How to Load Your Own Data on Meta Quest

Updated March 2026

A guide for ConfocalVR & AstroCubeVR running on Meta Quest.

Supported File Formats

App	Formats
ConfocalVR	TIFF stacks, NIfTI (.nii), DICOM (including JPEG 2000 compressed), folders of PNG or JPG image sequences, and ZIP containing any of these
AstroCubeVR	FITS (.fits), and ZIP containing FITS

Step 1: Upload Your Data to Cloud Storage

From your desktop computer, upload your dataset folder to a cloud storage service such as Google Drive, Dropbox, or OneDrive. If your dataset consists of multiple files (e.g., a TIFF stack or DICOM series), place them in a single folder.

Step 2: Download to the Quest Headset

Put on your Quest headset and open the Meta Quest Browser app
Navigate to your cloud storage service and sign in
Locate your dataset folder and start the download — the service may compress it into a .zip file automatically
The file will download to the Quest’s Download folder

Step 3: Open in the App

Launch ConfocalVR (or AstroCubeVR)
In the file browser, use the left-hand directory selector to navigate to Download
Select your file or .zip — the app will extract and load it automatically
The dataset is also stored in the app’s local storage, so it will appear directly in the app’s directory on future launches

Tips

Voxel dimensions matter. Quest headsets have limited memory. The maximum supported volume is 512 × 512 × 256 voxels — this is the binding constraint, not file size alone. A dataset that fits within these dimensions loads as-is; one that exceeds them is automatically downsampled. The ~500 MB memory budget is a rough guide, but always check your voxel dimensions.
Auto-downsampling is on by default. If your dataset exceeds Quest limits, the app applies Gaussian filtering to fit it into memory and notifies you in-app. For full-resolution viewing, a desktop version is available — see immsci.com/volumeVR.
Crop before loading. If your data is much larger than the region of interest, crop it first in ImageJ/FIJI. The voxel budget goes further on the part that matters.
ZIP files are supported. If your cloud service compresses folders before downloading, that’s fine — the app handles .zip extraction automatically.
Google Drive works well for sharing datasets between desktop and headset. Upload once, download from the Quest browser.

Getting the Best View

Isolate one channel first. Use the Visible buttons on the Channels Panel to enable only one channel at a time — much easier to tune without interference from the others.
Set focus on the channel you’re adjusting. Make sure the Focus checkbox is checked on the active channel.
Cut background noise with Min Rescale. Adjust the Min Rescale slider upward until background noise disappears, leaving only the signal you care about.
Min Rescale or Min Threshold? Both discard everything below where you set them, so either will clear background noise. The difference is what happens to what’s left: Min Rescale restretches the surviving signal across the full brightness range, which is why the image looks clearer, while Min Threshold leaves the values exactly as they came off the microscope. Use Min Rescale for exploring — it gives the best picture. Use Min Threshold if you’re going to measure intensities, so the numbers stay as recorded.
Repeat for each channel, then turn all channels back on to see the full overlay.
Adjust depth and quality. Zoom Depth controls the sense of depth in the volume. Render Quality sharpens the image — but set it conservatively for large datasets, as high values can cause visual stuttering on the headset.

Preparing Your Images (ConfocalVR)

Most data needs no preparation. ConfocalVR reads multi-channel TIFF stacks and separates the channels for you, each with its own brightness, threshold and color controls. Color (RGB) images are split into three independent channels automatically. A folder of single-slice images is assembled into a stack. In most cases you can upload your file exactly as your microscope saved it.

Use ImageJ/FIJI first in these cases:

Your microscope saves a proprietary format. Open it with the Bio-Formats plugin, then Save As → TIFF, NIfTI or DICOM.
You have a color DICOM or NIfTI and want the red, green and blue channels separately. These two formats load as a single combined color channel. Save as TIFF instead and the three channels are separated automatically.
You have a time series. Save each timepoint as its own file, all in one folder. They load as separate channels and can be played back as a short movie.
You have more than 30 channels. Thirty is the maximum; reduce to the ones you need.
Your folder holds several different datasets, or several multi-channel files. Give each dataset its own folder — one dataset per folder loads most reliably.

A note on file types in a folder: ConfocalVR loads one type at a time, so avoid mixing TIFF and NIfTI in the same folder. Compressed NIfTI (.nii.gz) should be uncompressed first.

Channel descriptions (optional)

Put a text file named ChannelDescriptions.txt in the same folder as your data and its contents appear on the Channel Info panel in VR. Use it to record what each channel shows, and to credit whoever produced the data — useful when sharing a dataset with colleagues or students.

Saving Your Work

ConfocalVR saves your settings next to your data, not inside the app. Tune a dataset once and it comes back the way you left it.

To save

Press Save VR Scene on the Settings panel.

To get it back

Just open the same folder again. Your settings are reapplied automatically — you do not need to press anything. Load VR Scene is there for when you have changed things during a session and want to go back to what you last saved.

What gets written

Alongside a dataset in MyStudy/DAPI.tif, you will find:

File	What it holds
`DAPI.ConfocalVRch0.txt`	Sliders for that one channel — brightness, contrast, opacity, thresholds, colour weights. One file per channel
`ConfocalVR_GlobalSettings.txt`	Zoom, depth scale, render quality
`ConfocalVR_VRSceneSettings.txt`	The rest of the session — which channels are visible, blend settings, the excluder, reference axes, histogram state

They are ordinary text files. Copy the folder to a colleague and your setup travels with it.

Tips

Keep the settings files with the data. They are matched to your data by filename, so if you rename or move an image file without bringing its .txt file along, its settings are lost.
Saving overwrites the previous save without asking. There is no undo.
Working from DICOM? ConfocalVR converts DICOM into a faster internal format before displaying it, but your settings files are still written beside your original DICOM folder, which is where you would expect to find them.
Distance measurements are the exception. On Quest they are saved into the app’s own storage rather than next to your data. If you need the file off the headset, email support@immsci.com.
Panel positions are remembered per headset, not per dataset. Arrange the panels how you like them once and they stay that way.

Moving and Sizing the Volume

Everything here uses the grip buttons — the ones under your middle fingers — not the triggers. Triggers are for pressing buttons on the panels.

One hand

Squeeze the grip with the pointer on the volume to grab it. Move your hand and the volume comes with you; turn your wrist and it turns.

Two hands — the one worth learning

Squeeze both grips at the same time and the volume follows both hands together:

Move it — move both hands in the same direction.
Rotate it — turn one hand around the other, as though holding a steering wheel.
Resize it — pull your hands apart to make it bigger, bring them together to shrink it.

This is the fastest way to get a specimen into a comfortable position, and it is how most people end up navigating once they are used to it. Depth scaling follows the zoom, so the volume keeps its proportions as it grows.

Moving yourself

Push the right thumbstick to fly. You travel in the direction the right controller is pointing, including up and down, so aim where you want to go and push forward.

Panels

The Channels, Sliders, Analytics and Channel Info panels can all be grabbed with the grip and moved wherever suits you. Their positions are remembered between sessions.

Working Together

Two to four people can share a session and examine the same specimen at the same time, from anywhere.

Before you start

Everyone needs their own copy of the data on their own headset. The specimen itself is not sent over the network — each headset renders its own copy, and what gets shared is everything you do to it. So send the dataset round beforehand and have everyone put it in place. It is the one step people miss.

To start or join a session

Press Start Collaboration.
A keyboard appears. Type a room name — anything, as long as everyone uses the same one.
Press Enter Room.
The status text changes from Not Collaborating to In VR Room: yourname.

Everyone joining types that same room name. There is no browsable list of sessions, so agree the name over email or chat first. Maximum four people per room.

To leave, press End Collaboration.

What everyone sees

Once you are in, the session is genuinely shared. All of this is synchronised:

The position, rotation and size of the volume
All channel sliders — brightness, threshold, opacity and the rest
Which channels are visible, and channel blending
The excluder box
Distance measurements and count markers
Each person’s head and hands, as an avatar
Voice chat, which is on as soon as you join — just talk

So when you grab the specimen and rotate it, everyone else sees it rotate. When you adjust a threshold, everyone’s view changes with it. Point at something with your controller and the others can see where you are pointing.

What is not shared is where each person is standing. You each move around independently and see each other’s avatars, which is what makes pointing at things work.

Tips

Load the data before you join, not after. Everyone should have the same folder open.
First time on a headset, Quest will ask for microphone permission. Accept it or nobody will hear you.
Anyone can drive. There is no host or presenter mode — whoever touches a slider changes it for everybody. This works fine with colleagues and less well with a class of twenty.

Preparing Your Data (AstroCubeVR)

AstroCubeVR reads standard FITS files natively. Most radio astronomy pipelines (CASA, CARTA) export FITS directly — no conversion needed. Upload your .fits file or a folder of .fits files to cloud storage and follow the steps above.

If your FITS cube includes WCS headers, AstroCubeVR uses them automatically for coordinate display and voxel scaling. Velocity axis conversion (Hz → km/s) is handled via the RESTFRQ header when present.

Need Help?

Acknowledgements: StarGateVRTM was developed by Immersive Science LLC, and refined through a ongoing discussions with Rochester Institute of Technology.