I have previously used the Tobii system, and currently use the Eyelink 1000. I much prefer the Eyelink, in terms of transparency in how it works, and in terms of the skill level of the tech support.

SR Research Eyelink 1000/2000 can be run in remote mode at 500Hz without chin/forehead rest. You just put a little silver sticker on the kid's forehead, and as long as they keep their heads basically within the bounds of the monitor, it tracks just fine.

I would also recommend the SR EyeLink 1000. I use it in remote/desktop mode without a chin/head rest and have found it to be quite reliable and robust. I mostly test adults (including older adults and adults who have had strokes), but I've collaborated on a few projects testing kids with the same equipment and it seems to have worked well.

The main other eye-tracker I've used is the ASL 6000 with Flock of Birds head tracker. I think the biggest advantage of the EyeLink is easier calibration and much better performance for participants with contacts or glasses. It also has higher sampling frequency (250Hz vs. 60Hz) and simpler head tracking (a forehead sticker instead of the magnet headband of the Flock of Birds). I've also found their tech support more helpful than ASL and Tobii (I never really used the Tobii system because we couldn't get it running and couldn't get any help from tech support, so we gave up and got the EyeLink).

I have nothing but wonderful things to say about the SR Eyelink 1000. I've used this tracker in projects with infants between 18-24 months of age, and in children between 3-10 years old. It captures excellent quality data, and there are useful tools for experimental presentation and data analysis. SR Research has been highly responsive to all of my questions and seems to understand my research needs.

In addition to the EyeLink system (which I also can highly recommend) there is the LC Technologies EyeFollower. The system allows movements within a certain range in front of a screen (see for details http://www.interactive-minds.com/eyefollower), has an excellent spatial accuracy and should be less expensive than the EyeLink system. The only disadvantage--depending on the research question--it samples with 120 Hz only.

Many thanks for your really helpful answers. It's good to learn about what else is on the market, and what others are using. Oftentimes decisions on what system to use aren't actually based on picking the best system out there, but what systems one has used in the past, or those that colleagues prefer (Which isn't necessarily a bad thing, given greater ease of use and potential support)

I have heard good things about the SR Research 1000 as well. I was wondering, has anyone who has used it, also used other trackers, and might be able to elaborate on the pros and cons of either system? In this light, many thanks to Sebastian for pointing out the LC Technologies Tracker. The Tobii X120 we are using in our department samples at 120Hz only, too, and so far that hasn't been a problem for the kind of research we are doing, but it may be for others.

I have previously used the Tobii system, and currently use the Eyelink 1000. I much prefer the Eyelink, in terms of transparency in how it works, and in terms of the skill level of the tech support.

I have used the SmartEye system; calibration is easy and it works quite precisely.

http://www.smarteye.se/

http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2842.2011.02219.x/abstract;jsessionid=A7B45712673E856C64D8274755459069.d01t03?deniedAccessCustomisedMessage=&userIsAuthenticated=false

I am using the Mirametrix S2 with children with autism and am very happy with it. My needs are not clinical or related to basic research. I am doing interface design for 3D virtual learning environments. A highly portable system like the S2 is ideal for my work. I do not have a lab; I set up my equipment where my participants are available.

I use the S2 along with the bundled viewer software to capture video. The viewer software also captures all gaze/saccade behavior to a CSV file that can be imported into other analysis suites. In addition, I use the open source Ogama analysis software from the University of Berlin (http://ogama.net/). It's not perfect and requires power user proficiency in Windows, but is sufficient for my humble needs.

The accuracy is quite good, even with some head movement. The calibration remains sufficiently accurate for my needs even after participants leave and come back. I do not need pixel-perfect accuracy though. I am generally interested in regions that are at least 100x100px.

The pricing is quite good, although the EyeTribe may give them a run for their money.

I have used the Applied Sciences Mobile eye for many years, coupled with external motion analysis cameras. It is light, accurate and has an excellent frame counter. It outputs very good videos of the eye of the performers and their gaze as they perform. I use our own software, Quiet Eye Solutions Inc to couple the gaze and movement videos together precisely in time. The software requires frame by frame coding. Not as time consuming as you might think as most motor skills are of short duration.

We recently used the SMI mobile eye tracker and it is superior to the ASL in terms of calibration and worked better than the ASL ME in surgical knot tying, but in tasks where the visual field is not just straight ahead (as is common in most sports tasks) it is limited in recording all of the gaze of the performer.

So, given these two excellent eye trackers, the key is to understand your motor task and the visual field of novice to expert participants and purchase accordingly. Or better still, buy one of each.

It depends on the type of task. If your research question requires participants to translate and rotate their heads a lot (e.g., in a driving simulator), you will probably have to employ a multi-camera system with head tracking capabilities (e.g., SmartEye)

If your experiment could be performed in more rigid conditions behind a PC, I would recommend the Eyelink 1000.

Hello! I know this question was asked originally a while back, but I wanted to contribute my experience, since I'm sure people still see this while searching for answers. I worked with a Tobii T120 and a Tobii T60XL at two different universities, and used Tobii Studio software. We studied infants 3-18 months, and found tracking to be really good (with the exception of the 3-4 month olds, which is just a little trickier). Infants are notoriously fidgety, and the Tobii eye trackers are great at tolerating movement. Tobii Studio is super easy to use - I learned how to use it on my own as an undergraduate.

In full disclosure: I do work for Tobii Pro now, which should be a testament to my experiences with the technology, employees, and company as a whole. The T120 and T60XL are now discontinued hardware products, but the Tobii Pro Spectrum (up to 1200 Hz) would be a great solution for testing infants and young children. It has a large tolerance for head movement and the best accuracy/precision metrics on any of our hardware ever built. I'm attaching a product brochure for the Pro Spectrum, as well as one of my articles using our Tobii T60XL system for reference. https://www.tobiipro.com/product-listing/tobii-pro-spectrum/

Please feel free to reach out to me directly if you have any questions! My email is marisa.biondi AT tobii.com.

Dear all of you.

I have learnt you were using a S2 Mirametrix eyetracker. We have one in our lab and we are experiencing some troubles. Infrared emisor flashes and video image displayed in the PC monitor is not stable. Have you ever experienced something like this?

Regards,

Francisco