Color Management

Have you ever had to deal with faculty who tell you that current projectors do not match the resolution or color quality of 35mm slides? Do they still demand slide projectors in their classrooms, 8 years after Kodak stopped making them? Do they insist that ten years of technological advances still have not reached the quality of the classic slide projector? I hear this often from faculty I work with, and for the most part, they are correct.

Over the past few years higher resolution projectors at more affordable costs, have helped with delivering a better resolution. However, most institutions can not afford 15k projectors in every classroom. In talking with our faculty, the resolution is something that they have learned to live with. Certainly the convenience of digital images, along with the brightness of modern projectors have helped convince them in moving to the digital realm.

In my experience color rendering is a much bigger issue than resolution. In order to understand the issue better I recently took a walk around an academic building with a faculty member and looked at the same image on multiple screens and monitors. In fact, in every single room the image looked different. Suddenly I understood how difficult it is to teach fine details, such as how a specific artist puts golden glows around their main subjects, if that golden glow does not show up on every projector.

I dove into this problem headfirst determined to understand what was happening and learned a lot. My first stop was at the Imaging Center we have on campus to talk with our resident photo expert. He introduced me to the topic of Color Management. The basic concept of color management is simple, all colors have very specific definitions, and all computers agree on this definition. However, for various reasons (age, use, quality) the displays we connect our computers to don't all show these colors the same way. So, what we have to do is determine how "off" our displays are and have our computers adjust their output based on that number.

There are some fairly easy tools on the market to do this. We used a product call the "GretagMacbeth". The product contains a receiver that we set on a tripod and point at the screen. Then your computer, projecting on the screen, runs software that displays a series of known colors. The beamer determines how "off" these colors are, and develops a .icc profile for that display. This file tells your computer how it needs to compensate for the projector in color, brightness and contrast.

This ten minute process can dramatically correct color problems in your classrooms or presentation spaces. Our faculty loved the changes and are now more ready than ever to change over to digital. However, there are some problems and considerations. First, you need to remember that this process does not make an image look better, it makes the image look like it is supposed to look. Second, it is ideal to design systems so that the computer recognizes all attached displays. If you have rooms in which your computers feed into a matrix, switcher or splitter then it is likely that your computer does not know that it is plugged into multiple displays and you will not be able to set different profiles. Therefore, managing the color on the projector will make the local monitor colors wrong. In many cases this may be acceptable, as the lecturer or presenter knows how the image is supposed to look, and they are more concerned with the audience's experience. However, in a scenario where both need to be correct, or perhaps a room with multiple monitors of a different model, you will need to provide the ability for the computer to recognize all the displays. Due to the fact that you will be using your own .icc profile, full EDID information is not needed. All that is needed is recognition that the monitors are connected.

On a mac, these settings are extremely easy to use and very intuitive. When the mac sees a projector or other display, you can open the display preferences and get a separate set of preferences for the second display. Under the color tab of those preferences you can easily set the new profile. Under the settings for your second display, you can set a different profile. On the windows side, I have not found an easy solution for this. With Windows XP, you need to download an application, "Microsoft Color Control Panel Applet for Windows XP". Even with this application installed I was not successful in selecting separate profiles for the two separate displays. If anyone has had any luck with this application, please leave a comment and help the rest of us out.

A third issue to keep in mind is that displays experience color shifting over time. As lamps get older or heat alters the panel, there are slight changes in the color. For your pickiest user, you may need to do a re-calibration once a semester to ensure you have the most accurate colors possible. For some of the less picky faculty you may be able to use a single profile for one model of projector. The .icc profiles are cross platform, so this allow you to have a single profile that can be shared with multiple users, on multiple platforms, for a given space.

If you need a real upper, on a bad day, profile one of your classrooms then call in some of your art history faculty. You will put a smile on their face for sure. Going through this simple process is also a great way to continue to get the most out of your investment. It is an easy, yet very obvious way, to take your service to the next level.

Video Streaming and Time Lapse

I was recently tasked with providing a live web feed of a construction process on campus. The customer also wanted to have a time lapse video of the project after it was completed. My first reaction was no problem, people do this all the time. After asking several colleagues about what they do, I realized there is not much of a standard out there for this type of project.

Most of the people I spoke with recommended Axis Network cameras. I researched several of these but ran into one major stumbling block. The built-in web servers would only serve 20 people at one time. While this may be suitable for security purposes, it seemed that this feed, at times, would get more than 20 viewers. I could imagine that many people are like me, and would open the web page in a tab, and then just leave it running, eating up one of our connections. My customer, and I, thought it was unacceptable to have people turned away. Next we looked at using a free service through the web, and ended up choosing Ustream. We connected a consumer level mini-DV camera to an out of service computer and started streaming. You can see the results here:

Bates College Garcelon Field Construction Feed

For several weeks this has provided us a reliable streaming service. As an added bonus to our web people, it also provides metrics. Our customer has been very happy with the result, and we have heard from people who view the stream and they love it.

The time lapse portion of the project also required some research. Our intial thinking is that we would store all the video, stick it into Final Cut Pro, and speed it up. We quickly learned that even though this is a short construction project (5 months) saving all that video would take too much space. We also did not think we had the processing power to create the final video.

We did some research on various cameras, but learned that most of them that did time lapse saved it as a QT video. We did not want to work with video that a Point and Shoot camera had compressed. Finally, we learned that Canon cameras come with a piece of software that manages the camera and activates the shutter on a schedule defined by the user. This software is only available on the higher end Canons. We purchased a Rebel XS, on sale for $500. If you go this route, you will also want to purchase a power supply for the camera. I forgot this step and had to scramble to get one when I realized the camera was not powered by USB from the computer. We have done a few tests of the time lapse, again with great results. Look back here in October and I will post a link to the final time lapse, along with a final critique of how this project went.

Rooms That Talk

If you would like to see this whole presentation, in better detail, follow this link:

http://www.screencast.com/t/NjQ4Mjg2

blogger.com makes it really hard to embed regular sized video.

Home Built Control Systems?

In early May I attended a NERCOMP conference hosted by UMass Amherst. A group from Wesleyan University presented what they are calling RoomTrol. It is a control system that they have developed in-house with student programming. The system has a web (Java) interface, and you can use any touchpanel with a USB connection. The control system is a mini-computer (like a mac mini, but a PC) and is running linux. For RS-232 ports and IR ports they use USB-RS232 or USB-IR conversion cables. The input from the touch panel to the "control system" is USB. The group reported that the entire "control system" and touch panel cost them under $1,000. That is a significant savings from what we are used to. If you buy a similar product commercially, you are going to spend around $5,000 for a control system and touch panel. Also, they are using the switching functions built into their projector. So they have no need for other switchers, again another saving. They have had this system in pilot mode for one year in a room, and over the summer plan to deploy it to three spaces. This initial deploy will cost $12,000 less than other installs.

You may be interested to know that this University has a very interesting system of supporting rooms. In fact, all the support is done by student staff. The entire design and programming for this new system has also been done by students. So, here again they are saving money by not having their professional staff creating this system.

I was excited at the potential cost saving, but left wondering if they really know what they are getting into. Obviously, someone is writing the drivers for this system. What do they do every time they have a different model projector, LCD panel or IR device. How much time is spent writing the modules for these? And, even with student labor, is it worth the saving? What about when things simply don't work? Do they have the time to do the troubleshooting on a home grown system? Finally, what about all the things that professional systems provide that this system does not? In my programming, I use video sync info from the processor, I use touch panel activity indicators and all type of other useful tools that are provided to me. How much time would it take to have someone program all of this? Finally, I will be the first to admit that when all heck breaks loose, I like having someone (a company) to point a finger at. Who does the finger get pointed at, when you created the system?

When I thought more about this on my 4 hour drive home, I began to wonder if I just saw the first ripple in a tidal wave coming at the A/V industry. What Wesleyan is really doing, is creating open source control systems. Is there a future in this? Would having support networks in place, like those that exist for moodle, sakai and others applications answer my questions? Certainly, the questions I have posed are the exact questions my school (and many, many others) have asked and answered before switching from a commercial product, like BlackBoard, to an open source one, like Moodle.

What do you think? Have you ever considered such a revolution? Is it possible that in five years a system like Wesleyan is building could be common place in our "hang and bang" classrooms? Or, is this a project that works very well for one institution, but will not be easy to replicate at others? Leave a comment and let me know what you think.

Thoughts from a weekend emergency

This weekend there was an event on campus that involved a presenter bringing their own laptop. I got a panicked call on Saturday night, because it was not working. Luckily, I was only minutes away from campus, so I was able to stop by and see what was going on. The VGA input for the laptop had been damaged by a previous user and therefore was not working. As I scrambled to find a solution, a couple thoughts came to mind.

1. On the inside of every rack should be a schematic drawing. In an emergency trying to trace back a wire, without a schematic gets to be very stressful. Having a schematic would have allowed me to relax a bit and have the time to work through the rack properly. I imagine that I am one of the last people to think of this, but better late than never!

2. Is all the Cat5 we are using for installations really an advancement? On Saturday, my saving grace was disconnecting the VGA from the rack, at the wall plate and plugging the laptop in there. In all of our new installations, this connection would not have existed. All that would be at the wall plate was a Cat5. I could try all night, but I don't think I could have plugged the VGA into the Cat5. While these installations are certainly easier to install and design, they also take away a lot of our learned experiences on how solve problems in emergencies!

Can AV Managers Learn from their IT Counterparts?

A/V technology has rapidly changed over the past several years. The technical changes have been blazing fast and demands from our users grow at an almost equal pace. In the education world, A/V shops have transformed from staff who push around A/V carts to highly trained technicians, system designers and programmers. Despite these changes our basic charge has stayed the same, "provide the technology that I need and make sure it works when I need it". Many technology managers feel immense pressure to somehow manage all of this change, manage the exponential growth in equipment on our campuses and provide the same level of service that we did when we had more staff and less equipment. Bates College, for example, has experienced a 400% growth in A/V equipment and installations over the past 6 years, while losing one full time position. Unfortunately, many of the leaders of our institutions are unaware of how quickly A/V has evolved and still view the position in A/V departments as low skilled. At this critical point in the growth and importance of A/V in our institutions we can learn a tremendous amount from our colleagues in IT. After all, they went through, and continue to go through, similar changes with computing. Don't forget there was a time when the IT group was considered the "techies" whose only charge was to keep the computers working. Rightfully so, IT now has a role in almost all aspects of campus life.

At Bates College we have moved towards a model that borrows from the IT structure. Our first and most significant step towards this transition has been developing statistics on usage and problems. Talk to your IT guys, they can tell you how many e-mails come into campus on a day, how many help desk calls were made during a given time period, how many computers exist on the campus and they use this information to gain support for their work. At Bates, we are using Crestron RoomView for our reporting needs and the results have been phenomenal. Every month I am able to report detailed statistics to my staff, administration and faculty about A/V use on campus. Frankly, the usage blows us all away and makes us realize A/V is in constant use and is remarkably reliable. We currently have 50 classrooms on campus with full A/V installations. In those 50 rooms the A/V systems were used over 5,000 unique times in the fall semester of 2009. We had a total of 70 trouble calls during the first semester that were technical problems, resulting in a 98.9% success rate of technology in our classrooms. I don't want to lessen the importance and significance of the 15 problems, any time there is a problem it is serious, but I also use this data to point out to our administration the incredible job that is being done in keeping the equipment running. Compare this to what goes on in the IT side. If your CIO got 15 e-mails a month from people complaining about computing problems (and he/she probably does) they would simply chalk it up the fact that technology does not always work the way users expect it to, realizing that 15 calls is a pretty small number considering the usage of the network and computers. However, 15 e-mails to that same CIO about classroom technology problems would probably raise a "why does this stuff never work" response. If your CIO has this response, but has no frame of reference for the complaints, then what other reaction could be expected? You have to provide the data on a regular basis whether they ask for it or not. If done right, this data should be readily available for you to provide to your superiors without much of a time commitment on your part.

Another issue we grapple with on a regular basis is user support. For so long our mode of operation has been to drop everything, run to a problem and do anything you possibly can to fix the problem. I believe this can be traced back to the early days of classroom technology in which the equipment was a mix match of various technologies, much of which was custom built, and frankly, did not work consistently. Clearly, the technology has gotten to a level where this is not as much of a problem. Equipment is more reliable and many of our users are more comfortable with the equipment. However, there are still some problems that occur and training still needs to happen. Again, I suggest we look to IT to see what they have done that is successful and borrow what works. Most of us have borrowed the idea of a help desk from the IT side. I don't know of any institutions that don't have some form of an A/V help desk. How about classroom troubleshooting? What does a person do after hours when they can't get their laptop to work or there is no sound from the computer? Many institutions I have talked with either have no support for this type of situation, or they have paper printed and put in the classrooms. When was the last time you saw your IT department give a manual to a person when they delivered them a computer? They don't, instead they use the tools at their disposal, namely, helpsheets on the web. Use the same resource, or better yet, use the resource that may already be in the room, namely the touchpanel. In our classrooms, our touchpanels themselves are the troubleshooting tools. They tell the user whether the laptop is outputting a signal, and if not give the user the option of reading steps on how to ouput video from their laptop. Using current sensors we can detail some simple troubleshooting, such as whether the dedicated computer or the document camera is on, yes, we too get calls from people saying something didn't work only to find they had not turned it on. If someone presses the help button to report that there is no sound from the computer, the touchpanel shows instructions to check the volume on the computer, as well as the volume level on the touchpanel.

There has been a lot of talk on the A/V channels over the past few months about the digital transition and the host of problems that come with these changes. Particularly, how do we deal with super-high resolutions and connection types that are not standard in the classroom (i.e. HDMI, DisplayPort). While well intentioned, some people jump to conclusions like, let's have an emergency kit of every adapter type in every classroom. Again, that feeling of "we must make anything work" stirs in us. Yet, we know that we can not possibly achieve this goal and trying to will only result in more failure. A pile of adapters looks very messy, they will disappear and likely the users won't know how to use them. By making all these adapters available, you would be declaring support for all of them. We followed the lead of our IT department in this area as well. IT has a very clear list of software that is supported by the college. For example, we support Microsoft Word. If you choose to use Google Docs, Open Office or some other word processing application, the help desk will not support you. In the A/V group we published a list of supported input types (DVI, VGA, etc.) along with supported resolutions (XGA, WXGA). When someone brings a visitor to campus, we point them to this document and let them know what we support.

Finally, the last big thing we can learn from IT is about budgeting. This is a sensitive issue at the moment because everyone is getting their budgets slashed. However, I believe that IT does a very good job of letting people know exactly what it means to cut parts of their budget. Our network folks make very clear that cutting the preventative maintenance budget means they will not be able to replace switches and servers, and that it WILL result in more down time. Our desktop folks make it very clear that stretching out the desktop replacement cycle would mean that you end up with machines that are not under warranty, creating more work and costing more money. If the budgets in this area are cut, and there are failures, again, the network and desktop people do not "take the fall", and rightfully so, they point out that the failures are a direct result of budget cuts. Are we explaining very well what it means to skip a year on the replacement cycle for displays or control systems? Have you talked with your CIO about the change to digital interfaces that is well underway? Do they understand clearly that not replacing projectors and internal A/V system parts, that they will not be compatible with say, DVI or HDMI? Do you take it a step further and show them how this will affect faculty, staff and visiting presenters? If not, then it makes perfect sense that they cut your budget when the fiscal knife comes out. There are many resources available to provide you with data on the transition to digital. Use these resources to learn about these changes and talk with your superiors about why the equipment needs to be part of a regular replacement cycle. As technology managers, it is our responsibility to educate people in our organizations about the technical changes that have taken place in the industry and the correlation to change in job skills, responsibilities and fiscal support needed to provide reliable A/V on our campuses.

The IT guys did this years ago, now it is our turn.

Lecture Capture - Why So Expensive

Over the past year or so, I have been investigating Classroom Capture systems.  There are some great systems out there, but they are SO expensive.  I truly don't understand why.  Sure, there are some systems that are free, and others that are not incredibly expensive but require LOTS of configuration and in-house programming.  And most of those systems require some work by the professor to get them running.

I guess that the other systems are so expensive because the market will bear the cost, but I wonder why.  How can it make sense to pay $10k or more for one of these systems.  Ok, ok, I know why it makes sense to pay that much money, it makes sense if you are turning around and making a profit off the recordings that are made.

What about the rest of us?  What about those of use who want a system that fits the following requirements:
- no software on the computer, recording is done via an inline box
- simple to use
- no more that 1 day of server install time
- less than 5k per unit
- NO recurring software costs?  Why should I pay for something I already own!!!!!

Beyond the blatant (and powerful) use of this technology as a way to review class sessions there are so many other uses that smaller colleges, like the one I work at could explore.  Here are some examples:
- pre- lecture modules:  a professor puts up a recording of a particularly difficult subject, from the previous year's lecture and lets students review it before coming to the current year's lecture.
- learning modules:  a professor records himself/herself lecturing on a specific topic, with examples, and posts it in the school's LMS.
- learning differences.  There are so many examples here.  Students who have learning differences have oppurtinities to focus in class, without taking detailed notes, knowing they can review the lecture later.
- students with illness: During the recent swine flu outbreak, we learned that the best way to avoid illness is to stay away from others when we are sick.  Serious students may have a hard time not coming to class.  They may be more willing to stay home and get well, knowing that the lecture would be captured.

Because these examples don't bring in revenue, there is no way a small college can make this technology widespread at current costs of the equipment.  Plus, what is so complicated about this technology that makes it so expensive?  They are essentially computers, with a piece of software on them.

There is a big market out there for someone who figures this out and sells the product at a reaonsable price.  When that happens, I hope they call me first!