Hi Rita,

Ive always used the ASL mobile eye tracker and its always been a reliable piece of kit. It's easy to calibrate, lightweight and very easy to use. We have used this for a variety of tasks to (e.g., golf putting, football penalty kicks, dynamic rugby tasks, basketball shooting etc). One negative is we havent had much luck in collecting data from outside due to sunlight issues. I've never tried the others so can't comment.

I would be interested in seeing which one you go for (I will be looking to buy one soon too!)

Greg

Thank you Greg; could you tell me which model of ASL you were using? I'll definetely post my final conclusion here.

Hi Rita,

we have been using the ASL Mobile Eye; we had participants run for fly balls (inside conditions). Worked quite well. Contact me if you want more detailed user experience.

Best,

Frank

The most recent version of the one we use is the Mobile Eye-XG. This saves the gaze data to a SD card, via wireless or via a firewire cable connected to the laptop.

Hi, Rita.

We have an ASL eye-tracker (H6 model - I also used previous versions as 3000 and 5000). It does work well in indoor conditions; I have no experience in using it outside (I heard the system has improved to work stably in more intense light conditions). One aspect I believe you should pay attention is the acquisition of head movements, depending on how important is the eye-head integration for your studies. If you want to have 3D line-of-gaze descriptions, there are options for simultaneous recording of the head (e.g., Ascension Flock of Birds, Optotrak) in the ASL system; additional calibration is required. We are now using our system with more controlled situations into the lab, but I've collected integrated eye-head data in table tennis context in which ASL system worked quite well. Feel free to contact me for additional details.

The Tobii eye glasses are quite neat and light-weight. Depending on the level of body contact (e.g. someone getting hit by another player and knocking the glasses out of position) it might be suitable. It can be easily fixed snugly enough to the head so that when running it wouldn't move. It also allows has some IR markers that you could affix to other objects (e.g. the ball, goal, other players, etc.) and works well in poorer lighting conditions (e.g. dark clouds passing over!). As with other mobile units you sacrafice accuracy and data rates; this now has HD quality too and can use SD cards similiar to those mentioned above.

http://www.tobii.com/en/eye-tracking-research/global/products/hardware/tobii-glasses-eye-tracker/

Hi Rita,

Perhaps you would like to consider the Eyetracking Glasses from SMI (Sensomotoric Instruments). We can highly recommend them, especially for their way tracking is done (there are no mirrors as for the ASL tracker)

While we have used them only indoors, we have had excellent experience with handling capabilities and, most importantly, calibration and data quality. They are spot on accurate after only a 1-point calibration. And they offer automatic parallax compensation (automatic correction when a person changes focus between a far and a close object), a feature that I believe is important in sports as well (e.g. when focusing on a tennis ball). According to their website, they are already be used for eyetracking studies in sports.

I am definitely also in line with Sérgio on that you might want to pay attention to head-tracking depending on how you want to analyze the data. The SMI eyetracking glasses offer similar integration options as ASL.

Let me know if you think I could help you further with more information!

Best,

Jan

Hi Rita,

He are using both the Mobile Eye and SMI here at the VU. I would say that the Mobile Eye gives you the most flexibility in terms of the number of different tasks you can examine. The design means that it can be used for most tasks (though as Greg says, it can be difficult outside). I would say that the key disadvantages though are that the proportion of participants where the data are unusable is higher than for the SMI, the calibration and analysis can be very laborious, the software designed to detect the pupil could be better, and we have had problems with the customer support from ASL. The SMI is fantastic, provides high quality video, and is easy to use (calibration is simple). Unfortunately though it has a very thick frame and so it can block the line of gaze, particularly if you want to examine tasks where you are looking up (compared to the direction of your head, for example in snooker/pool) or when you look down (for example when looking down to hit a table tennis ball). Also, I think the frame rate is lower than the ME (though this could have changed in the last few years).

So on balance, if you have a wide range of tasks you want to examine then I would get the Mobile Eye. If you have some clear defined tasks that you know the SMI would work for then I would say it is a better suited instrument.

Good luck!

Cheers,

David

http://www.youtube.com/watch?v=2NcUkvIX6no

Hi Rita,

The Dikablis eye-tracker is very light weight, can be worn over glasses and is easy to calibrate. It also links in very well to vicon motion capture systems to allow for synchronised body motion and eye movement capture.

Once the new version of D-Lab is available it should allow for some very nice data analysis features. I am using it in Clinical research involving walking. However, the above youtube video shows how it can be used in sport.

I would say however, that depending on the eye movement events that you are considering analysing you may need a system with a higher temporal resolution (>50Hz).

regards

Sam

It would seem that yu have already found the top companies in the market. Choosing one dpends on your desired application. After deciding on desired protocols and measurement accuracies, I would suggest examining for each: ease of set up & calibration (can be few min to half hour), accuracy of measurements (currently 0.1deg- 1-2 deg), how data is stored for later study (do you want to pack a PC on their back?), what sampling rate ( video 30Hz to even 300Hz), degrees of freedom (1D, 2D of 3D; yes some even can track torsion), etc. It all depends on your application, the lower the specs the cheaper it gets and even trackers used for games might be sufficient. Note that if you intend to study rapid eye movements (saccades) as part of the measurement then your sampling rate should be at least 200Hz, otherwise if you don't care about precision during saccades, and are focusing on fixation points or slow orientation, a lower sampling rate would suffice. However, never forget that saccadic data would be poor.

Hi Rita

I have used the tobii glasses quite a bit. They are lightweight and quck and easy to calibrate as Stephen said. I have found them to be very poor outdoors though where they just do not work to the level that you need for a research project. Indoors they work seamlessly and the studio software can be good for analysis too.

regards

Mark

Hi Rita,

I was facing the same question the last month. I have been to several conferences, where TOBII and SMI presented and I had the whole instruction by Dikabilis. I have not seen the ASL system yet.

What convinced my was the nice integration of an EEG for the SMI system and the qualitiy of the data I saw for the outside. Additionally, I do not need a system that combines some aspects you would be able to do with a stationary system (I will get an SR research system there) and a mobile one. The dikabilis system seems the best to have something of both worlds..

Last, if you want to combine various measures (like I will try to integrate EEG, eye-tracking and kinematics), then you should make sure those fit with each other...

Bests

Jörg

Hi Rita!

The first question is to know what kind of sport are you interest in.?

I already used ASL eye tracking system (H6) and sincerelly , if you are thinking of any kind of contact sports, forget it. The front mirror would put participants physical integrity at risk

Despite the boring calibration process, a chess match is a situation on which ASL H6 can be used.

If you are thinking on contact sports, and think that Tobii eye glasses are too fragile ( I have used it!!!). I would use the new SMI eye tracking glasses 2.0 , which are more robust and provides a binocular eye tracking performance of 60Hz.

The second question is: what do you want to measure?

fixations ? smooth pursuit movements? scanpaths???

You should pay attention to some small big details on the eye tracking sys you want to use, such as ease of set up, sampling rate, accuracy of measurement, data storage, analysis software, signal synchronization (integration with other measures) and verifiy if they are suitable to your research goals.

Hope it helps.

Cheers,

Pedro

Hi Rita,

I also recommend SMI eye-tracking glasses. They capture high quality and high frequency data, and are pretty sturdy for sports data collection. The only point to consider is the extra cost for purchasing the mobile data collection unit, which is an ipod-like recording equipment that captures data in real time and data can subsequently transferred to a computer. Also, a nice feature is that you can calibrate after recording, in case you see a gaze pattern is off. One last point, SMI has recently integrated their portable recording into customized Samsung S4 cellphones, and the performance seems fast and accurate. I hope I was able to help.

Dear all, thank you for all the helpful comments! Finally, I've decided to acquire the wireless eye-tracker from Ergoneer Dikablis which is coming out in February. The main points in favour were: 1) no obstruction of peripheral vision; 2) wireless and mobile; 3) mobile cameras with possibility of different lenses so large and adaptable tracking range; 4) the contact person was the salesperson and software trainer, he had an engineering background and could answer the questions well; 5) software appears to have similar functionality as other systems.

Other systems we looked at had different short-comings: SMI obstructed peripheral vision; ASL [has asked me to remove my comment]; Eyelink didn't suit because it isn't mobile but the support people clarified this; Tobii obstruction of peripheral view.

In the next few weeks we should receive the system and try it out as well as the software so I can post an update here. Thanks again to all who helped!

Great Rita! In case, in future if you need a fixed ET system you should check out this new one from gazepoint.

Cheers

http://gazept.com/

Rita, what was the conclusion? Did you find a solution? Our ASL XG is still going strong (although our 2 older models have now died), and we have an SMI on it's way... did we 'miss a trick' by avoiding Tobii?

Thanks again for all the comments! I hadn't heard of pupil labs but on a quick read this sounds like a good system to have for undergraduate students' projects (at least). The same goes for gazepoint. I'd be interested to know how these compares with other systems.

In the meanwhile we had a delay with receiving the wireless system so they sent us the wired system (which we can use in the non-sport current project) and we had a training day last week to go through the software system (this will be released with the wireless system together with its manual which isn't available yet but we got an advance). I was very happy with the software: it identifies the pupil as soon as you put it on and it tracks it to the extremes of the eye; the calibration takes 10 seconds; it easily identifies areas of interest (AoI) through printed markers on the setup and gives you data summaries for these automatically; but importantly it allows re-calibration during the analysis (after the recording) and definition of epochs for the analysis.

I'll be testing during Easter so will probably be able to identify the problems but so far training and piloting went all smooth with the Dikablis.

Dear Rita de Oliveira, if talking about sports you are alluding to the need to leave the athlete as much as possible free in his movements (as I think in your case), I would suggest the EOG (Electro OculoGram) which uses electrodes suitably attached on the subject's face in Horizontal and Vertical position (G.C. FILLIGOI, L. CAPITANIO, F. BABILONI, L. FATTORINI, A. URBANO, S. CERUTTI "Reduction of ocular artefacts in source current density brain mappings by ARX2 filtering" Med. Eng. Phys., 1995, 17, n.4, 282-290; F. BABILONI, L. FATTORINI, C. BABILONI, L. CAPITANIO, G.C. FILLIGOI, S. CERUTTI, D. LIBERATI, P. ONORATI, A. URBANO "Spatial distribution of single-sweep event related potentials is improved by combining autoregressive with n exogenous input filtering and surface laplacian transformation of the potentials" Proc. V° Int. Symp. on Biomedical Engineering, Santiago de Compostela (Spain), Sept.26-28, 1994, pp. 165-166). By means of EOG you can get just a qualtitative and not exact information on eye movement. Insted, if the head athlete's movement is not so critical you can use some infrared lighter and mini-TVcamera mounted in transparent glasses. By means of this technique you can get a more precise information on eye positioning, but at the expense of some head movement difficulties.

Dear Rita, 

We are planning to buy an eye tracking glasses and I was wondering if you are happy with your final choice of Ergoneers Dikablis? Any comment or advice on this would be helpful and appreciated.

Cheers,

Mohammad 

Hi Rita

Might you be able to comment on the performance of the Dikablis? I am unable to find ASL's website (perhaps they have gone out of business??) and the Dikablis seems to be the only other mobile eye tracker that doesn't obstruct peripheral vision.

BW

Sheree

Hi Sheree,

It works alright but sometimes it gets a bit fiddly. But then again, they mostly are... You're welcome to visit LSBU and try it out if you like! Or you can contact the guys who sold it to us as I'm happy with them and they'll go round for a demo if you ask them (i.e., I assume they also go out of London) http://acuity-ets.com/index.htm

Cheers,

Rita O.

We used Pupil Lab eye trackers for some simulated and real-world driving studies and found that worked quite reliable and accurate.

We are also very happy with Pupil labs.

Check out this video (there's a bunch of them):

https://figshare.com/articles/Supplementary_Movie_2_17_21_31_r_mp4/4498613

from a recent paper:

http://journal.frontiersin.org/article/10.3389/fpsyg.2017.00620/full

You can see the calibration being done in the full video:

https://figshare.com/articles/Supplementary_Movie_1_full_video_mp4/4498466

The most important thing about Pupil Labs is it's open source code so you can customize and improve things according to your needs if your team has computer science skills, and you get access to the raw eye images. 

0tt0