The Art and Science of Describing Images Part 2

Just another WordPress site

The Art and Science of Describing Images Part 2

– [Richard] Hello everyone, and a very warm welcome to you My name is Richard Orme from the DAISY Consortium and I am your host for today’s webinar, The Art and Science of Describing Images Part II Whether we are creating accessible documents, adapted learning materials, or posting to social media, we need to know about describing images It’s a skill applicable across many job Roles and there’s always more to learn In our webinar, The Art and Science of Describing Images, our presenters introduced four golden editing tips to help you craft effective descriptions And they brought these to life by exploring examples of popular image types, from Shakespeare to pancakes, via Freddy Mercury In part II our experts will dig deeper, covering techniques for complex images, tables, charts, infographics and maps It sounds fascinating, so let’s get started At this point I’ll hand over to our panellists to introduce themselves – [Valerie] Thank you, Richard Hi, this is Valerie Morrison And I work at the centre for inclusive design and innovation at Georgia Tech in Atlanta Georgia, and I manage the e‑text department We basically focus on converting text books and course materials into accessible file formats for students with a wide range of print related disabilities, learning disabilities or just individuals who access material using Assistive Technology – [Huw] Thanks so much Valerie I’m Huw Alexander I’m the Director of textBOX We are focused on providing image description services for the publishing industry and media in general Hopefully we can share tips with you today It’s lovely to be back in tandem with Valerie after our first session Trying to move on to the next slide, it’s not quite working As Richard said, this is the second part in the proposed trilogy You could say the empire strikes back in image description Hopefully I’ll find out who my father is at the end of it But as an overview we’re gonna talk about information and complex images, and how information is conveyed now We gonna talk through describing some complex images with some examples We’re also going to talking about complex images and how you can simplify them and break them down into their constituent parts, to make it an easier job to describe and an easier job to follow the information being conveyed And at the end we will take questions So please as Richard said, add your questions to the Q&A box as we go along, and we will be happy to answer those As we all know especially this year, the world is a complex place in a variety of ways And also in the digital space, it’s become an important area with digital learning and everyone accessing material, especially colleges from home and making sure that content is as accessible as possible What’s happened not just this year, but in the last decade or so, there’s been a real change in pedagogy, the way that information is conveyed to students Educational materials have shifted to a more visual form of conveying information If you look at a text book now and compare it to 10, 15 years ago, it was more text base then But now students are expecting and getting really fantastic text books coming through with infographics, autograph images, of ways of conveying the information in a consumable way That’s because students are demanding it They’re wanting to learn in different ways You can see that in a consumption of things like YouTube videos, and the way people are learning in education is portrayed through videos but also through text books as well So we are trying to help with that because otherwise there is a certain amount of content in there, a huge amount of content that is not being conveyed to the actual user So a vision impaired reader is gonna be missing out on so much great stuff For instance, when you’ve got a chapter and you have an infographic at the end that explains everything you should have learned in the chapter, if the alt text and image description just says image 31 it’s not very useful to the vision impaired reader

So they are missing out It’s almost like the book is censored Hopefully what we’re going to be talking about today will help you in describing those complex images We are talking about things like maps, infographics, diagrams, anything like that We will walk you through examples and hopefully take away that confusion so you can move very quickly in that description and create a description that is (indistinct) and mainly useful, so the students are not missing out on that really fantastic work that publishers are doing now in creating information in a very visual way So I’ll hand over to Valerie who will talk through examples I’m in control of the slides still, so it’s my fault if anything goes wrong So over to you Valerie – [Valerie Morrison] Thank you Huw, I appreciate that If I mess up I’m just gonna say it was all you That’s very comforting So I am going to run through a few complex examples of different image types to tell you how I approach them And please bear in mind that I have several years of experience in focusing mainly on writing alt text and how to describe and tear content out of the visual space and into a more accessible text space And I still find a lot of the images that I’m going to use as examples today, I find them daunting So really don’t think that this is beyond you because we are just gonna show you some examples and how we approach them, so you can feel more confident and more capable of doing this yourself So, the first type of image I wanna look at is a map This has a lot of information in it, this map There’s a great inset that shows you different types of information that are marked on the map like coal fields, industrial areas, ports, steel works, oil wells There’s a lot indicated on this map How I approach writing a map, and I think that there are all types of different approaches, but how I approach it is I’m always going to want to start with a very general overview sentence that sums up what is in the map So if the figure that you are describing doesn’t have a caption that proceeds it or a title, then I’m going to have a general overview sentence For example I would say a map of Europe, titled areas of industrial construction from the year 1870‑1914 period And then I might list using the inset as my guide, sort of like my table of content of the map I will focus on that inset and figure out what relevant details I want to focus on I won’t list every single country unless the context demands that I do I’ll just talk about the area indicated on the map in a general way, and then fill in details from general to specific as I go and as needed depending on my context If have space or the ability I would maybe create a list separate from the map So I might have my alt text be very general to describe the map in a brief way and then if I could insert a list after the figure of the different coal fields, where they are, list the industrial areas or the cities with high population counts to offer that information in a different modality Something that I think Huw is gonna focus on later or I have noticed looking at his slides he’s got a lot of very long alt text descriptions because some of these demands that you really go into detail, but if you are forcing someone to listen to that long description in alt text format, they don’t have the ability to pause, go back, rewind They have to listen to the alt text field in one shot So if you can take things out of the alt text and provide a separate table or list that might be another modality, another accessible solution for you So when I’m approaching a map again, general overview sentence and then the other thing that I wanted to mention

is that you don’t really need to focus too much on the colours of a map, what colours countries are shaded These purple diagonal bits You don’t have to say purple diagonal anywhere in your alt text description The symbols of the colours often have no significance and you can just focus on the meaning instead of the appearance of symbols Next slide please So here is a very very complex infographic with a time line and a little essay over to the left, and images and description of those images This is a lot This is one of those end of chapter, here’s everything you need to know situations that Huw was describing So again for this instance I’m going to use my tried and true approach which is create one general overview sentence where I describe if there’s a title, I’m going to talk about the title of the infographic, talk about it as a time line and then indicate the type of events listed on that time line So sometimes you’ll see here we’re focusing on events and people important to black history and then I will list the range of dates that the time line covers So that’s a brief description for all of this info One way if you want to translate all of this visual data from this image into text format, a great way to do that would just to be to creating a list You could list the events by date or you could create a numbered list if there aren’t dates indicated, if it’s just maybe eras You can have a numbered list and list all of the events indicated on the time line Next slide please So for bar charts, this is a very simple looking bar chart There are all kinds of bar charts There’s quite a range Line graphs, bar graphs, pie charts, all of them I approach the same way I begin with a title if there is one If there’s not, I’m just going to name what type of graph it is A bar chart titled in this case, Maternal Mortality in Selected States And yikes I am in Georgia And this map indicates that Georgia has a very very high maternal mortality rate In fact it’s more than double any of the other states I also noticed that not all of the states are listed And so from looking at this, I want to pay tribute to the fact that this graph was probably created to make a visual impact The different numbers are arranged in descending order and the highest rate is more than double any of the other states And so in my alt text description I’m going to wanna step back and pay tribute to that and mention that for people who cannot see the visual impact of this graph So often when people are translating data into graph format they are doing to highlight a particular idea or piece of data So my general approach to a bar graph is going to be mentioning the title, describe what is on the X and Y axis If I’m in a humanities context I may call them vertical and horizontal and not X and Y, because I still forget sometimes which ones which But if I’m in STEM I’m definitely calling them X and Y axis and then I will describe each bar in very regular predictable ways so that it’s very easy for the person listening to follow And then again, always stepping back, and sometimes a graph is just a graph, but in this case I think there’s definitely a visual impact the way it’s been arranged and the way that only certain states have been selected So I’m gonna want to mention that probably before I list all the data points If I’m going to go into specific detail, I wanna mention that before rather than later

Next slide please This is the kind of graphic that I really dread This is a supply and demand curve Once you have one, always copy and paste it into a document so you can save it for later and you don’t have to reinvent the wheel But it looks daunting but again it’s just a graph So I’m going to approach this by creating that overview sentence where I mention what type of graph it is It’s a supply demand curve or line graph with price on the vertical axes or the Y axes, yes? And quantity on the X axes And then I’ll describe the slopes or lines on the graph And I will note where these lines intersect at what points So it looks complicated It looks like you could get lost in a lot of word salad or letter salad with D, S1, S2, P1, P2 There’s lots of information here But it really can be much more simple A supply and demand curve with quantity on the X axes and price on the Y axes Two parallel positively sloped curves are intersected by a negatively sloped curve You could name them and get specific about the points So, I have a very regular way I describe these Next slide please This is a very beautiful, this is subjective because I love dinosaurs This is a very beautiful infographic but very complex, it has a lot of information on it We’ve got a time line down at the bottom that is shaded different colours to represent different eras And then above there’s a circle And at the centre is the earliest era And as we move outward to the edge of the circle the outer ring is the later eras Superimposed on top of that time wheel is a phylogenetic tree showing dinosaur evolution So as we progress out towards the outer ring of the circle dinosaurs are evolving later and later and later and getting more and more awesome I guess So I’m going to try and approach this the way I approach all my infographics All my complex alt text That one sentence that sums up everything and provides a framework for the person listening with screen reading software So I’m going to say complex infographic this was titled dinosaur evolution and then list all the parts that include a time line of eras ranging from this date to that date, with a phylogenetic tree and illustrations of dinosaurs and then work from general to specific from there filling in details as needed depending on the context It would be difficult to describe the appearance and anatomy and structures of every single dinosaur on this graphic So I may sum things up in one brief way saying the dinosaurs appear smaller and less complex in the centre at the earliest and as they evolve they get larger with more spikes, I’m not a dinosaur expert, I should be, the way I love them so Next slide, please This is a complex infographic that describes different cloud formations and has illustrations and arranges the illustrations based on their elevation The different cloud types and their elevation The way this is arranged as an infographic it allows a sighted user at a glance they can see all this different clouds types grouped together If we wrote alt text for this you would lose the ability to compare If we wrote it all out in a long paragraph

describe each of the cloud type and what they look like and their elevation it might get lost to the grouping ability So this is an example of an infographic that I would argue could function well as a table If you had a table where you mentioned the cloud name, the cloud description and then the elevation and then maybe the appearance as your column headings along the top, then you could have a table where you describe each cloud type and all of that information and someone would be able to compare it, and go back and tab through the table and it might be easier for someone to absorb that information in table form Next slide, please So, this is a complex STEM infographic about genetic code And I would argue that this is hard to parse even for a sighted individual who aced all of her science in college This is difficult This is a lot of information coming at you in wheel form It’s angry too There’s a lot of capital letters, a lot of bold, a lot of repetitive letters that kind of look the same, C and G look very similar So for someone who has dyslexia or someone who just didn’t want to be shouted out by a lot of capital letters this is hard to absorb So converting into a table with very specific column headers might be more accessible for more individuals Granted, someone has gone to a lot of trouble to create this graphic and visual learners might see information or see groupings or see connections in the visual format, which is wonderful and amazing, but a lot of people will have difficulty accessing this So, if you could advance to the next slide This is the same information translated into a table The goal is if you could provide multiple modalities for your learner, it will be accessible to more people You could have the visual and the nonvisual text form in accessible table form So here we have a table with 3 columns The column headings are amino acid, symbol and DNA codons And providing this data in table form will allows someone to tab through each cell or each column in order to access all that information In addition, I recommend if you do create a table that you have a title for your table and/or a caption to briefly describe what’s going to come in that table What the data is, what the table concerns so that someone listening with Assistive Technology can decide whether or not they want to enter into the table to read it They could skip it if they wanted to if they read the caption or a title Next slide, please And I just want to finish up by talking about structural alt text for tables So, in this specific example of a table we have a table titled physical properties of the giant planets And it doesn’t have a column heading in that very first column I think it wants you to assume that you know what those all are, what groups them together And really what that first column is about is physical properties So it’s in the title but I think to make it more accessible it should be repeated in the actual table itself So we have physical properties, and then the four giant planets and their different data points for each physical property So, if this were in accessible, this obviously on the slide is an image of a table But if you turn it into an accessible table I would recommend creating alt text for your table That is a thing that you can do in PowerPoint or Word

You can right click on a table just as if you right click on an image to edit the alt text You can add alt text to your table and what we do is we create what we call structural alt text where we just talk about how the table is structured So my example here is table 10.1 is titled physical properties of the giant planets It has 5 columns and 13 rows The column headings are physical properties Jupiter, Saturn, Uranus and Neptune And in that way I’m giving someone that overview of the table, so they know they have a framework before they start hearing all these numbers being launched at them They know the general layout of the table They understand and can get a mental map of how the table is arranged before they start listening to it So that’s not always a possibility Sometimes you are dealing with pictures of tables, but I recommend that your tables should have a title and a caption and alt text to be 100% accessible So, next slide please I’m going to hand things over to Huw Although, I don’t know if we wanted to pause for questions We could at this time or if Huw you wanna just describe your infographics for us – [Richard] It would be great if we could take a couple questions at this point in time, so we don’t hold them all to the end Huw, you are in control of the deck Maybe you take us back to the bar chart example while I ask Valerie the question from Oscar Thank you for your question Oscar Valarie, Oscar says, “You mentioned that a reader can’t rewind, “fast forward or pause the alt text “Is that a general rule “or is that based on what programme or technology “they are using?” – [Valerie] It is a general rule I’m using JAWS and NVDA to test the files that we make accessible And so based on the default settings you just hear the alt text all in one shot and you can’t pause it If you want, you have to listen to the entire description over again Now that of course is dependent on the version of the software that someone’s using or the settings that someone has There’s lots of people that can get into JAWS and find specific settings to get around that but the default settings you hear it all in one shot – [Richard] And the question specifically on this bar chart example comes from Jessica who says “For this bar chart example “it’s hard to be specific about the data points “because there’s a range if you’ve got the right one “of 10 between each label on the X axes.” Maybe it was a different bar chart “So if you are giving data per state “what would your approach be?” Would it be to say that Georgia has approximately 27? Or how would you represent that when maybe the number itself isn’t so clear to you? – [Valerie] So I would not want to say approximately every single time because it could get redundant and it’s a long word to take out each time So I would probably say the following data points are approximations and then list them all And then do my best to line up a piece of paper or something and list them Georgia 46, Iowa 18, Kansas 18, Utah 17 So that I’m not saying approximately over and over again I would just say it once and list the data points – [Richard] That’s very clear Thank you for that Huw, you have the comm, and you are picking up I think, from the slide to do with– – [Valerie] Infographics – [Huw] Just a quick point on bar graphs and bar charts I’ve been doing hundreds of these over the last couple of months The best thing publishers could do is provide data points Cause It’s so valuable to everyone besides readers as well As it was said, a lot of that is very very similar You can say these are estimated data points, but I spent so much time recently going back and trying to find the actual data

Going through census websites and things like that to find the information so that you can actually build the table It’s absolutely valuable content for everyone So if publishers are thinking about data points it just makes it a lot clearer to provide (indistinct) I was just thinking then about how this might be quite daunting for the audience Cause these are complex images And it’s okay not to get this right the first time you are writing a description As the great Terry Pritchard said you’re telling yourself the story The first draught of you writing it, the description of the story is you telling yourself the story You’re are taking the data points and information from that image collecting them and organising that into a description You can edit it You are not doing it live for the person So take your time over these descriptions because as you have seen with things like complex STEM graphic or (indistinct), a lot of information Take your time and work through the steps and look at the fantastic description at the end of it So describing infographics, description of an infographic tends to reference the following elements And as Valerie said this is areas that we’ve covered and is the title of the infographic and talk about the structure of the infographic And the numbers of the infographics are so wide spread and popular now And they are fantastic for conveying a lot of information in a snap shot It’s divided up into sections cause a lot of infographic are 4, 8 sections Tell the user the number of sections up front so they are aware and can visualise as they work through the infographic description Talk about the images and the diagrams incorporated into the design But make sure that you do include decorative images A lot of infographic include images that you don’t need to describe everything So if it’s got a pie chart, yes you need that data If it has a nice picture of a book because it’s talking about 100% of people read books, you don’t necessarily need decorative book image So talk about the information provided being at each section and repeat the text contained within the image Very often the case that image will just be an image is not readable for a screen reader So you need to extract that text from the image and that can be very straight forward if you got PDF, you can use OCR on that And you could pull out that data So you are not typing everything out each time As I said I need to describe the relevant images and number each list element within a section to help the user keep track of where they so they can jump back and forth and say I’m in section 2 and this is a list of 4 elements, and they will know exactly where they are, so you don’t get lost, especially with large descriptions For instance this is an infographic of infographics This is all the information people convey in infographics So it’s divided into 10 sections and so immediately you are visualising or creating that structure that framework for the user So it’s a design section on the left and a content section on the right And then it’s got sections underneath it, 5 sections within each section So it’s a lot of information here You start off and then provide a structural, the over view as Valerie talked about, the overview introduction and then talk about the structure and then go to details So you move from the scene setting to storytelling So that story telling is the details You end up with a massive description like this one I’m not gonna read the whole thing out cause I think everyone would leave Again we have that over view at the start of an infographic entitled infographics of infographics The introduction reads as follows And there’s an instruction brief here that helps you include that You take the text out of that and you include that in the description right upfront That’s the author’s introduction so you follow their lead Again the graphic is divided into two main section Design on the left, onset on the right Each main section is further subdivided into 5 sections The sections and subsections are five And then you’re basically creating a list Lists are your friend They organise things and provide structure

which is easily relatable by human beings We like creating lists of things Favourite albums or whatever We enjoy lists It helps us understand information in an easier way I’m not going to go through the whole thing, but you see each section is divided up into a list and you provide the data points related to that This is the description continued So you can see there’s a lot of information there You are providing it in a structured way to provide that information to the end user So describing choropleth A choropleth map is displayed quantitative values for distinct definable spatial regions on a map The description for choropleth map needs to reference the following data, actually the title of the map Again the structure of the map and the map being used, the text key to the colouring used to measure data Colouring is important for these types of maps We can look at a bar chart which is done in different colours and that’s not necessarily that useful You can still understand it from the data point But in a choropleth map the colouring can be very important indeed So you need to convey the colours and provide a key to the user A measurements scale if the work is not available You can provide trend analysis with examples if possible Especially if it’s a huge amount of data, massive data image, you are not gonna provide every single data point, but you can provide trend analysis And this is to say Choropleth maps use colour to define or prioritise information that should be translated into text Mainly because if an unsighted user is working in a peer group they need to be able to say the blue colours on this map or the red colours so everyone can relate to the same thing Here is an example I’m sure many of you have seen a graphic like this in the previous couple of weeks So this is titled State of a Nation It’s a Choropleth map of the election results from the US presidential elections in November This choropleth expounded around on things quite a lot The U.S. is very good for creating these types of maps because of the state system So the description would be, Biden won And that’s all you need to say Not really Description would be a map coloured coded to reflect the winner party in each state with blue represented democrats and red representing republicans A horizontal bar chart, positioned above the map, illustrates the overall result with democratic nominee, Joe Biden, receiving 306 votes, and incumbent Republican president, Donald Trump, receiving 232 votes The results for each are presented in the following table The states, listed alphabetically, again you’re creating that structure allowing the user to visualise the content and information they are gonna receive, are presented in column 1, the winning party is listed in column 2, and the number of Electoral College votes for each state are presented in column 3 Split Electoral College votes in Maine and Nebraska are represented by the two rows for the respective votes That’s included because the user might get to the point in the table where Maine being included twice, and they’ll be like what’s happening here? But it’s just because of the way they do Electoral College votes So you end up with the data as follows And then you just have a table Three columns, then there will be 52 rows cause District of Columbia is included in there and they got a header row as well And then you have all the data, and the user can move through the table and discover who won in each state, which party and how many Electoral votes were voted for I’ve gone into maps, describing maps The following elements need to be referenced when describing a political map We’ll have this up in a moment The main subject of the map must be included in the brief description in the alt text The subject of a political map will be a global continental country or regional effect So we’re talking about the date When was that map created? Cause that’s always quite important if that’s provided in the information you have Emphasis and context The map is being selected for a reason and the context and usage of the map should be addressed in the description Include things like places of interest, things like main cities

Usually main cities are included especially capitals, seas and rivers, mountain ranges, anything like that It’s not necessary to list every place of interest but it does provide a certain amount of detail that creates emersion and just interest in the description I like to include edge boundaries It’s not applicable to every map but sometimes when you have a map and it’s focused on a particular area, I will show you an example next actually The map here is a map of Europe It has around the edges Russia is on the right North Africa is at the bottom So I like to create edge boundaries just to increase the visualisation aspect, so you know the North Atlantic Ocean is on the left to the East (indistinct) to the west and Russia is to the east and things like that So you’re creating those boundaries and they come in quite useful also with things like street maps as well So if you’ve got boundaries of certain streets, North, South, East and West they can create a framework for the user Describing maps continued In most cases colour is used to differentiate between countries It’s not generally necessary to include those colour details If France is brown colour and United Kingdom is orange You don’t need to know that information, we’ve just done that to make a little bit more clearer But that’s not necessary for the description Include things like scale ratio and scale measurements Talk about spatial description We are gonna go through the sector approach to describing maps because it can be quite useful So for this part of Europe you can divide into quadrant because that’s a lot of information in a map If you divide into quadrants you can deal with each quadrant separately So you’re creating sections and talking through them For instance, this one is a political map of Europe in 2012 The map shows each of the countries of Europe The 4 quadrants cover the following countries So then you go through North West, North East, South East and South West listing out the countries in each sector Then you saying that each capital city is been included major rivers are included as well And then you talk about the edge boundaries at the end there to really frame that description And back to the sector Theory or method of describing this type of complex images Basically you are using quadrants or a clock face or a compass So you’re describe north, east, west, south of that image or you’re dividing that image into 4 parts or could be more than that if necessary, or you are dividing a particular detail, illustrations and diagrams, you can divide them into a clock face So you get 12 sectors So you can say between 12 and 1 this is what’s happening Between 1 and 2 this is what’s happening That can be useful for very complex and detailed paintings in our works for instance But it can be applied to maps and all kinds of things So on the left we have a 2-page illustration from a (indistinct) star dust There’s a huge amount happening in this kind of market place So you can use the clock face approach and divide into 12 sections So you can really get the richness and detail that the author is trying to convey On the right-hand side we have the Flemish marketplace I think by (indistinct) Again it has a huge amount of detail Lots of great things happening in this city You can go on this and arts gallery and you can get into detail and see all the faces It’s so detailed the way they replicated it But you can create a quadrant effect there so you can divide into sections, talk about the left section or the right and go through and break it down in an easier way This slide is entitled the UK divided It’s a map of the UK regarding the geology of the UK You can divide this because of the nature of the UK You can divide it into sectors but one on top of each other So you are working down through north to south So you end up with a geological map of the UK and Ireland The map plots the geological makeup of the United Kingdom and Ireland And uses colour to define the geological properties of the regions A colour coded key for sedimentary rocks reads as follows and then you can provide the colour key And then you can divide the map into 6 distinct regions So you go from Scotland to Northern England,

to Midlands, Southern England, Wales, and the Northern Ireland at the bottom And then you can have a note at the end if you want saying the rock formations in Scotland match Northern Ireland And the rock formations in Wales match those in Southern Ireland Just an interesting thing, you can see where (indistinct) has the same geology And just to recap, think about structure, organise as lists Lists are really your friend Especially with things like bar charts, pie charts You think about where that bar or pie chart came from, it was originally a spread sheet document and was created from that and translated into a pie chart So basically what you are doing is translating it back Present things as tables It really helps independence for the user If you provides the data then they can explore themselves Make sure to subdivided into sections for things like infographics It just breaks it down about huge amounts of data that’s been provided in infographics, you can break it down into an easily consumable way And as Valerie says talk about user path way and go from an overview to the details Moving from the general to the specific Visualisation, always start with the over view or that focus point and move through the description using a pathway through all the details or using the structure as a start and providing that story telling aspect with the details and providing the immersive aspect and end up with a fantastic final description Quality control quickly At the end always think about stepping into the user’s shoes Are you provide all the information that they would want? Precisely if you were to look at this diagram you would get all the information quite quickly (indistinct) the information I need Are you doing that, are you repeating that in the description to enable a visually impaired user to get exactly that same information? And try to re‑create the image from the description It’s always such a useful exercise If you’re not sure read your description back and see if you can re‑create the image from the description without looking at the image And see how well you’ve done It’s always a useful exercise So I hope that’s helped unravel the complexity of complex images I know it’s by nature a complex business, but we hope that that gives you an over view and a starting point to think about describing complex images and creating that level playing field for all users As we say, we realise it’s quite a complex field So as we said at the start we are doing a trilogy of these talk So we’re very open to your thoughts about what we would cover in the next session So I think at this point we are gonna asking Dave to take control, here we go, a poll I’ll move from my screen Valerie, do you want to take over? – [Valerie] Great presentation, Huw We are looking to get your feedback and your input into what we cover in our next webinar on alt text description And we are interested in learning about what kinds of images you would like to be covered in a future webinar And so we have some ideas for what we haven’t covered yet that we were thinking of including So those are available choices and I’ll list them out Artwork, graphic novels and cartoons, anatomy and physiology text books or images, tests and how to describe an image without giving away the answer if a student is gonna be tested on that or other And please respond in the Q&A if you have other examples that you would like to see us cover next time – [Richard] Thank you so much Valerie and Huw We’ve got time for some questions while people are voting A good mix of questions here And as Alex said already, that was an awesome presentation Thank you both of you So quite a practical question from Cassie here which is maybe related to, you both talked about bar graphs for example, bar charts If you are listing values in an image, do you repeat the units or are you specific about the units of measurement after each number like kilometres and so on? How do you handle that so that it’s both clear and yet not repetitive? – [Huw] That’s a good question

Generally if I’m using tables I would include the measurement in the table header and then just the number in the table, in the row as it were So you are not creating that, especially if there’s a lot of data points it does creates a repetition which is a bit boring If there’s only a few examples data points and you are not using a table I would just list them out and after that I would put kilometres or whatever it might be I generally also write them out as well just in case screen readers stumble over KM or whatever it might be I tend to write out the word just to make sure – [Valerie] I agree 100% I would make sure, especially if you have metres per second, if you have a slash or forward slash, or something I would type it out in words Metres per second – [Richard] Thank you for that Sue asked a question, it’s really a clarification Because you’ve talked about alt text, Valerie, you were talking about what’s available within JAWS and you kind of hear that all in one go Could you help Sue with some clarification around what’s better or appropriate to include in the alt text field, verses in an extended description which may is navigable And how practically do you include that extended description? I guess that second part of the question relates to exactly what is the document and what format is it in and how will it be read? So that distinction between alt text and the extended description So Valerie, are you able to start with that? – [Valerie] I am and it’s a really hard question It depends on a lot of factors It depends on the type of material you are describing and your audience So, working at CIDI, a lot of the decisions that we made are based on what the student prefers So the student through their disability service provider is asking us for either brief description or long description, and we have different products that we provide based on those preferences So, a lot of students prefer the brief description But in other cases the student will say, no, I want more comprehensive detail and all images described in every data point So that’s how we are making our decisions I would say as a rule, you don’t wanna go beyond 3 or 4 sentences of alt text description on any image because depending on the person’s settings in JAWS or NVDA or whatever, technology they are using, it could get cut off You have the option that maybe they have chosen brief description They don’t have their settings on verbose description So you run the risk of someone not hearing all of the hard work that you are doing So I would say 3 or 4 sentences, 5 maximum and then you wanna think about moving that information out of the alt text into the body of the document – [Richard] Got it, and Huw some of the examples you gave where you had nested lists and tables and so on, so clearly that’s not alt text You don’t have the benefit of knowing the preferences of the end user’s because you are doing this work, as I understand it, with publishers So what’s you’re take on this and how are you including those? – [Huw] As Valerie says, it’s a short brief description No more than 3 or 4 sentences Generally works out with things like graphs It’s quite a brief alt text It could be 1 or 2 sentences With a photograph you can generally get most information that you need into the alt text, but that’s fine it could be longer But as you say, I work mainly with publishers So generally working with the EPUB format and I think this is probably a nice segue into a session we’ve got coming up in terms of actually implementing alt text and image descriptions, those long image descriptions into EPUB Cause EPUB really handles it quite well I should say you can’t use lists and things like that in alt text It won’t work

You’ve got to point at list and those kind of things and nested tables But it does handle that really well in EPUB So the majority of publishers I work with are using tables and lists as long descriptions within EPUB So it keeps things simpler for me I’m not dealing directly with the end user But the EPUB format provides an excellent way of delineating between alt text and providing the opportunity to have long descriptions which have all that information and have the way of conveying information through lists and through tables, which can be independently examined by the reader So EPUB, I think you have a session coming up quite soon and that will be fascinating to see because it’s always developing with long descriptions and the way that they are handled – [Richard] Great, well I’ll mention that session in just a moment But maybe you can reassure Orikas who has this question, “You are working with publishers, “so they are not considering these images descriptions “and extended descriptions “as a corruption of their materials “from the publishers point of view “Is there a question from the publishers “on that side of things?” – [Huw] A corruption? My work is a corrupting influence, I like that – [Valerie] I like that a lot – [Huw] It’s a good point It’s a really good point, because the author is conveying information You have to be impartial You’re not saying I don’t put descriptions (indistinct) You are always being impartial, you’re looking at the context of the work around it You are providing a very neutral description This is information being conveyed in an image You are not making an interpretation of the image If there are any instances where I’m not sure what the author is trying to portray or there’s an error In some of the books I’ve worked on, they’ve got hundreds and hundreds of graphs, so sometimes you come across an error in the data You work with the publisher, you go back and say I found this or I’m not sure about this I even work with the authors themselves because the publisher is not sure So we create that kind of path way between me and the author So I always want to make sure the author’s story is being told in a correct way So wouldn’t kind of go in and say this is what I think So you have to be very impartial At the end of the day it’s the author’s work and the publisher is paying me to describe it Hopefully I’m not corrupting it, but I do like the idea of it – [Richard] Thank you for that Huw And thank you also for mentioning that we will be having a webinar we’ll be announcing shortly, which will cover the technical detail of how extended descriptions can be included Including some latest published best practise examples And maybe indeed we can include some of the images and descriptions that we featured in this webinar as examples to show working within EPUB there So that will be exciting to see The results of the poll are out And we see that for example (indistinct) asked a question but didn’t have a chance to get it answered The most commonly requested one is for images that relate to tests and testing scenarios without giving away the answer So we will be sure to feature that and the others that are covered in the poll results So thank you for those, we’ll be analysing those We’re coming to the end of this session So thank you to Valerie and Huw for sharing great information and insights Next week we will be featuring Kirsi from Finland as our host, for the session, Do more with WordToEPUB Point and shoot, and you can start making accessible EPUBs with Microsoft Word and WordToEPUB by DAISY But you can do a lot more than simply click and follow through This session will demonstrate the very latest version of the tool It will introduce you to the newest features and give a taste of what is coming soon You can register at, where you can also sign up to the webinar announcement mailing list If you would like to suggest a subject for a future webinar, or if you are considering presenting a webinar yourself, then please email us at [email protected] Well, I hope you will join us again next week In the meantime, thank you for your time and have a wonderful rest of your day, goodbye