Hi. Is there a way of downloading the Neon Scene video with the gaze overlay visible? At the moment I can only see this on the Cloud itself. Once I download the Scene video the overlay disappears.

Can I create a scanning recording to use in Pupil Cloud with a Neon and still have the reference image mapper use data collected with an Invisible?

Hi all! My team is in the process of designing a study using wearable eye trackers and I had some questions about viability of using Neon. Who would be the best person to reach out to (or would posting here on Discord suffice)? Thanks so much!

Hi Neon team, is there a way to save the settings of Neon Player, and batch process multiple recordings using the same set of settings in Neon Player?

Hi all, in order to make our order from the University i have to register this company (Pupil Labs) as a supplier with our University. Can I use one of you as a contact for the form? @user-480f4c @user-d407c1

eye

Hello. I wonder about the average gaze point producing frequency on the current companion device. I guess 200 Hz is a fairly high amount for a mobile device which will run with a neuro network. I was hoping that you could share this specification. Thanks..

Hello Nadia. Yes, the eye cam operates at 200 Hz. However, I was curious about the approximate frequency of the pupil detector on these phones, during real-time usage, not from recordings. I guess I got the answer as 200 Hz for that too. Thank you.

Hi @user-7b683e 👋 ! To expand on my colleague's response, you can adjust the inference sampling rate through the Companion App. Simply navigate to the settings page, where you'll find options to set the rate at 100Hz, 33Hz, or even disable real-time inference altogether, in which case you will obtain the gaze data in Cloud at 200Hz post-recording.

We were recording during the day out in the sun. Could that be the problem?

We got this error for 2 recordings and both those recordings did not get uploaded to the cloud.

Hi @user-37a2bd, Could you open a ticket in #⁠⁠troubleshooting? We'll continue the conversation there.

Hello, I'm Wiss from Thailand who having a question about un-uploading clip to a PupilCloud

I did record by NEON on last Friday 3rd May 24, while disappeared without any uploaded to a PupilCloud. What should I do, coz I need to see the analysis so soon?

Plz kindly suggesting

I still have 2 record files in a companion device (one record from world camera, and another record of subject's eyes)

Hello – I have a question about the coordinate systems for the gaze and imu data. We are trying to calculate a gaze angle that also accounts for the head position (so if the participant’s head is tilted up/down, it would change the gaze angle). Our initial plan was to get pitch from imu.csv and the gaze elevation from gaze.csv and then add them together. In the documentation it says “The IMU's coordinate system is rotated by 102° around the x-axis in relation to the scene camera's coordinate system.” – but it’s not clear how to actually convert the coordinate system to get things properly aligned. Do you have any suggestions on how we can combine the pitch and elevation to get gaze angle relative to the horizon? Thank you!

Hi, I was wondering whether it is possible to download the gaze raw data AFTER doing an enrichment (mapping to 2D). Otherwise the gaze data coordinates are always in relation to the video of the recording I guess and can't be compared between wearers?

Hi @user-861e5b ! Which enrichment are you referring reference image mapper or marker mapper? After performing an enrichment, you will get a separate gaze.csv file which contains the remapped gaze in reference image or surface coordinates for all the recordings enriched.

Hello! I am having issues with getting the reference image mapper to function. I upload the scanning video, a still image (set a few AOIs), and include the test video and each time it throws an error after processing. I think I am following all the instructions, but I am not having any success

Reference Image mapper error

Hello. API question. Is it possible to have the realtime API and RTSP servers visible when you connect to the phone's hotspot? Right now it is only visible when the phone connects to the hotspot of another device (e.g., laptop).

Hi @user-d407c1,

we are running a study with pupil labs and other sensors using LSL to synchronize all data streams. For no obvious reason your lsl relay stopped working lately.

Error message: [05/09/24 12:48:47] ERROR Raw gaze data has unexpected length: [email removed] gaze.py:78 [email removed] \xbe\x9a;\xe8>\xa9\x10,?e\x01U', timestamp_unix_seconds=1715255328.2962165) Traceback (most recent call last): File "C:\Users\XXX\PycharmProjects\eyetrackerlsl\venv\lib\site-packages\pupil_labs\realtime_api\streaming\gaze.py", line 75, in receive cls = data_class_by_raw_len[len(data.raw)] KeyError: 65

Did you change the variable with an update?

Thanks.

Next time it might be helpful to write here that there is an update: https://pupil-labs-lsl-relay.readthedocs.io/en/stable/history.html#id1

@user-d407c1 I did the update and apparently it works but I get a huge number of Warnings:

       WARNING  Dropping unknown gaze data type: EyestateGazeData(x=834.7101440429688, y=469.3448181152344, worn=True, pupil_diameter_left=4.754848957061768, eyeball_center_left_x=-27.83203125, eyeball_center_left_y=11.42578125, eyeball_center_left_z=-48.80615234375,                                   relay.py:92
                         optical_axis_left_x=0.1795196384191513, optical_axis_left_y=0.21891407668590546, optical_axis_left_z=0.9590877890586853, pupil_diameter_right=4.134486675262451, eyeball_center_right_x=34.16015625, eyeball_center_right_y=9.1845703125, eyeball_center_right_z=-52.77099609375,
                         optical_axis_right_x=-0.2444096803665161, optical_axis_right_y=0.21104899048805237, optical_axis_right_z=0.9464260339736938, timestamp_unix_seconds=1715258636.4630961)

@user-d407c1

Traceback (most recent call last): File "C:\Users\XXX\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "C:\Users\XXX\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in _run_code exec(code, run_globals) File "C:\Users\XX\AppData\Local\Programs\Python\Python310\Scripts\lsl_relay.exe__main__.py", line 4, in <module> File "C:\Users\XXX\AppData\Local\Programs\Python\Python310\lib\site-packages\pupil_labs\lsl_relay\cli.py", line 16, in <module> from pupil_labs.lsl_relay import relay File "C:\Users\XX\AppData\Local\Programs\Python\Python310\lib\site-packages\pupil_labs\lsl_relay\relay.py", line 7, in <module> from pupil_labs.realtime_api.models import Component, Event, GazeDataType, Sensor ImportError: cannot import name 'GazeDataType' from 'pupil_labs.realtime_api.models' (C:\Users\XXX\AppData\Local\Programs\Python\Python310\lib\site-packages\pupil_labs\realtime_api\models.py)

Or wait.

Apologies, it is my fault. I imported GazeDataType from models instead of simple.models, I updated the branch to properly import it

Now it seems to work without Warnings. Let's hope that the data is fine too.

Software testing helps 😉

The data should be the same. Definitely! I'm sorry for the trouble. I didn't have the chance to test it here right now, but I wanted to allow you to continue with your work as soon as possible. 🫡

Hi Neon team, I encountered the exact same problem as described here (https://discord.com/channels/285728493612957698/1047111711230009405/1210375932682965002) and I tried the solution suggested in the response (https://discord.com/channels/285728493612957698/1047111711230009405/1210531886087012363). Since then, when I connect NEON to my smartphone, it displays “device name: N/A” and when I take a measurement, it just displays “No scene video.” Is there any solution to this symptom? Thank you.

@user-d407c1. Does the new version use more resources on the phone? It overheated during the measurement and stopped all apps.

Actually it still recorded the data but displayed the warning

It would be really great if you could release an adapter that lets me connect the eye tracker directly to a pc. Then I wouldnt have to deal with this extra phone and wifi streaming

Hey all, should i contact the sales email address for a quote? We need an official quote to add Pupil Labs as a supplier into our univerisity system

I sent an email but havent heard back

we are so so so excited to get these new glasses!

Hi @user-82f555, thanks for reaching out 🙂 Could you send me your email address in DM just to make sure that we have received it? The Sales team will reply to you asap.

Hi @user-a5a6c3 ! No worries, let's start with this example and explaining some of the asyncio functions.

Here, asyncio.Queue is used to temporarily store video frames and gaze data as they are received. This ensures that the incoming data doesn't get lost if processing can't keep up immediately. Essentially, it acts as a buffer, holding data until it's ready to be processed.

asyncio.create_task() in that code schedules tasks to continuously receive gaze and video data without waiting for each operation to finish. This means your code can receive new data while still processing the old, keeping everything running smoothly and efficiently, without delays or blocks.

Depending on how you want to implement it, you might not need multithreading, but if you were to do so, note, that the Python’s Global Interpreter Lock (GIL) can be a limitation for CPU-bound tasks, as it allows only one thread to execute Python bytecode at a time. However, for I/O-bound tasks (like network data receiving or file reading/writing), threading can be beneficial.

Also note that. when using threads, data sharing between threads must be handled cautiously to avoid data corruption. Using thread-safe queues or locks where necessary is crucial.

If you plan to do so, your code could look something like:

gaze_url = device.direct_gaze_sensor().url
video_url = device.direct_world_sensor().url

queue_gaze = queue.Queue()
queue_video = queue.Queue()

gaze_thread = threading.Thread(target=thread_function, args=(gaze_url, queue_gaze, 'gaze', device))
video_thread = threading.Thread(target=thread_function, args=(video_url, queue_video, 'video', device))
processing_thread = threading.Thread(target=video_processing_thread, args=(queue_video, queue_gaze))

Oneplus 10 pro

We just bought them last year

Hi! I am trying to produce the same exports as in your tutorials, and specifically trying to export the gaze_positions.csv file. But what I get is only timestamp, gaze x and gaze y variables. How can I get the csv to include the index, confidence and the rest missing, like I see it in your sample_recording_v2? Thank you.

Hi @user-42cb18 👋 ! May I ask which eye tracker do you have? You can find Neon's data format here

Neon does not include confidence values, as it has a different approach to estimate gaze. Perhaps you are mixing up with Pupil Core's tutorials?

@user-42cb18 Pupil Cloud or Neon Player would def be the easiest path here, as you won't need to code anything (See this message https://discord.com/channels/285728493612957698/633564003846717444/1203972316896428053).

But I assume you would like to do other things over the frame image, and thus programmatically access both right? In that case, here was discussed a small snippet to do so https://discord.com/channels/285728493612957698/1047111711230009405/1187427456534196364

Hi @user-e74b04 ! Apologies, I missed your message. There is indeed an increased workload with the real time estimation of pupil size and eye state. While our latest Companion Device, the Motorola Edge 40 Pro, handles this increased workload without any issues, the One Plus might not achieve the same performance when streaming and computing all this data in real-time at 200Hz.

If you're observing overheating warnings, you can try lowering the gaze sampling rate in real time and accessing the 200Hz data later via the Cloud. You can do this in the app settings.

We know that you might want the 200Hz gaze data in real time, and we are working on an app update to include a toggle to disable pupillometry and eye state features if they're not needed, allowing you to maintain a 200Hz gaze rate with the One Plus device.

Being able to use Neon with a computer in the way you mention is not on the roadmap, but we hope the new app version will alleviate the issues!

Hello, I am having the same problem opening the recording data(replied). The same message is from the file neon_player.log(attached file). Its Ubuntu OS uses neon_player 4.1.3(latest). And when I drag the recording folder to the neon window, The Neon Player quits unexpectedly.

I used recording data, which can be downloaded from the Neon official web page.

Hi pupil team, I'm trying to extract the optic_flow_vector using your scripts. I was wondering whether it is supposed to work when I download one of my neon recordings and set the directory as 'rec_folder', and feed it to the functions you developed (e.g., "rec = Recording(rec_folder)" in line 96 in helpers). Thank you for your answer in advance!

Hi There, I had a quick question about analyzing AOI data for this product I ran a few trials for a study I am working on and I set the AOI points on your website. Once I downloaded the data for a specific image that a few viewers looked at I am noticing that for one specific viewer (participant), their AOI fixation times are identical for different AOI points For example, one person is looking at the nose and eyes of a participant and in the excel file it seems that the fixation times for that viewer for nose and eyes r identical Was wondering if someone could help me understand this better thank you!

Hey @user-20657f, just to help me understand the issue, when you visually inspect the recording of this participant, does the participant's gaze go to other areas of the face, or does it appear stuck?

But when I download the excel file to analyze the raw AOI data I can see that for nose, mouth, eyes the AOI fixation times are identical w

which would not make sense

Nope different fixation count

Okay, this means that this participant visited each AOI a different number of times. Now change the metric to average fixation duration, are the numbers the same?

Again, they are all differtent for av fix dration

Would you mind sharing the data, so that I can take a look? You can send the data to data@pupil-labs.com

Will send now. Thank you

Hello! I just want to make sure I'm sampling accurately. Each eye camera is at 200hz, IMU is at 110Hz and Scene Camera is at 30Hz?

Hi @user-0001be , these numbers are correct.

Yes. You'll probably want to experiment a bit with interpolation options, but linear interpolation might be a decent first try.

What kind of network are you connected to? Local dedicated router, hotspot, or work wifi shared with other people? May I ask what the processing is?

Are you using Matplotlib to show the data in real-time or something else? Work wifis that are shared with others will lead to increased transmission latencies and potentially dropped data, so we recommend against that. It is best to use a dedicated local access point that is only used for your Neon and computer, like a dedicated router or, yes, the hotspot capability of your desktop. A Bluetooth connection is not supported.

The hotspot on your desktop could potentially be fast enough and is worth a try.

You can also use the USB hub mentioned in this part of the documentation to connect via Ethernet cable to get even lower latency.

Correct, no connection to the internet is needed for usage of the real-time API. You just need a networked connection between Neon and the computer.

If you are using the simple API with PsychoPy, then the display of real-time values will be locked to the refresh rate of your monitor (e.g., 60Hz) and that will limit the rate at which you can receive data from Neon. My assumption is that this, in combination with work WiFi, is producing your results. If you want to receive the data at sensor rates while displaying the data in real-time, then you could consider switching to the async version of the API and using threads, but if you explain a bit more about what you want to do, then it might not be necessary to go through that effort.

@user-f43a29 @user-07e923 I'm trying the hotspot from my desktop now. So my desktop is receiving wifi and the output is hotspot. The Neon mobile is connected via WiFi to the desktop's hotspot. How can we make this work as they are not on the same network?

aiohttp.client_exceptions.ClientConnectorError: Cannot connect to host 172.26.218.238:8080 ssl:default [The remote computer refused the network connection]

Hi @user-e74b04 👋 ! I'm briefly stepping in for @user-d407c1 here. Soon, we will release an update where you can enable/disable eye state computation, which will resolve your overheating issues. Regarding the problematic recording, can you open a ticket in 🛟 troubleshooting ?

@user-f43a29 it works now sorry I forgot to update the ip address. Thanks for your advice! My sampling rate still does seem low even with hotspot. What I’m processing is actually if the head velocity along horizontal (gyro_z) is within a certain velocity? Like more than 150deg/s. Would this be taking up a lot of processing power?

When paired with PsychoPy. Then at the end I print out my sampling rate for IMU and Gaze respectively. Alright thanks! I think it’s my code and I really have to look into it.

Hello, I'm kinda new with the device and also the way to explore the data collected. I have so some basic questions is it the right place to ask for some basic help that I did find in the doc (maybe it in it but I don't see it or else)? Like I labelled the data and I want to try the enrichment reference image mapper but in my project subjects observe in all the videos 3 different thing and I don't really know how to separe them (3 diff project?) or there is some where a way to select in the project x video you want to have this reference image and this reference video and repeat this for each thing observed?

If I'm not in the right place to ask, don't hesitate to tell me ! My mistake 😵‍💫

Thank!

Hi! Currently, I am performing measurements with Neon glasses. However, I am experiencing some issues with the usb-c cable. Is there any way in which I can change the cable as it is connected to the tracking part? Hope to get some answers so I can continue testing!

It just that each of my video are recording an observation of one of those 3 thing (not mixed in one video), so if I understand well it is better to separate projects

Okay thank! I was thinking that label was made for this purpose but it also great like this

Labels are meant only for visually organising your workspace.

okay ! thank for precisions 🙂

is it possible to correct the offset online (in the cloud)?

okay thank for your answer

I found it but how to add the "corrected" video to the cloud then?

Hello, Is there a possibility to rent ET-glasses for a short period of time? It's just for private use and I can't afford to buy one for just 1 or 2 days of experiment.

Greetings Patrick

Hope I'm in the right channel to ask.

Hello! If we use the gaze offset correction in neon player, will this apply to other recordings of the same wearer as well or is it only applied to the single video?

Awesome thank you! Another question, when I try to download Neon Player I get a message that there is a corrupted image file when trying to open the downloaded file. Any thoughts on how to get around this?

is there a way to know the openess of the eyes? (in one csv maybe?)

Hello! We are using currently the neon pupil labs device, we wanted to take off the lenses to avoid extra brightness during an eye-tracking experiment. How can we take them off? thank you!

the basic one for neon "just act natural"

thank you very much, no, it is only to use them without the lenses, just for trying

Hi @user-07e923 , is it possible to use Neon locally, plugged into my notebook (with a pupil capture plugin)?

The main point is to avoid problems with LGPD in Europe.

The second point is plugin. What do you think would be the "way to go" for me? Would be great if I could use my screen tracker plugin with Neon.

I have adapted the surface tracker to track the screen, no need for fiducial markers.

Not so fast, but fast enough

And requires some light constraints too

Does the companion app works without internet connection?

So, I need the plugin to run into "Neon Capture"

Is it possible?

Hmm, what I can think of is to use the Real-Time API to get the live-feed of the scene camera. This will be limited by the frame rate of the camera (30 FPS). You'd still get data of your plugin via the scene camera.

Can you clarify it a little bit? Does Neon supports plugins?

I mean the Companion App

Would you mind making a custom version of the Companion app for me? I sent I PR some time ago with the plugin

Thank you

@user-41f1bf can you maybe explain what is your planned research and analysis? This can help us provide more precise feedback.

Hi @user-480f4c , I am teaching a pseudo-alphabet for people. Some data suggests that people look into some letters and not others. For example, the first two letters, or the first and the last, and so on. I am planning to record/analyse eye movement patterns and correlate them with participant's performance.

Tracking the screen is necessary for homographic transformations. I need data converted to screen space. (A secondary contribution is to make making choices easier, participants could just look into the screen and press a button, no need of mouse and keyboard and less noise.)

I don't want to use fiducial markers because the pseudo-alphabet is made of black-white shapes. Participants will tend to look at markers too often.

Hi Nadia, I am not allowed to upload personal information for your Cloud (It requires an authorization that I don't have). So, I need to do everything offline.

Unfortunately, we don't offer a markerless solution to detect surfaces in our offline application, Neon Player.

Thank for your service Nadia, I am afraid I will need to write my own data analysis solution.

Hi @user-41f1bf! If I may chime in here. If you already have a markerless screen tracking approach implemented, then frankly, it is conceptually possible to adapt that to work with Neon in an offline context.

Here are some points: - You need to log into Pupil Cloud once to set Neon up, but after that, you can effectively disable all Cloud uploads. Thus, no data ever has to go into the Cloud. - You can then communicate with Neon via our real-time API. You can send events to the phone, or you can stream data, in real-time, over the local network, as @user-07e923 mentioned. - Since both the scene video and gaze are streamed, you can build real-time screen-mapping applications, e.g. like this one we did that uses AprilTag markers, or even experiment with other markerless approaches, although these might best be processed post-hoc. - So, whilst we do have options in Cloud, it isn't strictly necessary, and a custom version of the app that has plugins really isn't needed – you can do most things with the real-time API.

Anyway, I hope this helps and is useful for others too!

So, is it possible to use my plugin with Neon Player?? Or to load my Neon Recording into Pupil Player??

Hi @user-41f1bf ! If your plugin works with Pupil Player, it will most probably work with Neon Player. The Plugin system works the same

Nice, thank you

May I ask you to kindly test it before I complete my purchase? I did a PR with it

https://github.com/pupil-labs/pupil/pull/2341

I known what I need, but I am not sure where to go from here. What do you think? Could you merge this PR into Neon Player? Or the best is just to load the Neon Recording into Pupil Player?

Just want to test how this plugin behaves with a Neon Recording. You need a white monitor screen in the field of view, inside a room with dimmed lights.

Oh.. sorry about that. The plugin is here https://github.com/cpicanco/capture_plugins/blob/master/screen_tracker_neon/ScreenTrackerOnline.py

So, as I said earlier, the plugin was made for pupil capture.

Pupil Player requires more changes for this to work due to the background task runner

@user-d407c1 , if you take a closer look, you will notice that the surface tracker does not exposes the marker stuff (called by the background task runner) for the plugin system, am I correct?

Hi, newby here, and honestly asking for a friend. I'm from Humanities, and our neuroscientist is currently unavailable, and we need an estimate of time required to calibrate Pupil Labs' 'Is this thing on - Neon eye tracking bundle' for our experiment. This is to cost the time for our funding application (deadline looming, of course...🙄).

So in our experiments, participants will sit at a desk and look at a book, to read one page at a time. They will have four pages to read, some might be from the same book, some from other books. Each book is set at a desk, and participants should move from one desk to another, depending on the books of their choice. Reading each page will take no more than 4 minutes, at the very most (each page has a small poem and a picture).

We are allowing 20 minutes per participant for the reading itself, but I need time allowance for the calibration process for each participant.

The equipment and recording etc will be operated by a very trained neuroscientist. It's just the time for the calibration that we need.

Good morning guys.

Quick question: Is the NEON module in "Just act natural" frame weather sealed? We need to use them in rain, and are worried about the durability in such conditions.

Kind regards Claus S. Hermann DanCrash ApS

okay, it will be very interesting for the project I'm working on! looking forward to see this!

hi I opened the Neon app and it displayed the firmware version update, but it crashed. May I ask what the reason is? Thank you

Hi @user-4bc389 , could you open a support ticket in 🛟 troubleshooting ? Then we will continue communication there

OK

Hello. I am using neon's realtime streaming abilities. I used under the hood documentation: https://pupil-labs-realtime-api.readthedocs.io/en/stable/guides/under-the-hood.html to connect using RTSP. Everything works fine now except Worn flag which indicates whether the user is wearing the device. The flag is always toggled whether the device weared or not. I wanted to know if it is a known issue or I'm doing something wrong. I previously used Pupil Invisible and they worked correctly.

is it possible to cut video in the cloud ? (shorten a little bit)

okay I'm not sure how to do this it is just my scanning video it's 3sec too long 🥲

hi, my neon app shows low storge ~30 mins left to record, can i ask how to solve that propblem? and why?

thank you Miguel, very helpful

hi, I was trying to add scanpath with https://docs.pupil-labs.com/alpha-lab/scanpath-rim/#generate-static-and-dynamic-scanpaths-with-reference-image-mapper but as I followed the steps an error occurs when I select the file (dowload of the ref img mapper) am I doing something wrong? the error is the following: raise Exception( Exception: Wrong folder! Please select a Reference Image Mapper export folder

Hi @user-bdc05d ! does your folder contain the following files:

required_files = [
        "fixations.csv",
        "gaze.csv",
        "reference_image.jpeg",
        "sections.csv",
        "enrichment_info.txt",
    ]

?

yes

there is 2 mores: aoi_fixations.csv and aoi_metrics.csv

Hello! Do you know where could I ask information regarding the extension of warranty?

self.img_height = self.img.shape[0] ^^^^^^^^^^^^^^ AttributeError: 'NoneType' object has no attribute 'shape'

Hello! I've been trying to connect my PC to the Pupil Labs Neon using the Python API. While I can connect to the devices through the app's user interface (as described here: https://docs.pupil-labs.com/neon/data-collection/monitor-app/), I'm encountering an issue with my Python code. The code doesn't throw any errors, but it seems to be continuously reading the device. I suspect the problem lies with the Device(address=ip, port="8080") line. Any insights or suggestions on how to resolve this would be greatly appreciated! Thanks in advance!

`# The two lines below are only needed to execute this code in a Jupyter Notebook import nest_asyncio nest_asyncio.apply()

from pupil_labs.realtime_api.simple import Device

ip = "192.168.1.16"

device = Device(address=ip, port="8080")`

little question: in aoi_metrics.csv and aoi_fixations.csv when I compare the total fixation time for a recording (given in metrics, and I do a sum for fixation with the correct recording id and area) I don't have the same time in result, am I mistaking something ?

like for a zone A in recording 1 if I look in metrics.csv I have a number that is not the same that the number obtain by the sum of zoneA x recording1 in fixations.csv (I'm talking for total fixations here)

@user-07e923 Thanks!

you can no longer stream information from the web browser after subsequent re-opening of the browser? This is correct. I use my laptop PC as the mobile hotspot.

@user-07e923 My version is 2.8.15-prod already.

@user-07e923 Thank you for your help! My procedures are follows: 1. Set up the laptop's mobile hotspot and connect the phone. 2. Launch the Neon companion app on the phone 3. Open browser on the laptop (Chrome or Edge) 4. Enter URL (neon.local:8080 or IP address) <- first time after reboot it works 5. Close the browser 6. Open the browser again 7. Enter URL (neon.local:8080 or IP address) <- does not work

Now, I tried to do another PC and another Network, but I have same results. Please let me know if you need more information.

@user-07e923 Thanks in advance. Different device: Yes Eduroam: No :8080 : Yes

I see this in both browser.

okay thank a lot 🙂 !

hello, Effectively I try some this morning and number are much closer (still experiencing diff of 1 or 3 unit of difference but it is much better), thank for the reactivity ! 🙂

but it is weird now I have average fixation duration that are often above 1.1sec (sometime I up 3sec 😵‍💫 ), this is way to long compare to the know 100-500ms, is that normal or maybe I am mistaking the meaning of average fixation duration?

on average, how much time scanpath generation take? because my tests take so long for the moment I'm a little bit surprise

Hi Pupil Labs, in Neon Player, when I try to render the scene video with a gaze overlay using the manual offset correction, the manually offset gaze marker looks fine in Neon Player, but is not reflected in the final render (i.e., is there but shows the original gaze marker). Am I doing something incorrectly?

during the adjustement of circle size (scanpath) is it suppose to see the image? I just have a red background with one green circle?

(red circle take all the image in fact)

oh okay thank, I have seen that but has the point green nearly disappeared I don't press n in off

@user-07e923 Thank you for your help. I want to remove the transparent lenses from my PupilNeon glasses. How can I do this?

Hi @user-5c56d0 ! Just to add on top of my colleague's response. Lenses are tightly fitted, especially with the heat sink version of the Just Act Natural frame, and there is a risk of chipping them during removal and refitting. We recommend exercising extra caution if you choose to proceed.

With that said, here is how you can remove them: 1. Important to note: Remove the module first to avoid any damage. 2. To pop the lenses out, apply light pressure with your thumbs from the inside to the outside, starting at the nose side while holding the other side of the lens. 3. If they resist, you can heat the bottom of the frame up to 120° C to make removal easier. Please be very careful while doing this.

The same information is also sent to Miguel.

I am using the Neon player (4.1.4) which works fine on my laptop. When I installed it on another computer (Win 10 server), it keeps crashing with the following error message:

[13:49:31] ERROR    player - launchables.player: Process Player crashed with trace:                       player.py:766
                    Traceback (most recent call last):
                      File "launchables\player.py", line 585, in player
                      File "plugin.py", line 412, in __init__
                      File "plugin.py", line 434, in add
                      File "audio_playback.py", line 96, in __init__
                      File "audio_playback.py", line 157, in _setup_output_audio
                      File "sounddevice.py", line 1494, in __init__
                      File "sounddevice.py", line 817, in __init__
                      File "sounddevice.py", line 2660, in _get_stream_parameters
                      File "sounddevice.py", line 569, in query_devices
                    sounddevice.PortAudioError: Error querying device -1

           INFO     player - launchables.player: Process shutting down.                                   player.py:768

This happens after dragging and droping a recording folder into the window, just before the new UI is opened (i.e. the processed files in the local neon_player folder are created). The computer does not have a soundcard installed, I assume that is the reason for the error. Would it be possible to fix the error?

Hi @user-5c56d0 ! I removed one of your messages since they are public, so no need to duplicate them. 😉

The model you have is the oldest one without a heatsink, so, it should be easier to remove the lenses. You can follow the same steps as described before https://discord.com/channels/285728493612957698/1047111711230009405/1245753344346161253