Welcome to the Victor's Pictures blog. Leave comments if you like.

Saal Book

December 15, 2020  •  Leave a Comment

In October, I was invited to create a book printed by Saal Digital for review. I didn't have anything prepared for a book, so my first efforts were spent in deciding what to include in a book, selecting photos, sorting them, and arranging them in a way that made sense for a book. I spent some time doing some cleanup, but not as much as I would have for a productional product because of the limited time frame.

After selecting the photos, the next task was to arrange them in preparation for printing. Saal Digital has software to download to assist with the layout and organization of the photos and text. The software is quite capable, but it is not as intuitive as it could be.

Note: project files are saved in
~/Library/Application Support/SaalDesignSoftware/Local Store/projects

folder on a Mac. My system is currently dual boot (Catalina and Mojave). When I started the book project I was using Lightroom 6, which is incompatible (sort of) with Catalina. So I initially installed Saal Design Software on Mojave. In November, I signed up for Lightroom CC, so I'm primarily using Catalina now. To migrate project files over from one OS to the other, it seems simply copying the project folder works. The software also has project import and export functions which I could have used, but that would have required switching to Mojave and back to Catalina. By doing the copy, I was able to stay in Catalina. Note that once the book is ordered, there is an additional folder, orderedProjects beside the projects folder.

When I say the software is not intuitive, what I'm referring to is that things are not where I expect them to be. Most software, for example, has a menu bar at the top, with things like File, Edit, etc. Not so with Saal Design Software. Instead, you have to learn where all the panels are, and where the various tools are in each panel.

One of the panels is an Image sources panel. The way I worked this project was to add a batch of photos and lay them out, then add another batch of photos. Unfortunately, the Image sources panel doesn't automatically refresh when files are added. There is a refresh icon that you can click to refresh, but it refreshes the entire contents of the panel, which can take some time. Similarly, if an image turns out to need editing, saving a new copy to the Image sources area is not sufficient to refresh the project. You have to drag the image from the Image sources area over to where it's used in the project. Apparently, that's because the software keeps its own copy of all the sources in the above mentioned ~/Library/ folder.

The software provides a number of predefined layouts. You can use those, or you can create your own. Layouts you create can be added to the list of available layouts. The software has tools to align and size images. I found that there is a potential issue with images that abut each other. The software rounds sizes and locations to the nearest 1/100", apparently. This seems to sometimes cause thin white lines between images that were intended to abut. I didn't submit my project this way. Instead, for layouts with abutting images, I increased the size of the images by 0.01" without changing the position so that they barely overlap. This seemed to result in an acceptable solution. I reported this issue to the company, and they were responsive, so there may be a software fix in the future. Here is an example of what I mean. The weevil pictures are laid out to abut each other. Adjusting them to barely overlap avoided the issue, and there is no visible overlap.


The page layout shows a thin red line in the layout software, where the edge of the paper is. This is useful if you have full bleed designs, so you get an idea where the image will be cut off. I found the position of the indicated line to be very accurate, when I compare the printed product to the image displayed in the software. A good feature of the software is that if you lay out components too close to that red line, i.e., too close to the edge of the paper, and the components are not full bleed, the software gives a warning and gives an opportunity to correct these situations before the project is committed.

Very useful is the PDF preview feature. There is a setting to enable (default) or disable addition of a watermark on the PDF file. For internal use, there's no reason to have a watermark, so I disabled it for a more accurate image of what the result of the print would be. The resulting PDF file seems very close to the final product. Colors are accurate, and the layout is accurately reflected in the PDF file. However, one thing I didn't do but should have: I did not print a sample page on paper. I very much recommend doing this, printing to the size of a book page, especially if you use text. I wound up using a larger font size for most of the book than I should have because I was viewing the preview PDF at a smaller size than the final print size. Make sure to view your preview or a print sample at actual size in order to determine things like font sizes. The software has a preview function, but the scale is not accurate: 100% in the preview screen does not display the page at the size it will be printed. On my computer, 183% is close to the size of the final product.

My project was for an 8x12 photo book using matte photo paper (Fuji Crystal Archive paper) with a leatherette cover. There is an option to include a title on the cover, but I didn't see it before placing my order. There is an option to use a photo cover, but after reading another review, they said that the cover picture scratched easily when shelving the book, so I opted not to have a photo cover. I plan to add a cover title by using a foil tool to lay down a metallic layer displaying the title. I'll probably do it with gold foil, which will contrast nicely with the black leatherette.

The printed book has arrived. It's a lay-flat design, but I found it doesn't really lie flat. Here, the book is opened to the first page. Since the inside cover is glued to the cover, it lies flat, but the opposite page lifts up a bit.

If you open the book to a page, one page or the other sticks up a bit. It's possible that with use, this will resolve itself, and I haven't made any effort to exercise the book. The binding is nicely done. The leatherette has a nice feel to it. Since there is no design on the outside of the book, you can't tell which side is up. That's a problem I'll fix with the foiling. On opening the book, the first thing I notice is how thick the page glued to the cover is. I contacted support to find out more information about that aspect of the binding. The inside cover is used for images. I was told that the same stock used in the rest of the book is used for the inside cover. It's no thicker than any of the other pages. However, that does not seem to be the case. Here is a closeup of the spine, where the thickness of the pages is clearly visible.

The left side of this image shows the inside cover page. The top layer is the printed paper. Below that, there are apparently two layers of a different material. This three layer sandwich is glued to the cover. If I get an explanation from support for this design, I will update the review. In the mean time, this sandwich results in the inside cover feeling like it's a couple of pages that need to be opened.

Maybe the extra thickness helps to hide the lumpiness that would result from the edge of the leatherette folded over the edge of the cover, under the inside cover page. In any case, this may take some getting used to but shouldn't be an issue.

Here you see the inside cover angled so that the lumpiness I referred to catches the light. It's not really noticeable, but it would be more noticeable if those extra two layers were not there. Another option for binding would be to have the extra two layers stop short of the edge of the leatherette. I think that would be a better solution, and the best thickness of material would match the thickness of the leatherette.

I didn't have any two page full bleed layouts. This was as close as I got. Full bleed on two pages with a couple of extra shots with drop shadows. Speaking of which, I like the rendering of the drop shadows. There are some options for that, but I think I used the default settings.


Single page full bleed images can run up to the center of the book. The transition of one image to the other matches the crease in the page very well. I did not find any errors in registration.


There are a number of tools to handle text. I didn't experiment much with different fonts, but there is a whole suite of fonts available. I did experiment with text layout. Text is arranged in text boxes. Images are arranged in image boxes. These boxes can be opaque or transparent. Here, I've made the text boxes transparent so that a full bleed image shows as the background.

The text or image boxes have various properties. For example, you can set borders with square or round corners. You can add drop shadows, as I previously mentioned. You can overlap the boxes and adjust which one is in front. In the above image, boxes overlap, but the text boxes on top are transparent. I did not use any layouts with overlapping images, but that is possible.


Here I tried a couple of different layouts to fill the pages with a grid of images. On the left, I used black borders with image boxes abutted together. On the right, no borders with image boxes equally spaced.

One final note about the binding. Ideally, the folded edges of the pages should all align. There is a bit of misalignment that may affect the way the pages lie flat or not. There's also a spot of glue at the left, which can probably easily be cleaned off. This misalignment is not really noticeable when using the book.

Throughout the process of creating the book, I had occasional questions. I emailed support with my questions, and I generally got an informative reply within a few minutes of my query. The support was excellent.

Oak Work Update

March 05, 2016  •  Leave a Comment

About five years ago I decided to do some work on a Chinkapin oak in my front yard. My yard has two Chinkapin oaks, both apparently planted at the same time, before I bought the house in 2002. The one on the north side of the driveway was growing significantly faster than the one on the south side, which seemed to be languishing. Upon investigation, I realized it had some significant issues, relating to errors that occurred at the time of planting. Roots were extending east and west, but not north and south, with the result that the tree rocked back and forth in the wind. Excavating the roots revealed some other issues.

This post revisits the work I did and includes updated pictures showing progress.

Quercus muehlenbergii from northQuercus muehlenbergii from northThis tree came with the house I bought in 2002. Unfortunately, there were some issues with this tree, and I set about diagnosing the problem. Last year I concluded it was planted too low and had potentially strangling roots. Here you see the result of my efforts to jack it up last year. The tree has responded by thickening, especially where it is in contact with the brick. The boards will eventually rot. I will periodically monitor activity around the brick, and remove it when the tree has grown enough to not need its support. I may also leave it. I can't think of a reason it would be harmful to the tree there.

This picture shows the tree about a year after I jacked it up. I had placed wood and bricks under the jacked up tree.

Nearly 5 years later, the brick is still there but the wood has rotted away.

View from SouthView from SouthThis southerly view shows clearly the section I excised from the wrap-around root. There are new roots growing toward the east and toward the south. I will nurture that small root heading south. It may eventually turn into a major root for this tree.

You can't really see this very well, but the cut root closest to the camera has built up a callus where it is in contact with the brick. This helps to firm up the tree's anchorage. Unfortunately, there are no new roots growing from this cut root.

To the left is a picture showing where a section of root was excised. The two roots going to the right are on the east side of the tree. The nearer one is actually the root that was severed. It had grown over a lower root and grafted to it. My plan in 2011 was to take out that section of root and graft a seedling to the stub on the left, the south side. The problem the tree had was that there were no roots growing to the north or south. Grafting the seedling was intended to remedy this problem. In retrospect, I should have used an older plant with a longer root, but I didn't have one at the time.

The brick you see here is opposite the one seen in the picture above. I jacked up the tree by using the trunk as a lever, rocking it back and forth, each time scooting the bricks further under their side of the tree. In all, I think I elevated the tree 5-6 inches this way.

Southeasterly view of graftSoutheasterly view of graftAnother view of the graft -- also a good view of the callus on both sections of severed root.

The seedling is approach grafted. For the first attempt, shown here, I simply wrapped the join with tape. That did not actually work, I suspect mainly because of the tree rocking back and forth from the wind.

Here you see the graft a couple years later. I redid the graft by first making fresh cuts, then nailing the seedling in place and wrapping in parafilm. Another issue with the tree involves the roots on the far side. They are crossing each other, which is not a good thing. It's not as bad as it could be because they're fairly far apart. That top root arches completely out of the ground now that the tree is jacked up. It was underground before I jacked up the tree. I'm thinking that in a year or to I may switch those roots' positions, severing them both and reattaching them so that they don't cross. I'm not doing that now because the tree is actually finally solid, and it looks like that arching root is primarily responsible for the tree's stability.

South view of approach graftSouth view of approach graftUsing a gouge blade on my Exacto knife, I removed a section of wood from the cut root. I also removed a thin layer of bark from the adjoining portion of the sapling. I tied it securely with green tape and then wrapped that with parafilm (not pictured).

Front view of the graft. The small right to the right of the cut came from another plant and was removed.

Several years later, the seedling has grown up and branched out. I expect that as it continues to grow, it will thicken substantially more below the graft than above it. We already see a hint of this, but on the other hand, seedlings also show the same feature as you can see in the left picture.

View from EastView from EastFrom this view you can see how the roots are really parallel to each other, offering very little north/south support. You can see bricks under the tree both on the north side and the south side, and a board on the east side.

Side view of the graft site before the graft was made. This view is interesting because it shows more clearly than other views how much the tree has grown. Note the thickness of the roots compared to the size of the space between them.

To stabilize the tree I pounded a steel stake into the ground and held the tree with a wire. The tree grew over the wire, and when the wire broke, I added a rope. Today I removed the rope and cut off the wire. The tree is very solid now.

Tile Rotation Part 2

March 08, 2014  •  Leave a Comment

In my previous post, I describe a tile rotation app and a couple of ideas for implementation of a stone moving game. This post describes another possibility. 

The players take turns adding stones to the board. Only one stone is allowed on a tile at a time. After all stones are placed, each player is allowed to rotate tiles in order to create the maximum scoring possible. Points are awarded based on number of stones that are mutually touching. Following are some screen shots that illustrate.

So. how should points be counted? One possibility is to simply count the number of connected stones. So in the above illustration, black gets 6 points, and white gets 7 points (but maybe only two touching shouldn't count, in which case it would be 4 and 5, respectively.

Another possibility is to count the number of touches each stone has. For black, each stone touches 2 others in the square and 1 other on the left, so the total point count would be 8 for the square and 2 for the others, for a total of 10. For white, 3 stones touch 2 others and 4 stones touch one other making the total 3*2 + 4 = 10 points.

Or maybe the longer the string, the higher the reward should be. Powers of 2 would be the obvious choice, but powers of 10 would be easier for most users, so I'll describe that. Essentially, one 0 is added for each additional stone in the chain. 

So for white, there is a 5 stone chain (100,000 points) and a 2 stone chain (100 points) for a total of 100,100 points.

And for black, there is a 4 stone group (10,000 points) and a 2 stone chain (100 points) for a total of 10,100 points. 

Now let's look at a screen that has been completely filled with stones (note only one stone is allowed on each tile).

In the above illustration, the tiles have been rotated to form a checkerboard pattern. No attempt has been made to optimize score. Using powers of ten scoring, the score is:

Black: 10,000 * 2 + 1,000 * 5 + 100 * 6 = 25,600 points.

White: 10,000 * 3 + 1,000 * 2 + 100 * 10 = 33,000 points (note the 3rd white triangle at the top joins the third white triangle at the bottom forming a 3 stone grouping).


Here the tiles have been rotated in order to cause more stones to touch each other. I won't add up the total score, but I will point out the highest scoring grouping of each color, which likely determines the winner.

Black: 9 stones: 1000,000,000 points (the string of black at the top continues with the string of stones at the bottom.

White: 8 stones: 100,000,000 points.

Here black has a 7 stone string. White has a 6 stone string.


Here white has a 10 stone string.


Here black has an 11 stone string.



Tile Rotation

March 04, 2014  •  Leave a Comment

This blog post is a departure from previous posts because it doesn't include any photographs. Instead, it includes screen shots of an app I'm working on. The app, for now, is called Tile Rotation. The basic operation of the app is described on the app's website.

I've decided to start brainstorming for a board game based upon the second phase of that app using stone movements in addition to tile rotations. Here are some sample screens to show possibilities.

Here stones have been added to all the tiles that are on either end of the board. Black stones are on one end. White stones on the other end. Players can move stones and/or rotate tiles. To move a stone, it is dragged to the new location. White stones always sit on the white portions of the tiles they sit on. Black stones sit on the black portion. To rotate tiles, the user taps once to rotate it clockwise or double-taps to rotate it counterclockwise. In either case, either a whole column or tiles or a whole row of tiles is rotated in order to keep neighboring tiles touching with the same color.

Here is an example of what the screen looks like after each player has moved one stone and tapped one tile to do a rotation. At this point, I need to decide what moves should be legal. There are many possibilities. Perhaps stones should move only by one tile at a time. Perhaps they should be able to move multiple tiles, as long as they go in a straight line. That's what's shown, above. Perhaps captures should be allowed.

Here, each player has made several moves, and white has just placed a stone on the same tile that's already occupied by a black stone. Does this constitute a capture?

Anyway, the rules are not set yet. If you have an iOS device and would like to join the test team to help decide on what rules should apply, let me know, and I'll add you and give you additional instructions.

Here is a sample tic-tac-toe-like game. The board starts out with no stones. Players take turns adding a stone and optionally doing one rotation. First person to get 5 stones in a row along the same color wins.

Bluebonnet Update

December 02, 2012  •  3 Comments

Maroon Bluebonnet December Bloom I have an update to my previous blog entry about the bluebonnet with the strange inflorescence, where it looked like the flowers were turning out to be leaves. Now it is producing actual flowers.

The plant is in a bed where I planted maroon bluebonnet seeds in the fall of 2011. This particular plant sprouted almost exactly a year ago. Perhaps I'll search for older photos of this plant.

It bloomed for the first time in the spring of 2012. It survived the summer, and now it's blooming again in December.


January February March April May June July August September October November December
January February March (2) April May June July August September October November December
January February March April May June July August September October November December
January February March (1) April May June July August September October November December
January February March April May June July August September October November December
January February March April May June July August September October November December
January February March April May June July August September October November December
January February March April May June July August September October November December (1)
January February March April May June July August September October November December