When PDFs were edited, not commented

My preferred of proofing artwork to clients is to provide a PDF proof of the artwork from my Adobe InDesign file, along with specific instructions to open the PDF in Adobe Acrobat Reader, and use PDF mark-ups using Adobe Acrobat Reader’s comment feature.

In a perfect world, the markups would be returned from the client looking something like this:

This will allow me to take advantage of InDesign’s “Import PDF Comments” feature:

Or similarly via the Annotations plug-in from DTPTools:

Each method has its own strengths and weaknesses, but the goal is the same – to take the markups from a PDF file directly into InDesign to accept or reject alterations.

Unfortunately, we don’t live in a perfect world, and proofs can come back in a variety of ways:

  • Provided as a series of instructions, usually as bullet points in an email or given over the telephone;
  • Printed out by the client and marked up with a pen. This can be made worse if a red pen isn’t used; the client’s alterations are illegible or uses their own shorthand rather than proofreaders’ marks; or if the hard-copy alts are made into poor-quality digital images by scanning or photographing (or worse, faxing).
  • Submitted to an upload service that allows markups to be made on its platform; but not extracted and able to be imported into InDesign via Import PDF Comments or annotations (e.g. Box and Dropbox)
  • Markups are made, but using software other than Acrobat (e.g. Mac Preview) that have difficulty importing into InDesign’s solution or the plug-in;
  • Markups are made, but using markup tools that allow for subjective opinion (e.g. speech bubbles, arrows, drawing tools) rather than replacement, strikethrough or text addition. Speech bubbles have their place in alterations, but usually to indicate that a larger content change is required, rather than for small type replacements.
  • Markups that are duplicated and effectively “clog” the commenting panel (i.e. using more commenting than is required to take in an alteration such as the example below).
  • Rather than using markups, the client has actually edited the PDF with a PDF editor of some description and made the changes to the PDF itself (as shown below)

Most of these bullet points are a way of life with round-tripping of proofs, but the last point is the most frustrating when taking in alterations as:

  • InDesign or the plug-in literally have no markup instructions to take in, so no alterations appear in their respective alteration panels;
  • It can give the artist a false sense that no alterations were made as no markups are present;
  • If alterations are subtle, it can be difficult to tell where the alterations were made.

I’d like to say this outcome rarely happens, but the reality is that this happens far too often. My first impulse is to contact the client and inform them that the alterations aren’t usable and to use the Comment feature rather than the Edit feature, but that has the following drawbacks:

  • It is likely to frustrate and annoy the client, especially if many alterations were made. This is moreso the case if the client has followed the printer’s instructions to mark up a PDF but has mistakenly misunderstood the difference between the Edit and Comment feature of Acrobat.
  • Even if the client complies, it introduces errors such as alterations missed that were on the previously sent proof. It also takes time to prepare the alterations again, time that may not be available.

To use a card player’s metaphor, we have to play the hand that we are dealt and somehow compare the two files to determine what changes were made. I will also communicate to the client our preferred method of proofing to avoid similar incidents in the future.

But what ways can the two files be compared to take in the alterations?

Visual comparison

This can be done on-screen by either having both applications open between two monitors or one monitor with the windows split. It can also be done in an analog fashion by printing out the original and the latest alterations, then using a light table, overlaying each altered page over each original page and looking for differences. Unfortunately, both methods are time-consuming and subjective.

Visually overlay the PDF into the InDesign file

This involves placing the PDF of the alterations into the InDesign file, but on a layer above the artwork and with a transparency so that an overlay comparison can be made. To do this:

  1. Use the multipageimporter script with the following options to place all PDF pages into the InDesign pages on their own layer above the artwork.
  • Make a new object style with 25% normal opacity as its only property.
  • Use the following script by “Vinny” that will apply the object style to the imported PDF only. (This script works for documents less than 100 pages, but upon testing will throw a javascript error).

With overprint preview turned on, it will now become possible to see alterations that may have been made, and toggling the PDF layer on and off will assist in this process.

However this is still a manual, time-consuming and subjective task.

Kasyan’s comparison script

This is a script created by Kasyan Servetsky based on an article by Mike Rankin at CreativePro.

The technique in the article is used to compare two InDesign files by placing original and altered InDesign files into a temporary document applying different transparency settings to each file, and through the transparency settings being able to identify where alterations were made. This can still be applied in this use-case but an added step of an additional InDesign file that contains a placed PDF of the altered file and comparing between the two files.

Like the previous methods, it is still a time-consuming, manual and subjective task.

Dedicated file comparison software

Software such as Global Vision offers comparison software that loads both the original and altered files and performs a comparison that highlights the differences between the two files. It is worth looking at a video of the software in action.

It isn’t the only software that compares PDFs, and a brief search of the internet will yield several online services that perform similar tasks, such as:

  • Diffchecker
  • PDF Forge’s compare tool
  • Kiwi PDF comparer

That said, naming the sites above is not an endorsement, so if looking for an online option, make sure to perform all appropriate due-diligence before considering any provider.

Compare files in Acrobat itself

Acrobat does have a similar feature from the view menu where both original and altered files are compared between each other.

The results are highlighted, but the report and specific errors are not always as obvious as the results prepared with Global Vision’s software.

Using Acrobat’s compare files data as the PDF markup

There is a technique that can take the comparisons from Acrobat’s Compare files feature and treat them as markups. The technique is as follows:

  1. After the comparison is run, hit the close button on the top right.
  2. Navigate to the first page that has the compare results title page and delete it using Command+Shift+D.
  3. Save the resulting file.

The resulting PDF can then be imported using the Import PDF Comments feature from InDesign

Or by using the Annotations plug-in by DTPTools

Note that the plug-in displays the three changes that were highlighted in the comparison document, but InDesign’s Comment Import only displays two, while acknowledging that there is a third somewhere on the page.

Consider other round-tripping solutions within InDesign

There are several third party solutions available from the Adobe Exchange that allow round-tripping via InDesign such as:

  • GoProof;
  • inMotion;
  • PageProof;
  • ProofMe

The advantage for clients is that rather than opening the proof in Acrobat, clients are directed to a website where alterations can be made. This avoids clients inadvertently editing the PDF and instead allows them to provide changes that will need to be made by the artist.

Having tried some of these proofing systems, one thing in common was that alterations that clients could make was only in the form of comments, rather than strikethroughs or additions that are possible with the PDF commenting tools. These services usually require a log-in system which can be a hurdle, and are usually paid services.

InDesign’s Share for Review

InDesign 2020 and above does contain a feature called Share for Review that works in a similar way to these third party solutions, though the 2021 release allows for text highlighting, strikethrough and additions as well.

Another advantage is that clients no longer require Adobe Acrobat or a PDF reader to open the proofs, only a web browser. Check out Daniel’s video over at Bring Your Own Laptop to see this in more detail, along with other 2021 update features.

It is worth pointing out that this is not Adobe’s first attempt at a proofing solution, with an earlier system called CS Review introduced in May 2010 and then deprecated in April 2012. It is also worth pointing out that Share for Review is a feature offered in InDesign that – at the time of writing – has no comparison from competitors such as Quark Xpress or Affinity Publisher.

So far as my own work goes, this proofing method was not considered when Share for Review was released in June 2020 as the markups were limited to pin and drawing tools. Additionally, the release of the expanded tools happened during a peak-time in our production, and was too difficult to switch clients over from the PDF round-tripping method to this method in such a short space of time. It was also too early to gather other user input about the experience and bugs, so more feedback was required before considering this as a real-world solution.

Now at the time of writing with the expanded tools, I will begin trialling this method and report my findings once I’m confident there is enough to report.

Last word on this article

Up to this point in time, PDF commenting has worked effectively as a round-tripping solution from my perspective in the majority of my work, though it isn’t without its issues such as:

  • Establishing the process with clients, especially with staff turnover as the process needs to be established and explained to ensure that a client will not only mark up a PDF (rather than make changes to the PDF itself), but that the markups are prepared correctly and efficiently;
  • Proofing large file sizes;
  • Proofing to clients who are at the mercy of their IT department’s rules as to what software or websites they can or cannot access;
  • The Adobe Acrobat software itself, considering in a previous version the Acrobat team removed features that most casual users of the software would consider essential (much to my frustration until customer demand made them reinstate it) and how the software will be supported in future releases and for future operating systems.

What have your experiences with PDF comments been? Do you use a similar round-tripping method or something different? And are there any technologies in this space that haven’t been mentioned? Leave your comments below.

Why are food labels hard to read?

During an overseas holiday, I was asked by a fellow traveller what I did for a living. Because the term “prepress operator” is esoteric, my response was to say that I make food labelling and other printing. However, I was caught quite off-guard by the traveller’s response, which was “why do you have to make labels so darn hard to read?”.

That moment has stuck with me for some time now, and in this article I’d like to answer that traveller’s question comprehensively in an Australian setting.

In short: Limited real-estate vs too much information

That’s it in a nutshell – labels tend to be small in size, usually because the product that they contain is small in volume. Despite the small size, there’s also a lot of information to include within that size.

Details that have to be displayed by law

As a disclaimer, I am not a lawyer and this is not constituting legal advice on all specifications for a label.

If retailing food, the code developed by Food Standards Australia New Zealand (FSANZ) details what should appear on food labelling. More information can be found on their site, but in short a food label should have:

  • A Nutritional Information Panel;
  • The ingredients and additives by ingoing weight, and in some circumstances the percentage of ingredients. For example, if the label is identified as strawberry jam in the branding, then the percentage of strawberries actually used in the product has to be given.
  • Allergen information
  • Directions for use or storage
  • What the food is (e.g. is it strawberry jam, cookies, kombucha, etc)
  • The contact address of the manufacturer
  • The measurement of the weight or volume of the product
  • Country of Origin label

Some information can be presented on the label but may be presented elsewhere on the container, such as:

  • Batch or run of the product being produced
  • Best before or Use-By dates

In addition, the position of where the measurement of the weight or volume of the product can appear on the label is rather rigid and structured.

Country of Origin

From 1 July 2018, food products sold in Australia must display a Country of Origin label. This label has a specific appearance that must be adhered to, along with guidelines of how the label must be phrased.

An example of the Country of Origin label

This is enforceable at law by the Australian Competition and Consumer Commission (ACCC) – the Federal Government’s consumer watchdog which can – and does – give hefty fines to corporations and individuals for breaches of the Australian Consumer Law (ACL).

There is an online generator to create such labels, so having to maintain an array of assets for this task isn’t required – simply generate a label as and when they are needed.

Details that may also be required

Barcodes

Note that in the above code, that a barcode isn’t required by law. However, if the product is to be stocked on a large scale through distributors in a logistics chain, then it will usually need a barcode. In Australia this will usually be an EAN-13 barcode but in North America a UPC barcode is more likely.

Barcodes also have limited scaling, and GS1 (a provider of barcodes and related logistic products and services) indicates that a barcode should be placed at no less than 80% of its recommended size.

Specifications from GS1 relating to EAN-13 barcodes at 100% size

In GS1’s more detailed instructions on their website, they also recommend specific locations and orientations for the barcode, so can’t just go anywhere at any size.

These recommendations sound strict, but are done with the best of intentions so that a barcode will scan first time, every time, regardless of where the product was purchased. Remember, it’s a retailer’s aim to get that product from their shelf to your pantry or refrigerator with the least amount of fuss.

The size of the barcode can also be influenced by the retail chain that is selling the product and may recommend not proportionally scaling the barcode at all but leaving it at 100% size. It is best to check with the retail chain’s specifications first.

Container Deposit Scheme

As at the time of writing, all states and territories in Australia (except for Victoria and Tasmania) have some form of container deposit scheme, usually for soft drink containers. These schemes encourage consumers to return the containers to a collection facility in exchange for a refund for each item returned.

There are no national guidelines as to how the deposit notification should appear, so check with the relevant authority in each state.

Three items commonly seen on flavoured milk cartons (from top clockwise): The Country of Origin label; The Australasian Recycling Label; and The Container Deposit Scheme.

Alcoholic drinks

While a Nutrition Information Panel is not required for alcoholic drinks, other information is required, such as:

  • Alcoholic content (ALC/VOL)
  • Standard drinks statement. This can either be a written statement i.e. CONTAINS APPROX. X.X STANDARD DRINKS or in the form of a graphic with its own requirements.
  • As of 31 July 2020 and to be phased in over three years – a pregnancy warning label.
Pregnancy warning label, phasing in began on 31 July 2020.

Additional voluntary icons or graphics for marketing purposes

Health star rating

This is a voluntary system to aid customers in making healthier choices.

Third party endorsements

Endorsements from third parties that require an audit trail to guarantee the claims made about the product or its packaging, such as:

  • Organic certification, such as NASAA or Australia Certified Organic to certify that a product is organic;
  • Forest Stewardship Council (FSC) to certify the product or its packaging is from a sustainable source;
  • Cultural preparation certifications such as Halal or Kosher.

These types of logos won’t use a generic logo, but will usually contain a unique identifier such as a number that can be tied to the particular supplier.

These logos usually have strict stylesheets that have to be adhered to such as size, color, position in relation to other elements etc, and these guidelines have to be taken into consideration when juxtaposing the artwork for the label. The third party (or its auditing agency) usually requests to see the artwork prior to approval for print to ensure that the logo is used appropriately and within its guidelines.

Australasian Recycling Label

This is a private initiative to reduce landfill by informing consumers how the packaging can be disposed of once the product is used. The logos aren’t in the public domain and permission must be sought for use of this artwork.

And of course, our own branding

With all of the other details now on the label as required, it’s time to add the branding and design to the label… if it will fit!

This is just for food!

The details listed in this article are only describing what goes on a food label in Australia and aren’t exhaustive – this article doesn’t even mention regulations that concern pharmaceuticals or industrial products that may contain harmful chemicals.

Don’t even try being creative for cigarettes

While cigarette cartons aren’t labels, it is worth mentioning this as it is currently unique to Australia. Since 1 December 2012, any tobacco product (including labels on loose tobacco such as roll your own cigarettes) in Australia must follow the plain packaging guide, which – on a regular pack of 25 cigarettes – features such marketing gems as:

  • The brand and variant must be in Lucida Sans font in PANTONE Cool Gray 2C;
  • The pack surface must be in PANTONE 448C (a greenish-brown color);
  • A warning statement and graphic that covers at least 75% of the front surface without spaces separating the statement or graphic; and 90% of the back surface.

And this is just for Australia!

This article has focused on Australian food labels as they are the ones I am most familiar with and see on a daily basis at work. However, after spending time living in Canada, I recall seeing products that were sold nationwide (with exceptions of course) having to be in both English and French – good luck with getting all that information to fit now!

Can’t the label be bigger?

That can certainly provide more real-estate for the information to be displayed, but the size is often determined by:

  • Price of label production. Larger labels usually cost more than smaller labels that have otherwise identical print specifications such as stock, inks and embellishments.
  • Size of the package itself. An A4 page simply won’t fit on a cylinder that holds 100ml of a product;
  • How the labels are applied. For example, if machine applied, there may be specifications that a larger label can’t fulfill that a smaller label can fulfill.
  • Marketing. It may be more desirable for a consumer to see more of the product in the glass jar as opposed to a label that would obscure the contents.

Your thoughts

I hope this article has explained why labels can be hard to read – put simply it is to cram all of the required information into a tiny space. I’ve tried to cover as much as I could without trying to encompass all labelling, but no doubt there are items that I’ve missed or aren’t applicable in your part of the world. Leave a comment below on anything I’ve missed or if there is anything specific to labels in your country.

Planning your Wall Planner in advance

Between September and December, most of my working days are devoted to preparing school diaries and planners. One pain-point that I encounter with some diaries is the addition of a year-to-two pages planner. The planners usually look like this:

Apart from the events for each day that are on a layer above and updated each year, the base planner itself has two major wholesale changes:

  • The initials for the days of the week; and
  • The shading for weekends.

The initials are usually 365 threaded text frames that sits above a table containing the weekend shading. The text frames contain the initials of the days of the week, corresponding to the month and day they relate to, so changing the initials is easy. However, shading the cells containing Saturdays and Sundays on the table below takes time. Even if cell styles were used, they still have to be removed from the previous year’s appearance and applied to the appropriate year’s cells – subject to operator error.

Ideally, I was after a solution that would:

  • Easily update all planner base dates in one go
  • Shade all of the weekends without having to do this manually
  • Adjust automatically for leap years
  • Reduce any mis-dating via operator error

Didn’t I already write a script for this?

In 2018, I’d prepared a script to assist in the creation of wall planners that would allow for four types of layout configuration; and in 2020 was further improved to allow an output in one of 14 languages.

That said, this article isn’t meant to be a shameless plug of this script. The script is a great solution for creating new yearly planners. However, this solution requires updating existing planners. An additional solution had to be created.

The simple solution first

As I mentioned earlier, linked text frames that contain the appropriate days of the week initials. However, rather than use one paragraph style for these initials, each day of the week was given its own paragraph style e.g. Monday, Tuesday, etc.

The paragraph styles are based upon a base style that defines its basic appearance, with flourishes added to Saturday and Sunday in the form of paragraph shading. Each paragraph starts in its own text frame, and each paragraph style also refers to a next style feature to go to the next day’s style.

The days of the week are easy enough to prepare. The first way is quick but allows for operator error, and that is to type the 7 initials and their respective line breaks; and copying them 52 times, starting January 1st on the appropriate day.

The second (and more accurate way) is to make a new Excel file, type the start date in cell A1 and then with the cell selected, go to Fill, Series, and fill out the dialog box as shown

Once done, go to cell B1 and type: =CHOOSE(WEEKDAY(A1), “S”, “M”, “T”, “W”, “T”, “F”, “S”) and press return.

In the resulting cell, double-click the green square on the bottom right corner of cell B1 to automatically fill the remaining entries in the column.

From here, select column B, copy and paste it into the threaded text frames in InDesign.

Note – if it came in as a table, make sure to briefly change InDesign’s clipboard handling preferences to handle text and tables from other applications as text only.

Last thing to do is apply the paragraph styles. Take note of what day’s initial appears on January 1 (Friday in this case) and with the text cursor still visible in the text frame, select all type (Command A). From the paragraph styles panel, right click on the Friday paragraph style and choose Apply “Friday” then Next Style.

But this solution only fills two of the four criteria:

  • Easily update a planner dates in one go
  • Shade all of the weekends without having to do this manually
  • Adjust automatically for leap years
  • Reduce any mis-dating via operator error

This solution does not take February 29 in leap years into account. In these instances, an extra text frame needs to be threaded to take in February 29, so is not completely automatic.

The over-engineered solution

I did prepare a solution outside of work hours that would meet all four criteria; but the setup of the file took more time than simply adjusting the previous planner. It would also be difficult for other operators unfamiliar with my techniques to carry out the alterations unless they were trained beforehand. Ultimately, the solution wasn’t practical for my day-to-day work and was never implemented, but it is worth looking at the solution as it uses a variety of InDesign’s features such as:

  • GREP styles
  • Table and Cell styles
  • Rule above within Paragraph styles
  • Linking to an Excel file

Both the InDesign file and Excel file can be downloaded here.

First thing is to create an XLS file that contains the yearly information that changes and link it in InDesign. Linking to an excel file can be done via InDesign’s preferences.

Linking to an XLS file can be fiddly and I’ve found the best way to maintain appearances of linked tables that are updated is to make sure that table and cell styles are applied to the table that will contain the incoming XLS data.

A glimpse at the XLS file shows that it contains the initials of the appropriate days of the week. Note that the initial letter is represented by two letters rather than one (which will be explained shortly). The year is off to the right-hand side for ease of updating the planner by just adjusting the year.

An even closer look at the XLS file shows hidden columns. If we unhide these columns, the workings behind the technique are revealed. To compensate for leap years, the calculation for February 29 has an if-else statement that says if a date appears here, add the appropriate weekday initial, but if there is nothing, leave it blank.

Back in the InDesign file, all cells contain a paragraph style that have a GREP style instruction that makes the first letter minuscule in size and width.

Why two letters for the days of the week and not one?

It is to do with how the shading of the cell is automated. The first letter is used to identify a weekend with the initial S and that letter is colorised with an appropriate color for the weekend. If the letter was a weekday, the first letter is set to none. To control the background color of the cell, the rule above feature of the paragraph style is used as it allows the [Text Color] option to be selected, and this color is the first character that the paragraph style encounters. The rule above is then made thick enough to fill the cell, and then the offsets are applied to fill the remaining parts of the cell.

Why not use the paragraph shading feature? Unfortunately, this feature does not have access to the [Text Color] option from the color dropdown.

Updating the data can present a trap, as instead of simply changing the date in excel and updating the file, the art needs to be relinked and then selected with the show import options dialog to only bring in specific cells, as the date at the far right isn’t required. In this instance, it is changing the cell range from A1:AA32 toA1:Y32 as shown:

So this will make the base for the planner, and all the further information into the particular dates is done on a layer above using a table with the same cell dimensions.

As mentioned, this is an over-engineered solution that wasn’t implemented, but its use of combining linked Excel files, GREP styles and Cell styles may have application for other purposes.

GREPgraphing – Beyond 2000

In 2016, I developed a concept of creating bar graphs using GREP styles. Put simply, when a number was typed such as 1423, several different GREP styles would kick in to transform the number into a bar graph. In the above example:

  • if there’s a one for the 1 followed by three digits, make the one character 1000% wide;
  • If there’s a 4 followed by two digits, make the four character 400% wide;
  • If there’s a 2 followed by one digit, make the two character 20% wide;
  • If there’s a 3 as the last character, make the three 3% wide.

To successfully work with any whole number between 1-1999, this technique requires creating 29 different types of GREP styles and character styles. It uses a fixed width font and takes advantage of GREP styles to adjust the width of the numbers to represent them as their appropriate value as a bar graph. I’ve written about this technique in more detail over at InDesignSecrets.

In 2018, I expanded upon this idea and made a script so that anyone wishing to use the technique without typing the 29 character and GREP styles could simply run a script:

Beyond 2000

One catch with this technique is that it stops at 1999 as the character style to represent 2000 can’t be made by creating a 2000% width as the maximum character width is 1000%. However, there is a workaround – create a new character style called 2000, based on the 1000 style, but give it a tracking of 600.

Adjust the GREP graph paragraph style and underneath the 1000% GREPstyle, make a new one using the new 2000% character style that was made and give it the following pattern:

2(?=\d\d\d\b)

To test this, I have two different GREPgraph styles containing the 1000 figure and have put them side by side. In the text frame underneath, if I type 2000 into the text frame with the GREPgraph beyond 2000 style applied to it, it should be the same width, right?

Absolutely. If I make an additional 3000% character style based on the 2000% style but changed its tracking to 1200, and added a GREPstyle to the paragraph style to apply the 3000% character style to the pattern 3(?=\d\d\d\b), this will work too.

From here, a pattern can now be seen: to get to the next 1000 in width, add another 600 to the tracking of that thousand’s character style and add a new GREP style pattern to the GREPgraph style. This will work until the tracking hits its maximum of 10,000.

Doing all this in less GREP steps

As stated earlier, the original technique required 29 GREP styles to create a GREPgraph, but what if I said it was possible to achieve the numbers 1-1999 (actually a few more – 2047) using 12 GREP styles? This can be done too, but requires a technique in Excel, and that is to convert the original base-10 number into a 16-bit binary number. I’d actually described this technique back in 2011, but it was implemented differently and with more complications: this method is far simpler.

First, convert the desired base-10 number to a 16-bit binary number in Excel using the following formula:

=DEC2BIN(A1/256,8)&DEC2BIN(MOD(A1,256),8)

Now that the number is displayed in binary, copy the binary number into InDesign.

The technique basically works the same way as the original GREPgraph technique but has a slightly different implementation. First, the character styles need to be created that will change the widths, but instead of making 1-9, 10-90, 100-900 and 1000, only the character styles 0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512 and 1024 need to be created. For example, the character style 32 will only contain the width of 32%, and this pattern will apply to the other numbers mentioned.

Zero and 1024 are the exceptions. Zero will have a character style featuring a .1pt size, no fill or stroke, and 1% width. 1024 will have a character style of 1000% and a tracking of 14.

This will create a total of 12 character styles.

Now to make the paragraph style. To keep it simple, the appearance of the bar graph will controlled by the paragraph rule (though more advanced methods can be made using combinations of rule above, below, underlines and strikethroughs).

Here is where the magic happens – the GREP styles. Here are the styles and how they are implemented in the GREP style panel of the Paragraph style.

And here is the result:

So what’s happening here?

It’s important to understand how binary numbers are written compared to base-10 numbers. The following site gives a really good explanation, and that’s all that’s happening here – the GREP styles are being applied to the 1s in the binary number, based on their position in the binary number, and applying the appropriate GREP widths.

It can go larger

Just like the regular GREP graph, this too can be expanded to numbers larger than 2000 by creating an additional character style that has an added tracking applied. In this case, to represent the numbers 1-4095, a further character style can be added: 2048.

An advantage of this technique is that instead of creating an additional 9 character/GREP styles for each power of 10 using the base-10 GREP styles, the binary method will require a new style every time the total doubles e.g. 14th GREP style would be 4096, 15th GREP style would be 8192…

The obvious disadvantage with this technique: the numbers to be displayed as bar graphs have to be written out in binary first.

Is this something I would use? Probably not myself, as GREPgraphs are normally enough for my purposes, but I often hear the question “what if you need more than 1999 in a graph” so I felt that it was worth answering the question: It can be done, but it’s a lot of effort to get there.

%d bloggers like this: