Archives for category: Drawing

2014-Digital-Epigraphy-cover

The Oriental Institute has released a new book titled Digital Epigraphy, a manual for the methods that they are using to record items found in their Epigraphic Survey.

During the past several decades the Epigraphic Survey has refined its conventions and recording methodologies to fit with the widely divergent nature of the inscribed surfaces we record and the changing conditions in Egypt that are resulting in the accelerating decay of those inscribed surfaces. For the past two years we have been experimenting with new digital tools, software, and equipment that have allowed us to streamline our recording process while still achieving the highest degree of accuracy, the bottom line of any scientific documentation. It has always been our aim to share these conventions and methodologies with our friends and colleagues, and it is our great pleasure to present the initial results here now. The digital formats in which this manual is made available are particularly appropriate and will be updated and changed regularly, since the manual will always be a work in progress. The possibilities are limitless.

The free volume is available from their website as either a PDF or ePub book.

Much of the volume is a primer on the use of Photoshop for photo manipulation, and Illustrator for the final vector drawings. They cover layer structures, brush selection, curves, manipulating vector paths, pattern creation, and more.

digital_epigraph_epub_p108

Screenshot from p. 108 of Digital Epigraphy

In short: this is essential and the manual that I wish I could write for those creating digital drawings in the field or in the research lab. I plan to make it required reading.

This post was written with the help of Greg Tucker of the BSR, who is our new project CAD developer.

The new drawing workflow is working better than expected with TouchDraw’s unlimited paper size and ridiculous zoom levels. The iPad drawings this summer are produced directly on top of the AutoCAD files instead of being traced. This gives us an almost-spatially-correct drawing environment. Almost because we do use a local coordinate system which we have chosen to ignore in TouchDraw. So while the measurements are 1:1, the plan is not in the local grid, and as you will see below, north has changed.

The overview of the process is that the relevant area is exported from AutoCAD in svg format, opened up in TouchDraw, and then cleaned up and made ready for field drawings. Once the drawings are finished, they are again exported to svg, converted from svg to dwf using Inkscape and then back into AutoCAD. There are some minor adjustments to be made in AutoCAD to get the drawing to fit back into the spatial grid.

Here are the details:

Creating SVG files for TouchDraw (And reintegrating them with the CAD model)

This season we’ve decided to draw on the iPads in the field using TouchDraw, which can export in Scalable Vector Graphic (SVG) format. This makes integration of the drawings into the CAD model much quicker, just needing to assign layers to the features and drawing elements as opposed to digitizing either raster images or on a drawing tablet. Our site plan is being made using AutoCAD, which unfortunately does not recognize SVG for either import or export (as of the 2012 version).

The first step to drawing on site was to export a base plan of the trenches so that the excavators could begin planning directly on their iPads. After exploring a few options available we decided on DWG to SVG Converter 2011 MX by DWG Tool Software to create SVGs from the CAD models. Initially all our attempts to generate the baseplans for each of the four trenches resulted in files approaching 10MB, much too large to easily and quickly manipulate. Even when deselecting layers we didn’t want exported, deleting hidden lines, changing the compression, etc, the file sizes remained at this bloated level. The immediate solution we have been implementing is to copy the objects necessary for each of the baseplans (wall bases, architecture, trench extents, baselines, and four target points to ease with scaling and rotation) to a new DWG so as to remove the extraneous data. In the options menu in the DWG to SVG converter we select “Last Saved View” and adjust the output size to match the view dimensions in the CAD model. If we are viewing plan of trench 53000 for instance the dimensions of the view window could be 26.55×11.62m and the image size we would output would be 2655×1162. The resulting SVG has a scale of 1px:1cm and is small enough, <500kb, to be easily manipulated in TouchDraw. Additionally, to aid drawing in the field the CAD model is rotated so that the baselines are aligned vertically before SVG creation. The final output then allows the excavators to use the gridlines within TouchDraw to draw more quickly and efficiently.

After the excavators draw their plans the resulting SVGs are loaded in Inkscape, which is compatible with both SVG and DXF, and saved as a DXF. The DXF can then be opened by AutoCAD, the objects can be rotated and scaled quite quickly using the target points and then copied into the side plan DWG.

There are potentially ways to streamline this process even further and we will be exploring different methodologies throughout the season.

output from CAD for trench 56000

Once the svg file is exported from AutoCAD, it is transferred to TouchDraw via iTunes. In TouchDraw there is an import command to convert the drawing to their native format.

Objects in the drawings were pulled out of the flat svg file into their different layers in TouchDraw and assigned our more standard drawing conventions (orange dashed line for the trench outline and blue dashed line for the baseline, etc.).

Plan with layers in TouchDraw.

The final drawing is given a name to indicate that it is the baseplan. That is, you do not open this one and start drawing. This one is duplicated and the resulting file named “trenchnumber plan x”

When the drawings are finished, they are saved in the native TouchDraw format. Then the Walls, Trench Outline, and Scalebar layers are turned off. The plan is exported into svg format with only the registration points, baseline, and any features or SUs drawn in. The resulting file is given to the architect for importing into the CAD plan.

The native svg output and TouchDraw files can be downloaded from here.

This is also my first post from the iPad. I am testing blogsy.

While playing with TouchDraw the other day I realized that the unlimited page size would help us get rid of another piece of complexity in our field drawings. Our drawings are usually done at 1:20 scale since this is the largest scale that will fit on an A3 sized sheet of paper. The standard method of drawing is to have something static, like the sides of the trench or a baseline, and measure an object from those static lines. Measure from two sides, find out where they collide, and you have a point that you place on your paper. Measure enough of those points and you can connect them with a freehand line that represents your object.

The problem is transferring the ruler measurement to the paper. You could use an architect’s scale to handle the math; or after enough time, you can do the math in your head. Whichever you do, you have to convert the measurements on the ruler to the measurements on the grid paper. Last year we worked in pixels, not centimeters, so things were slightly odd, but this year we can use the unlimited paper size and units capabilities of TouchDraw to help us even further.

The test drawing below is a 1:1 drawing of a trench with an SU in TouchDraw.

There is still a little bit of math to do. The numbers on the rulers are centimeters, not meters. But almost everything we measure ends up in cm.

Drawing at this scale is a little tricky. The lines, for instance, are anywhere between 15 and 30 (I assume pixels) wide. The text is 300 (again, I don’t know the unit, but I assume pixels or points). But dropping the necessity for scaling makes converting the ruler measurements so much easier.

Note the blue lines above. Those are my ruler lines. If I make the strictly horizontal or vertical (by using the function lock, the little fx on the lower left that acts as a shift-constraint), I can type the length of the line in the overlay (the gray HUD on the right). Where the two lines join, I can make my feature. When I don’t need the lines, I just delete them.

I was able to add the elevation symbol and the SU name to the library easily enough so they can be used over and over. The elevation symbols are a group of three lines and a text box, but if I lay one on the page and double tap it, I can edit the number on top very easily.

I worry some about the line thickness. This is something that you don’t get in CAD environments, but you have to deal with in straight vector or paper drawings. As you can see from the second screenshot above, the lines are roughly .5cm thick. This is roughly the equivalent as the thickness of a line on pencil paper, which is why archaeologists tend to aim for centimeter accuracy. But that might not always be desired. If you take a look at the top image again you can see that we can remove yet another leftover from the old pencil drawings and that is our dependency on pencil lines altogether. The SU shape has no line, just a semi-opaque shape.

The original TouchDraw document can be downloaded from here for experiments.

We leave for the field in a couple of weeks. In the meantime I am gathering some comments from the field supervisors and getting ready to make and test the last additions to the database before we go.

This spring we are re-evaluating our software for the summer excavation season. The iPad was so new last year that we went with what was available and while it worked, it could have worked better. Now we have another whole year of software development with some more sophisticated apps. Since drawing is the most difficult task to learn on the iPad, I am starting with those apps.

The PARP:PS team in Cincinnati will be meeting to start evaluating these apps as a group. The first one that we will look at is a new drawing app named TouchDraw, which has a lot of features that I like, and only a few quirks.

This post is written in the style of a software review but with a specific purpose: for drawing technical drawings in the field. There might be some features that make one piece of software stand out among the others, but if it doesn’t do me any good, I don’t really mention it here.

I will be posting some homework assignments to the group, but they will be different from the assignments from last spring. Instead of tracing a pre-drawn trench plan, I want them to work with the software in the same way as they do in the field: to draw from scratch. That means some measuring and triangulation to get the items where they belong. To make it easy for them, I will ask them to draw their desk and some of the items on it. Here is mine (more difficult than most because of the curve). It isn’t exactly what you would find in the field, all the items are pretty geometrical and not irregular, but it allowed me to learn a lot about how the software works.

Documentation

The documentation for TouchDraw is pdf based and needs to be downloaded before reading. There is a Quick Reference Guide and a User Manual. The manual is well written and includes screenshots of the features mentioned. Therefore, I did not include any screenshots in this post.

Beginning

Documents are easily transferred to and from the iPad via iTunes. There is also a cloud option, which reads and writes documents to a Dropbox folder. You can move or copy drawings from the local storage to the cloud directly from the iPad. A unique feature is the ability to save as… so you don’t have to leave the drawing, go to the browser, duplicate the drawing and open it back up to start the next one. Save as… will let you start on the new drawing immediately.

Setting up Scales, graphs, layers

We usually only create a single new drawing from scratch for each trench at PARP:PS. Once the base drawing is complete we duplicate that drawing for other uses. The basic setup for each document is done in the info menu. There we can select the units for the drawing. We use cm. TouchDraw acts more like a CAD environment than a page layout program in that it does not have a page size. Drawings are essentially unlimited in size. This makes it easy to have drawings of various scales, so you might want to make that explicit in the title box for each drawing.

The rules on the graph paper dynamically appear based on the level of magnification, which is nice.

TouchDraw has a freakishly large zoom level. I reached 216,437,286% before I got bored. At any rate, you can zoom to your heart’s content.

Interface

The tools are lockable. That is, if you tap on the rectangle tool and draw a rectangle, after that is complete the pointer tool is selected by default. But if you double-tap a drawing tool, it lock that tool until a different tool is selected.

In the tools menu there is a snap to grid option. That appears in most drawing programs and is handy for drawing the scale bar. In the tools menu under the general settings section is a toggle for snap to lines. This doesn’t appear in the documentation but it allows you to snap the ends of two line segments together. The line segments don’t join, but they do connect. There doesn’t seem to be a way to snap the end of a line to the middle of a line segment. Nor is there a way to snap closed shapes, only open lines. Snapping also won’t work if the line you want to snap to is locked.

Also in the tools menu are a couple of useful features in the drawing view section. Easy drag mode allows you to move shapes without fear of moving just one handle of a shape. Crosshairs puts full-screen crosshairs when you touch the screen. These extend all the way to the rulers and are quite useful. The show overlays option puts a heads up display (HUD) on the screen with information on the object you are creating/editing. This feature is incredibly useful for technical drawings, as it allows you to see the size of objects as drawing, but there are aspects of it that can be improved (more on that below).

Adjusting the elements of objects is fairly simple and is surprisingly absent from other drawing programs. Once an item is drawn you can select the adjustment menu item and change the position, width and height of the object. There are three oddities to this behavior. One is that I find the precision to be much more than I need. I don’t exactly need to know that the width of my rectangle is 0.89493525362219 cm. When making a technical drawing, I sometimes draw the rectangle at any size, call up the adjustment box, and hand edit the dimensions. In order to change that number to, say, 1.95 cm, I have to hit the delete key numerous times to get to the start to type in my dimensions. The second oddity has to do with the HUD feature mentioned earlier. For example: when you draw a line you see the line length and angle. But once you draw the line, you can never see the line length again. You can see the start and end points in the x/y grid on the HUD, but not the length. When you call up the Adjustment box you can see the overall width and height of the rectangle that the line occupies, but you can’t change the line length. The last oddity is that the position given for the object is always the upper right corner. There are times that I would like this to be the center position of the object, but maybe I am being picky.

There are an endless number of things that one might want to do to a vector object and various drawing programs have different methods for allowing access to those tools. Most of them use a contextual menu for items like duplicate, copy, paste, delete as well as some other more advanced functions like insert points, change to paths, etc. The standard way to get these is to hold your finger to the element for a second or so and then lift up. My experience is that this works only some of the time. But TouchDraw has a contextual menu item in the left menu bar that calls these commands without the finger trickery. That itself will save lots of drawing time.

New to us will be the Group option. We didn’t have that last year but the ability to group elements so that we can move them all at once is a good feature. Likewise the ability to lock and unlock. Last year we had to make sure items were on different (locked) layers if we wanted to avoid tapping on them inadvertently.

Another incredibly useful feature is the function lock (from the tools/general menu). This acts like a shift-constraint in most desktop programs. Activating it will keep the proportions of an item the same while scaling, and lock the direction of lines to horizontal or vertical.

The last great feature worth mentioning is the library. You can save graphic and text elements to the library for reuse. For example: elevation marks, title boxes, that sort of thing. This is something that we wished we had from the beginning.

Some of the terminology might be confusing. The handles that change the direction of a line at a point in Illustrator are called direction lines. In TouchDraw they are called quadratic curve segments. This took a little while to get used to but it is well documented in the User Manual.

The color picker is handy but I would also like to be able to store colors for reuse. Some other drawing apps have that feature and I was looking forward to using it this year. TouchDraw won’t allow that.

Auto Save

One of the big questions that I have is “what happens when the app crashes.” Some apps save the data anyway, some don’t. TouchDraw didn’t crash on me during testing, so I have no idea.

Ending

The documents can be exported as jpeg, pdf, svg, png, and visio.

Whether exported as pdf or svg, the documents open up in Illustrator just fine for additional editing. The layer structure, however, is not preserved in either format.

Since we are excavating an urban center with existing architecture, the main environment for spatial documentation is CAD at PARP:PS. This doesn’t exclude the use of GIS, as they are easily interchangeable, but CAD is where our base documentation lies. We have always used AutoCAD but, like most projects, get stuck with sharing the dxf or dwg files that AutoCAD produces, with both people who don’t have AutoCAD or people on Macs. Our old workaround was to use Illustrator to open those files, but we lose the accurate measuring ability. There have been two recent events that have changed all of that within AutoDesk.

The first is their Education Community. Here students can get a free version of AutoCAD to help them learn the software. It even includes the new AutoCAD for Mac. With that I can send documents to our trench supervisors (who are all grad students) and they can use the files to help them write their reports. The downside is that once a document is opened and saved with a student version, there is a permanent watermark on the file that can’t be removed, even with the professional version of the software.

The second, even better method is their free AutoCADWS website. I have written about AutoCADWS and their iOS app before, but I haven’t focused on the web version. Here you can upload your drawings in (they promise) a secure environment. You can invite others to look, comment upon, and even edit the files directly from the website. The interface is also much, much friendlier than the traditional AutoCAD interface, and thus can be used by mere mortals.

This has already changed our workflow. While we still house the main drawing on our file server, we host a copy on the WS website as well. Our architect can work directly off of the WS website version (a plug-in for AutoCAD allows connection to the WS hosted files), and everyone else on the project can view and comment on the document. If they want, they can also download the document from the WS website and keep a local copy (we traditionally put the AutoCAD drawings into Illustrator to clean up for publication). The website has group features that allow for a chat session while two or more people look at a document at the same time, and has a timeline feature to allow you to step back to previous versions of the document.

The only downside to the WS website so far is that it doesn’t handle 3D. I assume this is temporary but since we have only published plans so far, this hasn’t affected us yet.