
However, tools like Photoshop or Lightroom can be expensive, and performing edits that require masking and multiple techniques can take a long time to master to use them effectively and efficiently, and then performing them on images can also take a lot of time.
Enter Luminar AI.
I have been a user of Luminar for well over a year (versions 3 and 4) and I found them to be effective and efficient compared to doing the same editing tasks in Photoshop (which I have owned and used since before I can remember – 1991, at least – to process images). Luminar AI raised the possibility of making my workflow even more efficient by reducing the number of individual steps and complex masking needed to achieve a desirable result.
Since downloading Luminar AI in mid-December last year, I have had to adjust my workflow, but once I settled into a groove my productivity has improved immensely, saving me five to fifteen minutes or more per image to get to a result I am happy with.
One happy outcome is that I now look forward to the process of selecting and editing images whereas before it felt more of a chore. Anything that gets in the way of sitting down to write is an obstacle to be overcome.
The workflow below shows the process I went through to produce an image used in the body of the Missing page. The Missing page is a way of failing gracefully: It’s like a 404 except I know that an old URL (from Maven or the Jamroom Archive site) is no longer valid for some reason, and those URLs can be redirected to this page, which explains what happened and what steps to take next.
Original Image
Edited Version Used on TheChocolateLife
Because the Missing page is about the process of moving stories from Maven and Jamroom to Ghost, the keywords I used when searching Unsplash were migrate and migration. It usually takes me less than two minutes to find a suitable image but sometimes it takes longer – the time it takes to find the right image is always well worth it.
Workflow
After finding and then downloading the original image from Unsplash I open it in LuminarAI.
- Composition. Step 1 is always to use the Composition AI tool to suggest a rough crop. Images used as the header (or listing) image are cropped to 16:9 as this is a good compromise for the three different crops used in this theme. However, because this image was used in the body of the story, I wanted to emphasize the horizontal-ness and so chose a 21:9 aspect. I was not entirely happy with the recommendation, so I moved the crop so that the three zebras in the center of the image were just above the rule-of-thirds guidelines the Composition AI tool displays.
- Enhance. Step 2 is always to use the Enhance AI tool. There is a general Enhance slider as well as one for the Sky. I am not looking for photorealism here, I want the image to pop. One of the things I like about the Enhance AI tool is that it alters contrast, sharpness, and other parameters at the same time, in a balanced way. I rarely take this all the way to 100 but I almost always push it past 50 (waaaay past 11). If the tool does not detect a sky in the image that slider is not activated, but in this case there was a sky, so I used it to pop the blue.
What I do next depends on the image, but in many cases I consider my work done after just these two steps. - The Sky. In this image there’s not much going on in the sky, so I went to the Sky AI tool to add in the clouds. There are controls to tweak the sky/horizon interface and I used those here as well as pushing the intensity.
At this point, I felt that much more would be gilding the lily and any more “enhancement” would lead to diminishing returns. - Export. Luminar AI has the ability to open any image directly in Photoshop (or Lightroom) without having to save it first, a real time/step saver. So I “open” the image directly in Photoshop and then resize it to the desired width (1920 pixels) before saving as JPG, looking for a quality setting that gets the image to around 500k in size.
And I’m done. From the time I opened the image in Luminar AI to the time I clicked the Save button in Photoshop, the total elapsed time working to go from the original to the image used in the story was well under five minutes.