Wednesday, May 27, 2020

A toolkit for efficient educational video production

This blog post describes some prototype software, which, teamed up with the right hardware, can make producing decent quality videos at your desk slightly easier. If you want to jump forward to see the demo rather than read about it, scroll down to the section with heading "A Demonstration".

Introduction

A couple of years ago, my colleagues Shazia Ahmed, Ruth Douglas, Sue Milne and I did a presentation "Creating Effective Educational Videos: a toolkit for a quick and low cost approach" at our University's Learning & Teaching conference. Most of our presentation was about maths support videos, created with the help of a piece of software I created for our Blended and On-line Learning Projects, but the last bit was about a more advanced version of the software, which combined with some other software and hardware makes a powerful toolkit for creating videos or live on-line presentations. In the current situation, with the pivot2online, I think it might be worth revisiting that bit of the presentation, but as a blog post for a wider audience.

The version of the software my colleagues used for maths support videos was designed specifically for Khan Academy style videos, where the teacher writes on a virtual blackboard. My software has multipage virtual blackboards, graph paper overlays, and the ability to insert a web page or application and draw over that.

The more advanced version of the software, UPresentTo, was partly developed with the intention of starting Vlogging, and trying to make that as efficient as possible. I had found when making short training and information videos, that the time taken editing the video was often substantially greater than the time taken to record it. Editing together different parts of the video takes time, and, for me, editing out pauses as I was trying to think what to say wastes a lot of time. To address this my software is designed to help minimise editing time by providing a tele-prompter window for a pre-written script, and the ability record multiple pre-planned shots in a single take.

I have also realised, from studying with MOOCs, that videos which consist of just a voice and static PowerPoint slides are rather unengaging - a small amount of movement on the screen, such as hand annotations and the PowerPoint slides make the video much better, but having different shots, for example showing the presenter's face when there is no other useful visual information, makes for a much better video. So, one aspect of my software is making sure it is easy to put together different shots, rather than just virtual blackboards and slides.

Of course, my software is not a complete solution - it is part of a toolkit, consisting of both hardware and software, that can be used to create a certain style of video.

Hardware

Hardware for the video toolkit

The (mostly) affordable hardware consists of:
  • A Novation Dicer MIDI controller, which is used to move through slides and shots, and to increment the tele-prompt.
  • A Wacom Cintiq or standard graphics tablet.
  • A decent quality webcam.
  • A decent microphone and stand - the difference between a £3 and £30 microphone is dramatic, so don't skimp on this one, but at the same time, there's no need for something very expensive.
  • A standard Windows 10 computer. (My software is Windows only unfortunately.)
  • An LED garden floodlight - proper photographic lights are fine, but a white painted ceiling and a daylight floodlight is just as good for WebCam video.

Software

Software, UPresentTo and OBS

Just one other piece of software is required alongside UPresentTo when recording a video. OBS studio is used to actually record the video. UPresentTo sends keystroke messages to it to switch between recording the desktop (for slides, virtual black/white boards, applications and web sites), the WebCam, and any pre-recorded bits of video that have been added as scenes to OBS. In this screenshot you can see the main UPresentTo window on the left, the tele-prompter window at the top and OBS on the right. The desktop recording part of UPresentTo is being displayed on the Wacom Cintiq tablet.


Creating the script and storyboarding video is currently done in a separate web application, which was originally intended to become a web slides tool with markdown used to write the slides. By adding a few extra tags I was able to make it suitable for creating the input files for UPresentTo. Eventually I intend to get all this functionality into a single package, but this was a quick way to get going. In this screenshot you can also see how equations can be included on slides using MathJax.

A Demonstration

Here are two short videos, recorded at the same time which show the toolkit in use. In the first video, recorded on a video camera next to my desk you can see me making the video, reading a teleprompter at the top of my main computer monitor and controlling it from the MIDI controller next to the tablet. The second video is the unedited recording from OBS studio, showing how UPresentTo switches between shots, and makes a video which requires very little editing before it can be published.




Final thoughts

This is unfinished software - the University project was cancelled long ago because it does not fit within the remit of my new department. However, it isn't far from being a usable setup for creating decent quality videos reasonably quickly. By using the virtual camera plug-in with OBS, it would also be possible to use this software with Zoom, and some minor modifications could make it more appropriate for tutorials. (By making the navigation between shots less linear.)

With enough demand this project could be resurrected as a University project, but for now it's just my private video making toolkit.


Wednesday, May 06, 2020

Protecting the Hive

A couple of weeks ago I saw this tweet from Belle & Sebastian - an invite to collaborate with them on a new song. So, naturally I started trying to make my own version.
The first thing to do was of course download the different parts of their recording from SoundCloud, and stick them into an appropriate piece of software. In my case, the software is Cubase Elements, but there are plenty of free alternatives that would do the job - I also used Audacity to make a simple mix that I could use to work out the chords, and the band suggest using Garage Band (though obviously that's only appropriate for Mac users.)

So far my efforts have been a bit frustrating. Belle and Sebastian's version includes a finished vocal track, which also has a flute solo on it, two rhythm guitar tracks, a very basic Wurlitzer piano track and a drum machine. (I don't think the full band have been involved yet.) For me the first task, after working out the chords and structure, was to improve the drums, by cheating. I stuck the song into band in a box, found a suitable style then exported a MIDI version. In Cubase I trimmed out a bit, and teamed it up drum sounds from my Planet Phatt. Next I started adding in some of my own playing, and this is where things have begun to go a bit wrong - my guitar and keyboard playing, never that great, has got rather rusty with lack of practice of actually creating recordings.

However, I have been having fun and will continue to try and create a finished version of the track with at least some of my own keyboards and/or guitar along with my cheating drums, and either real or cheating bass. And to save others a little bit of time, here are the chords that I'm using (which may be slightly simpler harmonically than the ones that Belle & Sebastian used)


Verse 1:    G / Am / | Am / G / | G / Am / | Am / G / | 
            G / Am / | Am / G / | G / Am / | Am / G / | 
            Am / G / |
            
Verse 2 :   G / Am / | Am / G / | G / Am / | Am / G / | 
            G / Am / | Am / G / | G / Am / | Am / G / | 
            Am / G / |  
            
Chorus      C / G / | D / G / | C / G / | D / G / |
            C / G / | D / G / | A       | D / C / |
            C / / / | C  / / / |
            
Verse 3 :   G / Am / | Am / G / | G / Am / | Am / G / | 
            G / Am / | Am / G / | G / Am / | Am / G / | 
            
Chorus      C / G / | D / G / | C / G / | D / G / |
            C / G / | D / G / | A       | D / C / |
            C / / / | C  / / / |
            
Flute Solo: G / Am / | Am / G / | G / Am / | Am / G / | 
            G / Am / | Am / G / | G / Am / | Am / G / | 

Verse 4 :   G / Am / | Am / G / | G / Am / | Am / G / | 
            G / Am / | Am / G / | G / Am / | Am / G / | 

Chorus      C / G / | D / G / | C / G / | D / G / |
            C / G / | D / G / | A       | D / C / |
            C / / / | 
            
Verse 5 :   G / Am / | Am / G / | G / Am / | Am / G / | 
            G / Am / | Am / G / | G / Am / | Am / G / | 
            
Chorus      C / G / | D / G / | C / G / | D / G / |
            C / G / | D / G / | A       | D / C / |
            C / / / | C  / / / |
Once I'm happy with my version, or senile enough not to care, which ever comes first, I'll post it to Soundcloud and tweet as instructed by the band :

Saturday, March 14, 2020

Planning to move teaching online (suddenly)

So it seems that we will be moving to distance learning, not the baby steps that we've been taking (in many iterations) over the last 23 years, but suddenly, all at once. A whole moderately large, very traditional University, with no time to plan, switching to distance learning... Exciting times!

Being a University, nothing can happen without a committee being formed, so naturally there's a committee, but I'm not involved in that. 

However, I have been working with learning technology for around 30 years and the vast majority of the almost £500k grant income I've had in my career has been related to on-line and distance learning. I've also got a degree from the Open University, so I've got a bit of relevant experience from the student perspective. Oh, and I also helped facilitate a cMOOC once, and have done a lot of distance collaboration as a member of IMS working groups. So, although I'm not an expert in suddenly switching a university to distance learning, and not really an expert on distance learning, maybe my background means that I can provide some useful insights and hints.  

It seems to me that there are probably going to be two distinct phases to this on-line experience - the first, starting on Monday for the University of Glasgow, will be largely unprepared. The second will start with the new academic year in September, by which time there will be quite a bit of preparation done. I hope that preparation will include providing teachers with some training and technology to make it easier to teach from home.

One of the things I have been doing this year is working on a proposal for a digital skills course for academic and support staff. A feature of this course is going to be that it focuses on using free and open source software to do things well enough, rather than using expensive professional software to do things to a media professional standard.  Events in the world have rather overtaken me now, and this up-skilling is probably needed right now rather than in six months time.What I'm thinking now, is that I should just start writing this course, making bits available as I complete them.

What I don't know is what's to do first. I was expecting my course to go from basic image creation and editing, through audio recording and editing and ending up with video editing, with accessibility and web publishing being themes running through it. However, while that order makes logical sense as a course, it probably isn't addressing immediate needs.

So, teaching staff, University of Glasgow or elsewhere, is there any tutorial that you would like provided, to help you do your teaching online next week? Please let me know in the comments. The sorts of things I could cover include anything that will be in my digital skills course, or using specific features of Moodle.

Sunday, February 23, 2020

A sad list of cancelled and never started learning technology projects...


Recently I attended the first day of our annual learning teaching conference, which is spread over three widely separated days this year. Most of the presentations were really good, but it was the discussions that followed them that got me thinking about some of the projects which I probably would have been working on if I was still in the Learning and Teaching Centre. 

I've listed a few, that I'd be happy to resurrect if i got the opportunity, here. Some of these were cancelled when I was moved to IT Services, and others were proposals that were turned down. 

If you'd like to see any of these projects are progressing, please add comments to the blog.

ClosedCaptionHelper

One of the themes of the day was accessibility, and with distance learning being continuing priority for the University, Closed Captions for videos came up. One project relevant to this is my ClosedCaptionHelper, a Windows application that helps the user type up closed caption files with the correct time codes in them. This project wasn't actually cancelled - it just ceased to be a priority because money was allocated to get commercial captioning done for the projects I was supporting. I don't know if there is still a need - commercial captioning services are remarkably cheap - however, it could be resurrected if there is a demand.

Future peer review

Peer review, were students review and comment on each other's work, has been shown to have considerable learning benefits. (Nicol et al. 2013) Currently there are two peer-review packages in use at the University, Moodle workshop and Aropä, but both have limitations. Moodle workshop can be difficult to configure for each exercise, and has no support for group work, while Aropä does not integrate well with Moodle. A fundamental problem designing good peer review software is making sure that the requirements of different teachers are met, without making the user interface too complicated. The idea behind my Future Peer-Review project would be to create a new LTI tool with greater inherent flexibility, but with the user interface made simple for teachers by allowing them to choose from a limited set of scenarios created by learning technologists. The learning technologists would have access to the more complicated setup options, which they would use to create the required scenarios. Because this approach would eliminate some of the need to compromise between complexity and usability, the same tool would be able to support both individual work and group work modes. The tool itself would be an LTI application, launched from Moodle or another VLE.

Content Authoring Tools including Efficient Quiz Authoring 

The current process for creating e-learning content can be rather inefficient, and one area that I am interested in looking at is making use of simple syntax similar to markdown for creating content that can be delivered using different mediums. The most obvious area for efficiency improvements is writing e-assessment questions. Currently this is typically done using web forms, which are quite clunky. A more efficient method is to have a text format that can be automatically read into the quiz questions. A tool that use this approach was the CASTLE toolkit from the University of Leicester. My tool would go a bit beyond what was supported by CASTLE, as I would include standard markdown for text-formatting, and also support mathematical equations by including MathJAX.

The same approach could be used for authoring other types of e-learning content, for example books with interactive sections, or exercises were the student would select various actions, and see different information based on what they had selected. (An example of this type of exercise would be a virtual diagnosis exercise for medical or veterinary students.)

This would also be the starting point for writing IMS Common Cartridge content with extended functionality, which I'll describe in slightly more detail later in this blog post.

Desktop Presentaton software - UPresentTo

With an increasing emphasis on making use of blended learning for face-to-face courses, and also an increase distance learning, making short educational videos is becoming more common. As a distance student, I have found that the style of videos makes a huge difference to how easy they are to watch. I find videos that are basically just a voice along with static PowerPoint slides extremely unengaging, whilst just adding a face in the corner makes them feel a bit better. A second issue with videos is that editing together different parts can take quite a long time, so ideally video was created with as little editing required as possible. I spent a bit of time thinking about how to achieve this, and came up with a piece of software which combines web slides, embedded applications and web sites (with the ability to pause them and annotate the screen), optional telly prompting and a virtual whiteboard. It is also used to control OBSStudio, which allows the inclusion of web cam shots and pre-recorded video into the storyboard. Although it is still a bit clunky, when combined with the right hardware, this software does make creating okay-ish educational videos a lot easier.

NGDLE Prototyping

The Next Generation Digital Learning Environment (NGDLE) is a more distributed version of the VLE, making use of protocols such as LTI to connect separate components. Although called the next generation, it is to some extent already with us as we are increasingly using LTI tools to extend the functionality of Moodle. For example at the University our lecture capture software and library booklists are both linked to Moodle using LTI. Sarah and I wrote a paper on this, before the term NGDLE had become the standard name for the concept.

One of the things I would like to do is develop a process for creating the components of this distributed learning environment, using a mixture of product family engineering and generative programming for the fairly boilerplate aspects of the applications, and exploring potential extensions to LTI to make the combination of small, relatively simple, tools work as well as the integrated VLE is that we are used to. (Obviously these extensions would then be fed back to IMS for consideration for inclusion in a future version of LTI.)

To some extent this project would also combine some of the others here, for example the future of peer review system would be one of these components. The product family engineering and generative programming approach has also been used already for YACRS and a few other projects at the University.

xCC 

The IMS Common Cartridge is a way of creating learning material that can be deployed in different VLEs, however it has very limited capacity for including interactive content other than quizzes, with a fairly limited set of standard question types. (And some VLEs, for example Moodle, ignore the quizzes.) I have done some experiments with adding more interactive content to Common Cartridges, including creating specialist types of quiz questions for academic writing exercises,  and adding an H5P activity into a common cartridge. The limitation for both of these is that Common Cartridge doesn't have any support for custom server side code, so there is no persistence, i.e. when the student closes their browser there is no longer any record of what they have done. QTI 2.0, which I was involved in the development of, has a simple set of instructions that can be used to process and store quiz responses on the server, without any risk of being able to interfere with other processes. I'd like to design an extension to common cartridge with support for simple server side rules like this, giving it sufficient power to record students responses, and at marks to a gradebook. (Although QTI is part of the inspiration for this, I would avoid its XML syntax which is neither user-friendly, nor particularly easy for developers.)

 

Thursday, January 23, 2020

My thoughts on an IT Strategy for Learning & Teaching

This week we had a meeting of IT staff from around the University, with our director of IT telling us about the University's new IT strategy. There was a lot of good things there - major new investment in infrastructure to reverse years of underspending on IT, and a commitment from senior management that this investment will be ongoing, rather than a temporary fix. However, one thing that was missing was any mention of learning & teaching. Of course, many of the infrastructure improvements will be of benefit to learning & teaching, but it would be nice to have a strategy for the parts of our service that specifically support learning & teaching. The director did say that the strategy isn't finished yet, so maybe learning & teaching support is still to be covered. However, in the absence of a current strategy, here is my thoughts/wish list...

Our current support for learning and teaching seems fragmented and often contradictory. In addition, it leaves gaps, particularly when it comes to supporting the scholarship of learning and teaching. The fragmentation means that we have different teams supporting learning and teaching who technically have different roles, but whose roles seem to overlap. We need someone with a proper overview of all learning & teaching support to eliminate this overlap. The current structure also means that the expertise in learning and teaching that there is within IT services is buried under too many layers of management to have influence, so this role needs filled by someone who either shares that expertise, or is at least knowledgeable enough to recognize and use it. 

Supporting scholarship of learning and teaching, (teachers in higher education developing expertise in the pedagogy relating to their field and  contributing to educational literature and practice) is a University priority that we should share. Part of that scholarship is engaging with new or innovative use of technology in education, so we need a coherent structure to provide both technical and logistical support for that, with staff available to meet with teachers to discuss their needs, and then collaborate closely with them to develop solutions. Part of this means that we need to recognise that learning technology and business systems are fundamentally different and need managed in different ways - learning technology is an early adopter of new ideas, and is constantly evolving, whereas business systems are fundamentally conservative and aims for stability.

The next generation digital learning environment (NGDLE) is already present in higher education, even though it is still not really clearly defined. Broadly speaking this is a digital learning environment made out of small specialised tools linked together using LTI or similar technology.
Like every other University we are already moving to the NGDLE  through increasing use of LTI tool, both on and off campus. We need to embrace this properly, and decide whether we want to be leaders or followers. We also need to have a system in place to cope with the fact that the future DLE is going to be a network of different tools, with different licences, contracts etc. What we must not do, but are currently at risk of, is become the big name university that every dodgy supplier with more skills in sales than in IT targets for their customer list.

One thing that I have felt has been missing from IT services for the last few years has been a senior member of staff with a background of working in the teaching side of the University. In the past we had deputy directors of IT services who had started off elsewhere in the university, and even in one case, still involved in teaching. More recently our senior staff have all been IT professionals, which possibly is why our support for learning & teaching has lost focus. I'm not sure we need someone as senior as a deputy director, but we do need someone with deep knowledge of learning technology and related pedagogy in a more senior role. This person could then address the fragmentation of support, provide a suitable communication channel with teaching staff, and take charge of the NGDLE related changes.

At the moment there seem to be two rather extreme models of management in IT Services - one very much a hands on member of the team in the hardware sections, and in the software side, a very hands off hierarchical structure with managers also doing a lot of administration. I don't think either is ideal, but for learning technology I do think a degree of direct involvement is essential to remain on top of the subject - maybe the solution is to resurrect the idea of an Academic Related post, and get someone who's got some academic credibility into a senior IT role, with a professional administrator to support them.

At one time this was the type role I aspired to, however in recent years, the way my career has gone has not given me the opportunity to develop the skills needed for such a senior role - my management development ended when JISC stopped funding development projects. That brings me round to one final piece of strategy - we need to invest in our own people, and develop the future leaders ourselves. Otherwise, we risk ending up in a situation I've seen develop elsewhere, buried under an ever increasing load of bullshitters...

Wednesday, September 11, 2019

Office noise levels part 2

As a comparison to the office nose graphs I posted last week, at the weekend I set up my noise meter in my study at home and created the same graphs for that location. Like the previous graph, this shows a time when I was largely sitting working in a private room. As it was another warm day, the window was open, but there was very little noise from outside - no one nearby was cutting hedges or anything like that. The underlying ambient noise level (green lines) is similar at about 42dB, but the average sound levels and noise variation at home are dramatically lower.
Ideally it would be possible to get the noise generated by my own activities isolated, as that would give a clearer indication of the underlying disturbance noise. In the home standard deviation line (red) in the middle there is a two hour period when I was out, which is clearly visible, suggesting that most of the other sound variation came from my activities. In the work standard deviation line, the last two hours are also showing time when my office was empty, but clearly not as disturbance free.


We're due to be moved into a large open-plan office soon, and when that happens I'll try and run the tests again.

Monday, August 26, 2019

Office noise levels

A couple of months ago I sat a decibel meter on my desk for a few days, and recorded the noise levels in my office. Although I feel my office is very noisy, the figures I got didn't seem too bad compared to typical levels I found on-line. But one thing I did spot was that the most disruptive noises - people talking in the corridor - were also not particularly loud.

At the weekend I came across a paper, "Defining the Acoustic Environment of (semi-)open Plan Offices Acoustic Measurements leading to Activity Based Design for retrofit Buildings", which included average noise levels for various types of office. The authors of that paper found that the average noise levels in the modern (open-plan) offices they looked at were fairly consistent at about 51dB, but there was variation in the standard deviation from ~3 for an office filled with programmers to ~5 in an office with mixed tasks.

So, I went back to my records, picking one day look at, and plotted average sound level, standard deviation and lowest 2 second average in 5 minute segments from 7am to 6pm.

The minimum sound level line gives an impression of the background noise - the higher levels for a while in the afternoon are due to a graduation, with the new graduates and their families congregating in the quad outside my window, so this is an unusual event that can be ignored.

I suspect that the standard deviation line at the bottom (and the large variation in noise levels through the day) show why I feel my office is too loud, even though actual average sound levels are OK, the number of disruptive noise events is high, and this is what makes if feel like a noisy and unpleasant environment to work in.

It would be interesting to get similar records from other offices along with information on their occupants impression of how noisy they are to compare with this - maybe a nice wee research project for someone. I've not had much success in finding research on the impact of noise on knowledge work, other than the well known work DeMarco and Lister covered in PeopleWare, but that research should make this a major concern to managers.