 And I understand you're a fan of eye tracking because of what? Because your experience is actually quite complex and in the past it was very invasive, right? I understand the advantage of this is that it's perhaps more objective, right? Like, because we could just go to people and ask them, like, do you like the subtitles that you watched or are they too fast and so on? We could just ask people the rest. Of course you can ask them, like, did you have the time to read those subtitles and everybody would say, yeah, yeah, yeah, of course, right? Like, I didn't have the time. Yeah, so eye tracking actually allows us to verify whether they have indeed had the time to read the subtitles. And it's been found, not only in my studies, but in studies by several other researchers that when subtitles are too fast, people very often cannot cope. So they do not read subtitles to completion and this is evidenced by no fixations or fewer fixations towards the end of the subtitle. And this is really hard evidence that we can use, for instance, to modify subtitling guidelines. Okay, so I understand this shows us that somebody starts reading the subtitle, but never, like, never got to the second line or never got to the last few words. So this is, like, objectively, you're seeing that they haven't had the time, enough time to read this whole subject. That's right, that's right. There has been research showing that when subtitles are slower, there are more fixations on the subtitles, meaning people spend more time looking at them. And again, with faster subtitles, we have fewer fixations because these subtitles are just available for a shorter amount of time and people don't often get to the end of the subtitle. And a previous eye tracking research on subtitle reading focused mostly on subtitle level data because of technical constraints which are now going away. So previously, we had to use subtitles that were burnt into the image and as such, they were not visible as text to eye tracking software. So we had to draw areas of interest on every single subtitle and analyze them as a chunk, as a whole. Now what we can do is we can link particular subtitle files and we can analyze the reading of subtitles at word level without having to draw anything manually. So we are very excited to see this new development. So it sounds like there's less work now involved in designing... There's more work somewhere else, yeah. But there's manual drawing of various interests, yes. There used to be an exciting future for eye tracking research with all the technical development and so on.