When trying to render the animation it seems to create anew camera each time, meaning that I can't render from an existing 360 camera that I have added animation to. Am i right in thinking its impossible to do a moving camera render?

Render result has Jagged edge in blender 2.92.
But blender 2.83 works well.

My render result is giant when i watching it in skybox player on quest2.
I render it with 180degree.
I tried scaling model, putting camera closer to model, but no helps.

Nice to meet you! I was looking forward to this wonderful script. I downloaded it immediately and installed it in the Blender 2.8 preferences add-on, but eeVR Master is not displayed. I have been worried about it for a while, so I would like to know the specific method.

getting error while selecting oprtions....and I am using stable 2.80

Plugin disappeared in version 3.6

How to solve:
1-Open the plugin file "init.py";
2- On line 31, delete the text "support": "TESTING";
3- Save the file, and look again in the plugins list, it will be in the community tab!

when I render 360 image it is rendering 360 equirectangular format probably most web viewers are panoromic support only. Can you give any suggestions for 360 image viewers?

Blender version 2.93.1 LTS.

SBS mode works perfect.

But in top-bottom mode, top image and bottom image location is upside down.

Other people have experiencing same issue. You can find a commend saying that he used GIMP to fix upside down image.
: [(https://www.youtube.com/watch?v=ND8T5_VCcN0)]

I triggered an error message by typing a bogus location into the save folder location. I have since corrected the save location but the error message keeps coming up each time I try to render an animation and even with a single frame, the rendered output does save to the corrected location. Is there a way to clear the error warning?

I should add that I am running Blender 3.0 on W10.

Definitely, This script is helpful.

I'm new at VR and blender rendering.
Is there any recommendation setting for rendering to make 4K VR180/360 for Youtube, and recommendation ffmpeg command?
Please let me know..

There is sample of ffmpeg commands on Oculus blog, and it worked. But I'm not sure that it is good for eeVR(and eevee) and PNG files.
https://creator.oculus.com/blog/encoding-high-resolution-360-and-180-video-for-oculus-go/

Any suggestions would be help. Thanks!

when choosing aces shader, eevr seems to collapse

After enabling motion blur and going to render, blender will stop responding indefinitely. It doent even load anything into ram.

OS: win 10
CPU: ryzen 5 2600
Graphics card: 1080 TI
Graphics driver: 27.21.14.5206
Blender version: 2.90.1 (aslo doesnt work in 2.9 or 2.8 from what I can tell)

To make it work on my comp I had to rename gl_FragColor to fragColor and now it runs great!
Thank you for this code!

Btw, i have added dome support in my fork.

I get an error when trying to use this and render in Blender 4.0

Python: Traceback (most recent call last):
  File "C:\Program Files\Blender Foundation\Blender 4.0\4.0\scripts\addons\eeVR\__init__.py", line 57, in execute
    renderer.render_and_save()
  File "C:\Program Files\Blender Foundation\Blender 4.0\4.0\scripts\addons\eeVR\renderer.py", line 719, in render_and_save
    imageList, imageList2 = self.render_images()
  File "C:\Program Files\Blender Foundation\Blender 4.0\4.0\scripts\addons\eeVR\renderer.py", line 707, in render_images
    imgl, imgr = self.render_image(direction)
  File "C:\Program Files\Blender Foundation\Blender 4.0\4.0\scripts\addons\eeVR\renderer.py", line 674, in render_image
    bpy.ops.render.render(write_still=True)
  File "C:\Program Files\Blender Foundation\Blender 4.0\4.0\scripts\modules\bpy\ops.py", line 109, in __call__
    ret = _op_call(self.idname_py(), kw)
RuntimeError: Error: Cannot write a single file with an animation format selected

What am I doing wrong!?

Are there any possibility of adding a feature to render out Fisheye Equidistant and Fisheye Equisolid too?

Using 3.3.1

I ran a basic setup in a new scene just to keep things simple, set resolution to 1920 x 1920, output to ONG RGB 16bit and hit render image. This is the error message I get. I also saved the file into it's own folder on a local drive so the images would have a contained place to go.

Python: Traceback (most recent call last):
File "C:\Users\EKUltra\AppData\Roaming\Blender Foundation\Blender\3.3\scripts\addons\eeVR-master_init_.py", line 60, in execute
renderer.clean_up(context)
File "C:\Users\EKUltra\AppData\Roaming\Blender Foundation\Blender\3.3\scripts\addons\eeVR-master\renderer.py", line 592, in clean_up
if not context.preferences.addons['eeVR'].preferences.remain_temporaries:
KeyError: 'bpy_prop_collection[key]: key "eeVR" not found'

error log:

Traceback (most recent call last):
  File "C:\Users\ilia\AppData\Roaming\Blender Foundation\Blender\2.82\scripts\addons\VRRenderer.py", line 650, in execute
    renderer.render_and_save() 
  File "C:\Users\ilia\AppData\Roaming\Blender Foundation\Blender\2.82\scripts\addons\VRRenderer.py", line 588, in render_and_save
    imageList, imageList2 = self.render_images()
  File "C:\Users\ilia\AppData\Roaming\Blender Foundation\Blender\2.82\scripts\addons\VRRenderer.py", line 566, in render_images
    img1, img2 = self.render_image(direction)
  File "C:\Users\ilia\AppData\Roaming\Blender Foundation\Blender\2.82\scripts\addons\VRRenderer.py", line 472, in render_image
    tmp = bpy.data.scenes['Scene'].render.filepath
KeyError: 'bpy_prop_collection[key]: key "Scene" not found'

location: <unknown location>:-1

I am trying to render my scene here, but the stitch is not working correctly. Maybe I have camera setup wrong.
Can you share a sample blend file that is working with VR180 SBS output ?
Thanks you

Whenever stereoscopy is enabled the render gets saved to a file with a correct extension (be it JPEG, TIF, EXR etc.), however the saved file itself is always a PNG underneath (confirmed by looking at the file's header in Notepad++).
This happens regardless of whether I set the stereoscopy to be in side-by-side format or individual - the file created is always side-by-side and is always a PNG underneath.

When stereoscopy is disabled the output is okay - the file created is a real file of the chosen type.
There is a small secondary issue specifically with OpenEXR output - the color depth always changes to Float (Full) upon rendering with the addon. Would be nice to see this fixed as files with Float (Full) take A LOT more disk space while providing no actual benefit (Full is pretty much only useful for things like displacement maps. For regular renders there is no difference, even when pixel peeping while doing extreme color grading).

System information:

• OS: Windows 10
• Blender 3.6
• Graphics card: RTX4070 and RTX3090

For now I was able to circumvent the issue by just rendering the left and right views separately with two cameras physically spaced apart, but would be nice to be able to use stereoscopy to do this automatically with a single camera (it could even save the views separately if there's no way to make it work otherwise).

This script is really impressive!
I'm new at VR. I was just wondering.
I used Google Colab to render animation. But, it seems that eeVR doesn't use GPU so much. At "nvidia-smi" command, GPU use-rates are between 0 - 6%.

Am I doing something wrong?
or Is it expected behavior?

To Reproduce
Steps to reproduce the behavior:

1. Install blender-2.83.3-linux64 to Ubuntu 18.04.3 in Google Colab through VNC viewer
2. "vglrun ./blender"
3. start rendering animation with eeVR ( output resolution setting is 2048x2048 for VR180 )

Expected behavior
I hope it uses GPU near 100%.
( Normal eevee used over 60% )

System information (please complete the following information):

• Platform: Google Colab
• OS: Ubuntu 18.04.3
• Processor: Intel(R) Xeon(R) CPU @ 2.30GHz
• Graphics card: Tesla P100-PCIE-16GB
• Blender version: blender-2.83.3-linux64

ja lancaram uma nova versão ???????????????

Hello,
I have the camera inside cube with volumetric shader. When I use plain even distributed volume then everything is ok, but when I use cloud texture on it, every picture has different level of transparency. Final image is stitched from vastly different images. It looks like the different angle makes the volume look different.

Thank you for your help.

Hello,

Whenever I click "Render Animation", this error box pops-up and doesn't allow me to render the animation. What am I doing wrong?

EDIT: I've fixed it by using the correct FOV. However, now it just crashes during the first frame.

I've been trying to render a 360 with Pure Sky to use enviroment in a 3D scene but in the eeVR render the volumetrics don't show up.

When rendering animation with film set to transparent, background is rendered as black rather than transparent.

Describe the bug
Somehow rendering an animation is not working

Console log (if an error message is produced)
RuntimeError: Error: Image 'Render Result' does not have any image data

Describe the bug
This is kind of TL;DR problem (didn't even try read first sentences from REAME.md).
I just grabbed ZIP and tries to install it under Blender 2.91 addons and assumed that everything would be ok.
But as expected nothings showed up.
After that I little bit reorganized VRRenderer.py register- and unregister-method to new and shiny init.py and after that I was happy camper as it showed up on addons/testing and worked as expected

To Reproduce
Steps to reproduce the behavior:

1. Go to GIT repo down load master ZIP
2. Try to install with install from Blender
3. Nothing happens (there is no such addon)

Expected behavior
Downloading and installing ZIP or pulling git down to addons dir should work out-of-the box

System information (please complete the following information):

• OS: Linux
• Blender version: 2.91

Additional context
Just asking is this something that is wanted as PR as I did the needed changes?

Python: Traceback (most recent call last):
File "C:\Users\kinimod\AppData\Roaming\Blender Foundation\Blender\3.0\scripts\addons\eeVR-master\VRRenderer.py", line 711, in execute
renderer.render_and_save()
File "C:\Users\kinimod\AppData\Roaming\Blender Foundation\Blender\3.0\scripts\addons\eeVR-master\VRRenderer.py", line 638, in render_and_save
imageList, imageList2 = self.render_images()
File "C:\Users\kinimod\AppData\Roaming\Blender Foundation\Blender\3.0\scripts\addons\eeVR-master\VRRenderer.py", line 616, in render_images
img1, img2 = self.render_image(direction)
File "C:\Users\kinimod\AppData\Roaming\Blender Foundation\Blender\3.0\scripts\addons\eeVR-master\VRRenderer.py", line 588, in render_image
bpy.ops.render.render(write_still=True)
File "C:\Program Files\Blender Foundation\Blender 3.00\3.0\scripts\modules\bpy\ops.py", line 132, in call
ret = _op_call(self.idname_py(), None, kw)
RuntimeError: Error: Cannot write a single file with an animation format selected

location: :-1

it just doesn't render anything, the format is set to png and it doesnt give me any errors, there is just no output. I set the file location to a folder and it doesn't go there. My blender is 3.4.1.

Hi and thank you for this script.. It's quite helpful. I've only tested it for a few minutes and it works good, but I have a few suggestions.

1. Moving the panel from tools to a proper tab (bl_region_type = 'UI')

2. It would be nice if the rendered images would get the proper frame numbers instead of starting from number 1. I changed to this:
   image_name = "frame{:06d}.png".format(bpy.context.scene.frame_current)

3. I got an error when I had "ffmpeg video" selected. Is there no way to render directly into a video? No problem using ffmpeg to do it after render ofcourse like I would in production but it's one more step if it's just for a quick test.

4. The first time I pressed "render image" Blender (2.82 final) just crashed. Unfortunately I have no more information about this.

5. Some kind of feedback about what's happening while rendering. Right now it's like Blender just locks up. I can see it saving tmp-images in the console but there's no progress information.

6. Sometimes it renders images to 100x200 pixels only. Seems like it's resetting the output settings.

Unfortunately the plugin seems not to work on Blender 3.5 and also 3.4. Or is due to the Apple Silicon Architecture? The scene is nothing special just the blank default one.

`bpy.ops.object.camera_add(enter_editmode=False, align='VIEW', location=(0, 0, 0), rotation=(1.47306, -2.7808e-08, -1.41023), scale=(1, 1, 1))

Python: Traceback (most recent call last):
File "/Users/xxx/Library/Application Support/Blender/3.5/scripts/addons/eeVR-master/init.py", line 55, in execute
renderer = Renderer(context, False)
File "/Users/xxx/Library/Application Support/Blender/3.5/scripts/addons/eeVR-master/renderer.py", line 371, in init
self.shader = gpu.types.GPUShader(vertex_shader, frag_shader)

Exception: Shader Compile Error, see console for more details`

Samples are set at 64 samples but output is the same as when it is set to 1 sample. If I use the Blender render image for the same situation it works as expected:

Describe the bug:
For the image rendered using eeVR, I noticed it is a bit blurry or out of focus. I have no idea why the image quality has been decreased. All the view resolution in the eeVR setting has been set to 100%. Without using the eeVR addon, the image rendered using EEVEE will not be blurry.

The way I found to neglect the effect is to scale up the image resolution. From 4k to 12k, but this also increases render time. Which defeats the purpose of using EEVEE to render an animation.

I have included a Google Drive for the Blender files for you to see which settings should and shouldn't turn on; one is using the eeVR addon, and the other is using Cycle.
https://drive.google.com/drive/folders/1PpHxftQ1eO7Ql9sUyBaNmlJQmUDjKfuu?usp=sharing

To Reproduce:
Steps to reproduce the behavior:

1. Set EEVEE as the default engine
2. Set render image resolution to 4096 x 2048
3. Set Output as PNG with RGBA, Color Depth 8, Compression 0%
4. Set eeVR all view resolution to 100%
5. Hit the render images in eeVR.

Expected behavior:
It will produce a blurry image compared to the cycle.

Screenshots

The right one uses Cycle, and the render is much clearer than the left one, which uses the EEVEE with eeVR.

System information:

• Windows 10
• Processor: i7 12700F
• Graphics card: AMD Radeon 6800xt
• Graphics driver: 31.0.24002.92
• Blender version: 4.0.2

Whenever I'm trying to render I get a permission denied error.

Python: Traceback (most recent call last):

File "C:\Users\suzanne\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\eeVR-master\__init__.py", line 60, in execute
  renderer.render_and_save()
File "C:\Users\suzanne\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\eeVR-master\renderer.py", line 748, in render_and_save
  imageList, imageList2 = self.render_images()
File "C:\Users\suzanne\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\eeVR-master\renderer.py", line 736, in render_images
  imgl, imgr = self.render_image(direction)
File "C:\Users\suzanne\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\eeVR-master\renderer.py", line 676, in render_image
  bpy.ops.render.render(write_still=True)
File "C:\Program Files\Blender Foundation\Blender 4.0\4.0\scripts\modules\bpy\ops.py", line 109, in __call__
  ret = _op_call(self.idname_py(), kw)
RuntimeError: Error: Render error (Permission denied) cannot save: 'C:\temp_img_store_20508_back.tga'

Any solutions?

System information (please complete the following information):

• OS: Windows 11
• Processor: AMD Ryzen 9 6800HS
• Graphics card: AMD Radeon RX 6700S
• Graphics driver: 23.12.1
• Blender version: 4.0.2

Hi!
thank you very much for this awesome plugin, however there this huge issue that i experience..

https://imgur.com/a/Qe1DtlD
please look at the picture i tried to remove the Screen space reflection, i also did the overscan.. but nothing change i still experience this boring thing .

any help please its really frustrating

thank you :)

Color management settings seem to be applied to both the initial renders and the composited/merged/stitched together final render.

To Reproduce

1. Set up a scene and adjust the exposure and gamma on the color management properties. The more extreme you make these initial values, the more intense the result, as these settings seem to be applying twice.
2. Through the eeVR tab, render your scene.
3. Go to the folder where eeVR is rendering out your images (same as where project file is saved). You can see how the color management is applied correctly the first time, as the renders for each perspective appear right.
4. However, the color management is drastically off in the final render.

Expected behavior
Ideally, the color management settings should only be applied during the initial renders, and not in the final composite.

I have several semi-transparent objects set on the alpha blend. And there are parts of the image (middle left and right seams) where the render just clips out part of the object. When I switch to "alpha hashed" everything is normal. It started to happen recently, before it was fine. When I import models to a new file the problem stays.

The shape is always the same as if there is some object masking it. I tried like every combination of settings I could think of and nothing changes. Besides the alpha hashed, which unfortunately adds a massive amount of time in rendering.

Thank you for your help.

This is an FYI post as it is not a bug in this addon but in Blender.

If you are rendering in stereoscopic view with soft shadows activated in the scene, you will notice that in some cases the soft shadows are applied correctly in the left camera but not on the right. This will cause visual corruption issues when viewing in VR.

As a workaround, disable stereoscopy in the scene and create two separate left/right cameras separated by the local X distance of the desired IPD and render each camera view separately using eeVR. You can use another software to merge the left and right camera renders together, I used ImagicMagik as it is easy to script.

Someone else has reported this issue in Blender bug report and I verified that it still exists in 2.93: https://developer.blender.org/T84009

Current error log on a fresh Blender 3.2.1 file with Stereo3D enabled, png file is set, panoramic 90degree camera. I used to be able to render in 3.0, but it went back to giving me the same error. Even on a new Windows OS install after formatting drive.

Python: Traceback (most recent call last):
File "C:\Users\xxx\AppData\Roaming\Blender Foundation\Blender\3.2\scripts\addons\eeVR-master\VRRenderer.py", line 703, in execute
renderer.render_and_save()
File "C:\Users\xxx\AppData\Roaming\Blender Foundation\Blender\3.2\scripts\addons\eeVR-master\VRRenderer.py", line 636, in render_and_save
imageList, imageList2 = self.render_images()
File "C:\Users\xxx\AppData\Roaming\Blender Foundation\Blender\3.2\scripts\addons\eeVR-master\VRRenderer.py", line 613, in render_images
self.set_camera_direction(direction)
File "C:\Users\xxx\AppData\Roaming\Blender Foundation\Blender\3.2\scripts\addons\eeVR-master\VRRenderer.py", line 480, in set_camera_direction
self.scene.render.resolution_x = self.camera_shift[direction][2]
TypeError: bpy_struct: item.attr = val: RenderSettings.resolution_x expected an int type, not float

Describe the bug
Blender needs to be run as administrator. If run normally, the addon produces the error
RuntimeError: Error: Render error (Invalid argument) cannot save: 'C:\temp_img_store_28300_top.png'

The addon is trying to save images to a location Blender doesn't have permission to access, namely C:\. If Blender is run as administrator, it works correctly, with the final image written to the location I set in the Output panel. The problem is that I can't choose where the intermediate images, the ones with suffix _top, _right, etc., get saved.

To Reproduce
Steps to reproduce the behavior:

1. Install and enable eeVR addon.
2. Open a new Blender file and save.
3. Set render engine to Eevee.
4. Render Image in eeVR panel.
5. See error in console.

Expected behavior
Image saved correctly.

System information (please complete the following information):

• OS: Windows 10
• Blender version: 4.0.2

Describe the bug
If the camera is inside an object, the resulting image seems to be incorrect. Instead of stitching together all different sides of the object, the front side seems to be duplicated in all four directions. If the object has holes, the landascape shown through the holes does seem to be correct (not a duplicate of just one direction).

To Reproduce
Steps to reproduce the behavior:
Create a cube. Put the camera inside it. Cut some holes to the cube and make some differences to the four sides of it so that you can recognize which side is which. Render with eeVR. The resulting images will show only one side of the cube.

Expected behavior
I'm expecting the render to show all four walls of the object I'm inside, not just one wall duplicated.

Screenshots

System information (please complete the following information):

• OS: Windows 10
• Processor: Many different tried
• Graphics card: RTX2080
• Graphics driver:
• Blender version: 2.9

Hi! Your script is absolutely amazing and it is exactly what I searched for a long time. I know it may be difficult, but side-by-side format is more popular than top-bottom in 180° videos. I'm zero at scripts and can't rewrite it myself, can you please add this feature?

Also, I may be wrong but image renders with wrong eyes position (bottom should be for right eye and top for left eye)

And you posted a link to "VR Media Player - 360° Viewer" app on blenderartists, I used a lot of vr players including this one, and I do reccomend to try out "VRTV VR". It's fully customizable and gives best perfomance in both 180° and 360° videos

First, awesome script!

There are 2 key issues I'd like to mention:
1, when using camera motion(action), the rendering result will just be like a picture mixed with 6 same pictures. I think it's because of the keyframes. When using keyframes on Camera, one cannot really change the direction of the camera without inserting a new keyframe, and you will get 6 same pics from the 'front' view.
I hoped to fix this problem as following:
1>add a new camera obj, with the same parameters of the old one;
2>activate it;
3>using the new one to render scene;
4>remove it;
5>activate the old one.
but, I cannot find an API to do the 2nd step. Most likely to do this is "bpy.ops.view3d.object_as_camera()", but when executing, there will be a context error.

2. when the rotation Y of the camera is not zero, the rendering result will be distorted. I think it's because of the find_direction_offsets() method treats the Y value as zero by default.

By the way, the output file format should not be TIFF, the compositing node should not be used, or it will crash.

Thank you!