Improving the chart plotter

A short digression here from my usual blather about where we are, what we're doing, and what else is broken on the boat, to talk about, umm, something broken on the boat. In this case, it is the chart display, and, technically, it's not "broken," but neither is it working properly. Those not interested in the arcane details of marine chart plotters can skip this post entirely; I wanted to write this up for anyone else who finds themselves in a similar predicament.

First some background: We have no fewer than six chart plotters on the boat, and other than the two Furuno units which are literally networked together and sharing the same chart source, no two plotters ever show the same things. For this reason we are always astounded by people in go-fast boats zooming into unknown territory with supreme confidence in their single plotter. We, on the other hand, are forever living Segal's Law: a man with two watches never knows the time.

The "main" chart plotter on the boat is a general-purpose computer running dedicated plotter software called "Polar View." There are lots of plotter software packages on the market; we like this one because it integrates the Active Captain database of hazards, anchorages, marinas, and other services, runs on both the Windows and Linux platforms, and comes with licenses for up to five computers on board. It accepts cartography in a number of standardized formats, notably the ones in which all the NOAA and USACE charts are published.

What it does not accept is anything in a vendor-proprietary format, such as C-Map from Jeppesen, BlueChart from Garmin, or a host of other proprietary formats. Mostly these formats are understood by dedicated plotter hardware, loaded on encrypted memory cards, although some proprietary charts are readable by some higher-priced plotter software packages. The proprietary chart universe is a mixed bag -- no single proprietary cartographer has claim to the universally best charts for every region of the world.

When we were getting ready for this Bahamas trip, I had to scramble to find up-to-date cartography for the region. The best I could do in a format readable by Polar View was to buy the encrypted Navionics charts, for around $200. In most of the Bahamas they've been barely acceptable, but at the very least, they show the land masses and other fixed objects in the correct positions. The bathymetry is another matter, being hopelessly out of date or just plain wrong even along popular cruising routes.

We've generally worked around this by copying waypoints from our reasonably accurate Explorer Charts, which are spiral-bound in book form, and entering them on the plotter. We then use those electronic waypoints to plot our courses from place to place. I've also copied numerous hazards such as coral heads and shoals into the plotter as waypoints, along with recommended anchorages. It's tedious, but it works.

For reasons that are unclear, the western boundary of the Navionics Bahamas chart "cells" (tiles in the chart mosaic) starts at 79°0'W. Bimini, Cat, and Gun Cays, where we are headed next, are all west of that line. This problem first bit us back in February, when we initially crossed onto the bank north of Memory Rock, but with no land masses or underwater hazards where we anchored that first night, it was not much of an issue.

Where we are headed, it will be a much greater concern. We will have to navigate intricate channels through shoals, miss a number of underwater hazards, and enter narrow channels and fairways in significant current. The NOAA charts already loaded on the plotter are based on ancient charts from before the era of GPS precision, and even the land masses are off by hundreds of yards. Having the plotter showing us driving over dry land when we are in the channel is disconcerting at best, and dangerous at worst if we have to make quick mid-stream changes of plans.

We do have an up-to-date backup, which is the latest proprietary C-Map chart, based on Explorer data, loaded on an app on my Android phone. The display is tiny and can't be used outdoors, but at least the bathymetry is mostly correct. We've had the phone out and displaying these charts each time we've navigated our way into narrow or shallow stretches throughout the trip. At just $13, these charts were a much better deal, and more accurate to boot.

That does nothing, of course, to fix the main display. I really wanted to resolve that before we cross west of 79°, and today was more or less my last chance to do so. A search for any more accurate compatible cartography has come up empty, and even a plea on some online forums for a solution yielded nothing. With nothing to lose, I decided to try to roll-my-own.

One of the formats understood by Polar View, and almost every other computer-based plotter program, is "BSB," a format for "raster" cartography, which is a fancy name for a simple image of the region in question, with coordinate information attached. A free program called imgkap can import such an image in any of several common image formats, format it for use as a BSB chart, and position it on the coordinate system based on user-entered latitude and longitude of the NW and SE corners of the image.

Step by step, here is what I did:

1. I used my cell phone camera to snap a photo of a paper chart. A flat scanner would be a better choice, but I did not want to have to cut my chart out of its bound volume. I did my best to have the edges of the chart be square with the frame of the photo, but it's not critical, as this will be fixed in the next step. It's best to have the lens above one short end of the chart, rather than centered.
2. I used image-manipulation software to "straighten" the lines of latitude and longitude. I used Gimp, which is a free tool available for Linux, Windows, and other platforms, but Photoshop and other tools can do the same thing. The tool you want is the "perspective" tool, which is typically used to make the edges of, say, buildings, parallel even though they converge in the photo taken from ground level. The lines on your chart photo will have the same effect. Instructions for doing this vary by software and can be found online.
3. After getting the lines as straight and parallel as possible, I cropped the image so that each edge corresponded exactly to a specific tenth of a minute in latitude or longitude. The chart I used has a 0.1' grid, which made this easy.
4. My charts are in "portrait" format and I did all the above manipulations "landscape" (easier on my landscape computer screen), so I next rotated the image to north-up.
5. I saved the image in PNG format, one of several that imgkap accepts. It's easiest if this final image is saved in the same folder where the imgkap program lives.
6. I opened a command window and ran the imgkap program, feeding it the image I created along with the coordinates of the NW and SE corners, and the name I wanted for the chart cell.
7. I moved the cell into a folder accessible to Polar View, then added it to the chart database with the "Add Cell" button in chart manager.
8. Once the cell was loaded, I checked for accuracy by seeing whether my waypoints, copied directly from the chart, coincided with the printed waypoint crosshairs. They were close enough for our purposes.

"Printed" chart overlaid on the plotter, with strips of other, less accurate, charts to either side. Yellow squares with white labels are my hand-loaded waypoints.

I'm very happy with the way it came out, and we'll be able to make use of the two cells I prepared this way while we are in Bimini and at Cat and Gun Cays. The process is tedious and time-consuming, so paper chart copyright holders have little to worry about -- if I could have just purchased these files I would have. And I'm glad I only needed to copy two cells, although I might need to make one for West End as well, depending on where we head next.