Artificial Sunlight (official topic)

@dapperdanman Well retro-reflective surface don't perfectly return light, they just return a main cone back, and a few fainter cones in stranger anglers. If the incoming light is somewhat collimated, both cones could maybe cancel a bit.

Will post some pictures here when I get the time.

Stranger "anglers" - yeah, you have to watch out for those shifty blokes with large rods. 🙂

I've made worse typos...

@fhr interested to see it!

@pancooks sounds like a good find. I notice some of the Ali sellers also sell a variety of other brightness enhancement films that are significantly cheaper - unsure how their performance compares.

Do you have any sources for larger sheets? I'm looking to cover approx 0.6mx1m. Also curious how you are mounting them

@dapperdanman When I was looking I concluded the difference in the 3M and the other ones are that the 3M is dual layer / both X and Y prisms stacked together, and the other ones are just a single layer. 

I haven't found any good sources for larger sheets. I've been using the 30cm wide rolls in 120x30cm light boxes, with an additional two layers of the milky screen printing film to separate the 4500K light source into 4000K (for the direct light) and 5000k (for the scattered light). Looks quite natural! 

I haven't yet found a good way of mounting the film, simply putting it in the film stack of the LED panels I'm using, but it bulges a bit if I accidentally bend the flimsy panel when moving it. Not ideal. 

I'm planning a ceiling mounted mirror bounced ETC Source Four as another experiment, where I plan to place the film (just the scattering film) in a glass sandwich. I'll post an update on how that works out.

@pancooks sounds like the 3M is maybe worth springing for then, even if it's a bit pricey. Do you notice the light projecting more into the room when you use it?

I also have some LGT075J diffusion film I was going to use, maybe I'll run an additional layer of brightness enhancement film over the top. The diffusion film already has a light transmission of only 60% or so, so I'm keen not to lose too much additional light. Diffusing light efficiently (high transmission) seems to be a harder problem than I originally thought.

I was going to clamp the film in at the edges of the wooden frame with some aluminium strip. Not sure yet how well that will work.

Would love to see some photos of your setup if you get the chance. I'll also post some of my work in progress.

@dapperdanman Yeah, it throws light well! Here's a comparison I made of one of the 120x30cm panels aimed at the ceiling with and without the 3M BEF film:

A problem I did run into is that I have quite a large living room + kitchen space, and illuminating it with 120x30cm panels very realistically makes it look like I have a huge space with only a couple of tiny windows, even though the light coming from it is quite pleasant. (In the panels I'm using 98+ CRI Seoul Semiconductor LED strips.)

I haven't found a way around that yet, and am only using them in my smaller bedroom at the moment. Regarding the light loss, I ended up running mine at half the max brightness, as too much light with too little surface area was quite blinding. I'm now still running them at around 150W/m². I'm hoping to make a projection based artificial sun for the living room later, working with mirrors to increase the effective surface area, as I feel more surface area is my next point of improvement. 

Posted by: @nolo

↑

...

The LED head I bought also costed 220 euros (Nanlite Forza 60B) and has a CCT range from 2700K to 6500K. 

...

I agree that the use of a finished light source (heat sink, fan, color and brightness adjustment) and projection lenses make the light generation much simpler and i also wonder if building this myself is that much cheaper. I found the rollei (called colbor in other regions of the world) candela 60W or 100W bi-color with a optical spot to be even more affordable.

But there is one caveat with these - how do you control the light color during the day? Yes, there are apps an you can do it via bluetooth from the app. But how do you automate the light changing during a day (and maybe also switching off the light when no one is there as it consumes quiet some power)?

I think the standard in the video lighting is DMX but then again we are talking about a different price range. This might warant doing your own led controller and maybe only re-use the housing and optics of the professional studio lights.

@shavings I’d love to have 1 m of the film and live in Germany. You can reach me at marcus @ 4xb . de …

I found this one that has DMX, is only 40W, but has 3 lenses for 5°, 10° or 25° based on distance to mirror: https://www.thomann.de/de/cameo_q_spot_40_tw_black.htm . It can be remotely (e.g. from home assistant) dimmed and can also change its color during the day.

I guess what we are looking for is a "spot" with possibility to change the temperature of the white. The differences in "hard" light modifiers is explained nicely here:

UPDATE: I checked and the 4.5° angle cannot really be changed (it looks like its using a 3D Fresnell lense) and only diffusors can be added that do not provide a sharp frame. So i guess this only works if you have a lot of distance between this thing and the window that you want to shine through. Back to the "classic" permanent lights with optical projector mount from rollei or nanlite. 

FINAL UPDATE: I found something that sets you back only 100€: The 40W ULANZI or M60 device is 50 and the optical lens in front to allow to collumate it and become a "spot" or "beam" instead of a "wash" is just another 50€. It certainly is not the quality of the nanlite setup but its a lot cheaper to product the light we need. Key question is still the angle it puts out. But as it can project gobos it should be more around the 10-15° that i think are good for our case. Here is the link - the last option is the optical lense with a mini bowers mount: https://de.aliexpress.com/item/1005006756448852.html?spm=a2g0o.cart.0.0.39994ae4GPaBqn&mp=1&gatewayAdapt=glo2deu

FINAL FINAL UPDATE FOR TODAY: Last option with bi-color that is from a trusted manufacturer and combines the 30W LED with a movable lens. Its not the same sharp light compared to a projection mount but its probably good enough. https://www.smallrig.com/de/RC-30B-COB-LED-Video-Light.html

@bigfield How do your project go? We have a room with no windows which could do with similar treatment so I'm interested in how it went!

I noticed that the Circadian Sky produces an impressive 2,200-40,000k while the LEDstore 300x1200 is a more typical 2900K-5500K. I think that might be less range than my hue bulbs. I use Home Assistant with Adaptive Lighting so our lights are white during the day and warm in the evening - I'd use that for this panel too

Could we manipulate the spectrum of artificial sunlight to mimic the effects of a sunrise or sunset indoors, potentially affecting our circadian rhythms and sleep patterns?

[Advertising removed.] - Mods.

Hello folks!

First time posting here, but I've been reading this thread for about 2 years now 😆  I've been working on a project very much inspired by the original video and this forum thread, and I've finally blogged about it:

https://victorpoughon.fr/i-tried-making-artificial-sunlight-at-home/

Here are some quick pictures, but I go into much more detail in my blog post. Thank you @DIYperks for making a video that has inspired me to work really hard on this project and learn so many things along the way. Thank you to everyone who posted their progress in this thread over the years. I'm happy that I finally get to do the same myself and report on my progress.

Cheers to this community, and long live artificial sunlight projects!

Really nice work @victorpoughon! I agree, Matt's work inspire us all to be better.

@victorpoughon very cool DIY Project

Posted by: @tr47

↑

Building an artificial sun is an incredible idea, and you're right that optics are a big part of getting it right. If the program you found helps with light modeling or spectrum analysis, it could be a valuable tool for all of us.

I’ve read one article that mentioned the difference between full-spectrum and red-blue spectrum lighting—mainly that full-spectrum mimics natural sunlight and supports a wider range of plant processes, while red-blue is more energy-efficient and targets photosynthesis directly. But I’m not sure I fully understand the trade-offs. When designing your artificial sun, are you aiming for a full-spectrum output or focusing more on red-blue wavelengths for plant growth? What are your thoughts?

Yes, the program does help with light simulation. I'm still deciding between full-spectrum and red-blue—full-spectrum seems better for all-around plant health, but red-blue is tempting for efficiency. Maybe a mix of both is the best option?

New results, apparently you can bind titanium to glass with a standard etching laser

The titanium sample did show that blueish effect on the example with a weaker laser pulse

Hey folks,

I've been a long time lurker of this thread and I'm really amazed by all the discoveries you have made.

My question is, if the only goal of the artificial sunlight project is to have a light panel that produces a lot of light that travels and looks like sunlight, but without making the panel itself look like the blue sky, can it be done without spending a huge amount of money and time? I'm mostly talking about the broken TVs idea here, I'm wondering if a light panel big enough and designed to reduce the inverse square law of light, mounted on an angle in a corner of my living room which is about 30 square meters big, could sell off some decent sunlight?

TVs lack the colour accuracy of "artificial" sunlight. The panel would give off some light but it's unlikely to be anywhere near bright enough to light a room. Running cost could be high too, but that's likely moot.

Hi forum, would it be possible to attach a solar panel to the artificial sunlight and build a rack on one side of the four walls inside of a shed and install a solar powered 2 split air conditioner to cool everything down to an ideal temperature.

 To power a whole house you would need approximately 30 solar panel system with all of the power management equipment and batteries and wires. If the Artificial sunlight was built in a square box shaped and attached directly above the solar panels with some riser’s to elevate the light directly into the shape of the solar panel and no leakage of light. 

Parts:

1. 30 300watts solar panels.

2. 30 Artificial sunlight panels.

3. 120 riser’s, 4 on each panel.

4. Less cost to purchase electrical power cables that run from the solar panels to the PWM to the batteries. Because the whole system is going to be setup in an enclosed area, away from dirt and debris and also away from bird droppings.

Is my idea achievable?

Looking at at a few purchasable-premade options of this technology, I decided to start my journey into creating my own and eventually mounting it into my ceiling. My overall goals for this project are to try to get sharper edges out of a frame, and 2: not to electrocute myself! I've found a few LED's with built-in heat sinks that I'm hoping can expedite this journey for me. I'm going the inkjet screen route, but if it ends up not having the effect I'm going for I will possibly be using the water and soap method.

Posted by: @hexaclover

↑

Looking at at a few purchasable-premade options of this technology, I decided to start my journey into creating my own and eventually mounting it into my ceiling. My overall goals for this project are to try to get sharper edges out of a frame, and 2: not to electrocute myself! I've found a few LED's with built-in heat sinks that I'm hoping can expedite this journey for me. I'm going the inkjet screen route, but if it ends up not having the effect I'm going for I will possibly be using the water and soap method.

It may have been my own fault, but I applied the reflection tape on quite poorly, thus, the reflection was never quite even looking enough to convince me it was round like a sun. Round 2 came along, I bought expensive reflective/chrome paint from a specialty store. Same problem, not as smooth and reflective as I'd hoped, going back to buying another roll of vinyl mirror/reflective tape and applying it much more carefully this time around.

My alternative is small flexible mirrors that I can stick together to create the surface area with the curve.... No updates on that yet, arriving soon. 

My second problem is creating the "box" it will be sitting in. Given the satellite I bought is 70cm, the box that it would sit in would be really big. I'm considering possibly using aluminum frame painted black on the inside to create the focus/bounce light, as I don't have a wall or a room this will be going in, rather my ceiling. (Ideally).

I've been dissecting other company's lights to see how they set up their light boxes, and I've learned a few tricks I'd like to try once I get the satellite back into a reflective state.

Anyway, lots of failures so far, but lots of of things learned. Will update soon.

Good morning everyone! I'm a long time lurker, first time poster, and am enamored by the thought of creating artificial daylight. I've been working on my own prototype for a few months and wanted to share my findings. 

I've been going about my project a little bit differently than most of the other ones I've seen here. Instead of creating an artificial sun complete with parallax effects, I'm just trying to create a realistic blue sky. It could hopefully be a bit easier to product, still act as a fake window, have controllable color temperature, and hopefully be a lot thinner and cheaper than the out-of-the-box solutions from Yuji Lighting or Coelux (see here). I have a cheaper (still $300) 2x1ft version of these from amazon that sits on my desk and does a decent job of looking good. Problem is it's like 8 inches deep and can't be hung on my wall where it'd look much more convincing. It's been very helpful to take that one apart and see how it works. A super bright LED is collimated by a parabolic reflector and shined onto a piece of plexiglass that's very light blue (might be the same color as the transparent inkjet films mentioned before!). The inside of the box is lined with black velvet, minimizing reflections and giving the illusion of infinite depth. Simple enough, right? 🙃 

When I started, I tried to follow a similar design. Bright LED strip, black background, blue glass, easy peasy. But with the LED behind the glass and not pointing directly towards it, the brightness wasn't nearly high enough. And I was using some bright LEDs--1800lm per meter. By my calculations the one that those sky boxes use are a whopping 4100lm per meter. So my next thought was reflecting the light on a parabola. I 3D printed some slopes and picked up some reflective tape at the hardware store. Same story, even the reflected light is nowhere close to the sky box brightness. And I think the rest of the forums are right. Plain foil tape just isn't reflective enough.

A big break came when I got two slabs of blue plexiglass. One was 1/8in thick, the other 1/4in. Through experimenting, taking apart one of my old LED panels, and some help with chatgpt, I realized shining the LEDs directly through a polished edge of thicker plexiglass gives off brighter, more evenly distributed light than shining LEDs behind thinner plexiglass. See out the image below-- the left is 1/8in and right is 1/4in. Night and day difference with the same brightness!

Recently I got some absolutely insane high cri 3600lm/m LEDs from auxmer at the suggestion of QuinLED's youtube channel (not an ad, but those lights are legitimately the brightest I've seen on the consumer market). Shining them through each end of that 1ft blue plexiglass sadly doesn't make perfect blue sky, but it's the closest I've gotten so far! The two problems I see are 1) the light still isn't uniformly distributed through the glass, and 2) even this "light blue" plexiglass is too blue, making it look more like an aquarium tank than sky. Probably explains why the skybox glass is such a light blue color. This deeper blue also means the warmer whites does not get transmitted well for physics reasons. Check out the vid below where I sweep through the LED brightness intensity.

So to fix these two problems in my next design, I want to try getting a clear plexiglass that's the same 1/4in thickness but 6 inches wide instead of 12. With the LEDs as bright as they are, hopefully shrinking the distance the light has to travel will reduce the drop off enough. Then, I would get my hands on some of that inkjet film to give the face that same very light blue coat as the sky box. Warm and cool light could travel through the glass unimpeded before leaving through the film, giving it the desired light blue twinge. Fingers crossed, that's the design that finally gets to "blue sky" level! Do people have thoughts of different designs? A light guide plate could give me more uniform light distribution, but without a co2 laser cutter, it seems pretty hard to make a custom one. Hope y'all enjoy!

P.S. In my research I came across this company from china called YuerSky that seems to make sky boxes. They have a tiktok that shows off some of their stuff and it doesn't look too bad. Anyone have experience with them?

@vasqueso Awesome work! I'm also creating my own prototype, I'm just waiting for a parabolic mirror from China to arrive, as you've mentioned, parabolic shapes with reflective tape just aren't cut out for the amount of reflection needed to create something high quality. 

I've seen a few versions that have worked like yours that end up using a blue interior and using lightly frosted plexiglass to diffuse the light evenly across such as this video  .There is also this video, which is using only the inkjet diffusion layer on a mirror, which can give you a reference of how "blue" the inkjet diffuser gets, which isn't really very blue (even with multiple layers, I've bought some.) so to make up for it, painting the inside of the frame blue can enhance the look from what I gather, I haven't tried this method myself yet. 

I've actually inquired about Yuersky, their prices are much lower than most, 400USD for a small rectangle without the sun, and about 4000USD starting for the smallest of their "sun" series, which is common among other Chinese companies selling similar products. If you look at their sun series though, when they show the entire setup in their warehouse, you can tell they're just using an LED in the middle mounted on the back. It doesn't seem like they are using any parabolic shape to get that "infinite sun following you" effect like Coelux alleges they have. (Not many videos online of the Coelux sun movement.) Although I still don't know what type of diffusion layer they use.

That being said, I've recreated their sun series as-is on their social media, where the sun is stationary and moves only as you move, but I'm just waiting for that parabolic mirror to get here to test my final prototype version to see if I can actually pull off the infinity sun that moves, if it's a bust, I'll post my finding soon. 

Good luck with your prototype! Not many of us out there lol.

The thinner plexiglass with brighter lights were not the end solution 🙁 I had my 2ft of leds on either end with the plexiglass itself resting on a black fabric. My thought process was the fabric would act as a blackbody and give off the illusion of a far-off sky. The problem was the plexiglass reflected every little imperfection in both itself and the fabric. Putting the inkjet-esuqe plexiglass over this setup did not yield a uniform blue like I hoped. There was brightness right by the LEDs that dissipated much more rapidly than the light in the blue plexiglass did. You can see the setup with the white fabric AND reflective sheet from my LED panel here.

All is not lost though. This iteration brought me the closest to passable blue sky yet. With the black setup right by the LEDs, there's a blue bar that's pretty believable. It doesn't show super well on camera, but it's there. So, uhhh, back to the drawing board! I'm tempted to get one of the cheaper yuerlight products to see how they make their design so thin. 

How about Vantablack or something similar? Black fabrics are never really "black" when there's a decent light source nearby. It's largely perception that makes us think they are black.

@marcdraco matte black paint will already absord at least 95% of the color unless the angle of incidence is nearly parallel to the surface. Vanta black or soot will get you to 99.5 % but the difference would not be that impressive after everything. 

Are we talking measured reflection vs. perception of reflection? 

Human vision isn't that good at contrast and it's easily fooled. I'm coming at this from a photographer's viewpoint and I've seen some quite impressive demos that show us just how easily our eyes get stuff wrong. We can only see around 30-40 shades of gray which is nuts.

This is the Adelson illusion and it shows this in action.

@marcdraco measured reflection. 95 seems to the threshold for a paint to be considered black, some sellers claim 99%

https://www.musoublack.com/products/musou-black

https://acktar.com/product/magic-black/

https://www.iskweb.co.jp/eng/products/lushadeblack/

But in any case, those 5 % are unlikely to make much of a difference