Hi Max,

Thanks for the update.

My best wishes for your project.

Thank you very much, philosophy-chemical, for the feedback and reply.

It was a pleasure to look into it and to see how much you have established.

My best wishes for your project

Hi AlexC,

How about this (press play)
CV4_2025_drs_25_SNse_02.c4d

Selections that are set in " " can be used, as long as the naming of the Selections is well done. Then you can use a Field and have the interactivity you might want.

All the best

Hi philosophy-chemical,

I think the Wiki page
https://en.wikipedia.org/wiki/Gaussian_beam
showcases what needs to happen, and why I see some problems just to shape it and have the "photon-density" equally inside the whole body.
The heat maps in the link show this clearly.

As light spreads, its intensity lowers, while the spread here follows the Gaussian formula. I think spread and intensity is a phenomenon that applies universally. This might introduce a complexity you don't need for a visualization that ends up being 2D. Just a guess.

All the best

Hi AlexC.,

Please have a look here:
CV4_2025_drs_25_SNse_01.c4d

Node 1 - Select
First I select polygon 3 and 4

Node 2 - Subdivide
Selection is set to "all"

Node 3 - Extrude
Selection is set to Default

Is that what you are looking for?

All the best

Thanks for the reply, philosophy-chemical.

I checked my library, e.g., the book "Light Science", Physics and the Visual Arts (2nd edition, Springer 2020), and I couldn't find any entry there. Hence, my question: when they leave it out, I need to know more about it. Thanks for your input.

Perhaps check directly with Redshift's Tech-Support, and please share your needs specifically, and ask if Nicolas Burtnyk has any input for you.

I know from little talks we had that Nicolas Burtnyk, Executive VP, Rendering
https://www.maxon.net/en/about-maxon/executive-leadership
has an extreme deep knowledge of light science. Perhaps that is a better place to engage for your search. His knowledge makes me feel, even after decades of studying Color and Light, that I have not reached anywhere near his knowledge.

My feeling is that it needs a lot of attention to get the spatial intensity of the light correct if a Gaussian distribution is complete applied, if you ask for a valid, scientifically sound representation. Especially when, as you mentioned, optical elements are in the setup.

If any render engine is able to recreate this, it is questionable, as all calculations are based on RGB channels, and not on frequencies, which is needed to get to that precision.

Light is complex, and the simulation in a render engine is typically an attempt to get as close as possible to a sound result, while keeping the render time reasonable. Light in nature does not have such a problem, a near infinite bounces are done in "real-time" naturally.

Fingers crossed you get an actionable answer.

Enjoy your project

Hi philosophy-chemical,

The question would be, why not use two Spotlights? What is the result that you are after?

If you need that, perhaps use the "Share Your idea" here as a request:
https://www.maxon.net/en/

I found an OSL Shader for Gaussian "Light" (Just by googling it, but I can't support this here, as we do not share code here, for security reasons. I used -- Gaussian light osl shader for Redshift? -- sometimes writing redshift instead of Redshift worked, as I checked if Safari or Chrome produces the same result)

Perhaps ask in the Redshift Forum about OSL
https://redshift.maxon.net/

I explored how far I could get with a Cloner set up, but that didn't yield the desired results.

My best wishes

Thank you very much, Harry, for the ticket.
I'm not running the Site Issue forum, and I have no access to customer data. We take privacy here pretty seriously, way above what the law asks us to do. I wouldn't want to have access even.
So, I do not check here often in the Site issue forum.

The simple answer is that generators, like the SDD, generate new Geometry, and that doesn't include any MoGraph ID.

An alternative could be: (color or ID, both works.)
CV4_2025_drs_25_RSci_01.c4d
Enjoy your weekend

Hi Max M.,

I'm happy to look into it, but I have nothing to look into right now.

Images may contain some information, but looking at an image to identify what's wrong insight a project setup can lead to a list of questions. Over the decades, it has become clear that this is not an effective way to work.

As usual, images are supportive sometimes, but I need a Project file.

An empty file, such as a disc, and the material will suffice, typically below 1MB, and can be attached here as a C4D file directly.

Cloud services used here include Dropbox, Google, WeTransfer, Adobe, and Apple. Files were uploaded without compression, and the URL was pasted directly into the text without an HTML wrapper. For security reasons, I do not touch any other options.

However, I would create those patterns in Photoshop; hence, I currently have no idea how parametric this can be.

All the best

Hi high-spite,

Thanks for the feedback. Like with textures, where the suggestion is to have, at least, 1.5 times the resolution, to not noticeably lower the quality if the render (the 1.5 here means if the camera comes close, that this ratio is given, as it is nearly impossible with shaped objects to match each pixel from texture to render. Similar to that is the size of an object, since we have a very small one, typically with Particles. Here, the particle has a shape, which needs to "translated into the final pixel raster.

Like with any pixel movement that deviates from the next pixel position, changes that are longer or shorter will lower the quality. Lens distortion correction is a type of movement that affects quality.

The post effects can only "see" the scene through the rendering (the raster of the resolution), without any knowledge of the actual scene.

The Nyquist Frequency is a good way to explain it. Living in a cave and emerging only every 24 hours will give the impression that it is always night (or day, etc.).

Another point is that objects, especially particles, move, plus the camera might add screen movement for them as well. So they move through that raster of the final image. A typical example is Analog audio sampled into digital, which can lead to many problems if the resolution is not high enough.

But I get carried away here. I think the idea is clear, and I apologize for going on about it, but perhaps a few readers might find this supportive to know.

Cheers

Hi Harry,

Please have a look here.
CV4_2025_drs_25_RSuc_01.c4d

Also, please open a ticket to get access. I can't make exceptions here. Thank you for understanding.

Enjoy your weekend

Hi high-spite,

Last night (I'm on Pacific Time, and it was way past midnight), I made a note to explore this deeper today.

Here is a project file with a variety of resolutions. Please render each out, while activating the little square in the Render Settings Names.

The intensity of the dots on the right is set to 1.0, and the one on the right is set to 5.0. Rendering is set to 32-bit/channel Open EXR to preserve this energy.

CV4_2025_drs_25_RSfx_11.c4d

Next, load it into Photoshop by going to File> Script > Load files into Stack.
Scale each layer to fit the 10 K frame. 200%, 400% and 1000% for the CMD+T transform.

The layer names should tell you which one is which.

The key takeaway is that the smallest "disc" is limited to sharing its energy as much as in the high-res images. My impression is that most of it is gone based on sampling (call it AA) to the neighbor pixels. With this, the Post effects have trouble getting the same information, while the stronger "bleeding in the (low-res) larger pixels compared to the particle (disc) object creates the effect that the particle is larger. While the match of particle and pixel is not predictable, simply adjusting size and Illumination weight might not yield the desired results.

In the image, you can see that effect for the 1K, occupying the neighbor pixels the most (the grid of pixels is the 10K base, and the images are scaled to match), the Bloom is lower for the 1K!) The dot is from the 0.5 cm disc with the 5.0 weight of the emission. When you look closer, the "pixel-blocks" show up in each, the largest in the 1K.

To get good results, Oversample (perhaps twice what you need), then scale down.

All the best

Hi fordmar,

The slow down is typically done to produce 30fps (29.97) with the same duration as the 24fps (23.976), perhaps 25fps for different TV specifications. Your idea sounds different.

Either way, slowing down is not as simple as it is often "sold". It works great on high-speed footage used at low fps, where things are already slower. Typically Motion analysis is applied. Even then, some things can and will typically cause problems. (Crossing items, transparency, or patterns on reflective objects.

The Time stretch in Ae also needs an adjustment in the Frame-Blending.

With your relatively fast-moving content, the problems of this technique become very apparent.

Please render the sequence of the file below. Since it is an image sequence, set the sequence to any FPS you start with in Ae. Leave the C4D project settings, so I can see more what your results are. Then set up your slowdown workflow.
I hope my project setup can effectively demonstrate what's happening.
CV4_2025_drs_25_RStr_01.c4d
The quality is set to low, so I can see noise and the rendering is fast. PNG is typically not my format, but you mentioned it.

Thanks for the video. When you watch the second part of your link, go frame by frame with the "," or "." key, and you'll see that it repeats A-B-C-D-D each time. Here is the reason for the "jitter" as you call it. Even if you use a blend A-B-C-D-"D/A", the movement will stop because there is no progress added on D/A.

Blend> Pixel Motion (Compare the refections, they don't match the texture, just to name one example)

To perfectly "time stretch" each frame of the new sequence needs to change, which is often not possible.

Do I get your problem so far? Please let me know. Then we can take a look at what would work better. The question of Motion Blur might be a key point to render a higher number of frames…

Chhers

Thank you very much for the feedback, atomician!

My best wishes for your project!

Hi high-spite,

Thanks for the file.

There is no comparison to an image that has 36 times more pixels to define the result.

You might explore the Filter and Filter Size.

https://help.maxon.net/c4d/2025/en-us/Default.htm#html/VPRSRENDERER-REDSHIFT_RENDERER_TAB_SAMPLING.html#REDSHIFT_RENDERER_UNIFIED_FILTER_SIZE

Explore which filter provides you with a better result.

Explore oversampling. Meaning using 9K and going down to 1.5K as defined below:
To do that in one step often harms the outcome.
If you need a single image only, render in 9K and then set the Image resolution in Photoshop twice: first to 4500, then to 1500 (Photoshop> Image Size> Resample Preserve Details 2.0). You may also use three or four steps to scale it down.

In the black frame the downsampled image from the download, as described above.

Another point is the limitation of 8-bit/channel, which doesn't allow good downsampling, as the energy that you might have is gone. Some of your particles have values way above 100% or 1.0. This energy is visually absorbed (based on less contrast) to a certain degree with Bloom and Strike.

You have Bloom and Streak enabled as post effects, both baked into the render and applied as 8-bit/channel only, which will surely contribute to the problem.

Consider adding that later, not on the render directly (baked in), and use 32-bit/channel float (OpenEXR, Tiff, or PSD can handle this.) Later go to any format you need. During production I suggest avoiding 8bit/channel strictly.

Let me know if that works for you.

Cheers

P.S. Here is a simplified element example, or a little cube would do, as super simplified object.
CV4_2025_drs_25_SItc_10.c4d

It is pretty much the scene from the documentation.

Cache it and exchange the clone with your chain clone. Perhaps the fast response will help to find the best mix of Iterations.

(Or copy the simple object and plant it into the documentation file as is.)

Hi charity-chance,

I would suggest using the Preferences: Import and switching to Standard Physical, while leaving terms like trash out of this forum. OK!? Thank you.
https://help.maxon.net/c4d/2025/en-us/Default.htm#html/PREFSIMEXPORT-PREF_IMEXPORT_MAIN_GROUP.html#PREF_IMEXPORT_NODESPACE

Preferences> Import> Target

I tested this with a current Mixamo FBX, and it works just fine.

Take care.

Hi atomician,

Thanks for the file.

Please try the Distance tip first, the Search radius, anything else please explore if needed.
My main target here was the jumping off element where the Loop starts/stops.
CV4_2025_drs_25_SItc_01.c4d

.
The Settings Attribute Manager> Object> Create Settings> Mode: Distance

The Rigid Body> Collision> Search Radius: lowering it to 0.05
.
.
The Attribute Manager> Mode> Scene Settings perhaps to something smaller than 100cm, something between 20-30 cm.

The Dynamic Shape to Auto is often not usable. Try triangulate first, if that is not working, then the Complex shape.

Use the Attribute Manager> Mode> Scene> Simulation> Scene> Use the Draw options to see what is going on.

I left the Cylinder in it, as I used this as a second test object, it can be deleted. The Spline.Rail: I typically have better results with a rail.

Use the Attribute Manager> Mode> Scene> Simulation> Simulation: increase the Iteration to 2, and the Smoothing to 2 as well. If all works, and the final simulation is good, explore on a copy if these can be lowered.
.

My tendency would be to use a super simple object as a chain element, and then clone something more complex on it later, but if triangulaate (shape) works, that can be ignored.

Cheers

Hi high-spite,

Without knowing the project settings,

Here is my short demo:

The resolution you provided shows a 1:10 ratio, which means for a 2D object that is a hundred times more pixels in the hi-res than in the low-res. (I rounded your 1460 to 1500 for clarity.)

The energy (light) that a single pixel has in the high-res version, if no post effects are applied, can't grow over one pixel. It creates one pixel.
The energy (light) that a single pixel has in the low-res version, if no post effects are applied, can't grow over one pixel. It creates one pixel.

However, in the low-res version, that pixel has an area that is a hundred times larger. It will show different if both images are shown side by side in the same size.

This is the short explanation. Anything else, I need the file, reduced to the problem.

If it is larger than 1MB, then it can't be attached here as a C4D.

Please use a cloud service, but for security reasons, I will touch only Wetransfer, Google, Dropbox, Apple, or Adobe. Please paste the full URL into the text, no HTML wrapper. The file uploaded as c4d, no compression, no zip, and surely no rar. Sorry for the long request.

All the best

Sorry, I checked on the Liquids, Gloria,

The Speed of the particle is pretty high. Perhaps start with 75.

Place a Collider Modifier (from the particle menu) under the particle Folder.

• In the Attribute Manager> Mode> Scene> Simulation> Particles> Substeps is set to 1.0, explore, e.g., 60, then dial down.

The Substeps check if there is a particle that might collide. If that number is higher, it checks more often.

https://help.maxon.net/c4d/2025/en-us/Default.htm#html/OPBDSCENE-PBDSCENE_FP_GROUP.html#PBDSCENE_FP_SUBSTEPS

I hope that helps. I don't have a fast enough computer to fine-tune this.

I call it a day now. (…past midnight)

My best wishes