Atomic Model Editor

Page 6 of 7 Previous  1, 2, 3, 4, 5, 6, 7  Next

View previous topic View next topic Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 16, 2017 5:49 pm

LongtimeAirman wrote:
Nevyn, your post immediately preceding this has changed from
Airman, did you get my page working in Firefox or Chrome?
to some sort of matchmaking arrangement. Good Luck.

It looks like I have split personalities and one is setting up the other. Hmmm, that could explain a few things!

I think Cr6 has accidentally hit the edit button instead of the cite button. I'll assume a few beers were involved.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 16, 2017 6:07 pm

LongtimeAirman wrote:.
In a review of the nature of solids, Hume-Rothery Cool illustrates that covalent and ionic solids rarely involve more than six atomic bonds per atom and metallic solids may involve as many as 14 bonds per atom, if we consider the next nearest neighbors contribution of the body-centered cubic configuration as significant.

Cool Hume-Rothery, W., "Atomic Theory for Students of Metallurgy", Monograph No. 3, Institute of Metals, London, 1960

Six atomic bonds per atom make perfect sense, it agrees with the AV model’s cardinal directions: main up/down axis, and four carousal directions, left/right, front/back. I assume that the body-centered cubic configuration means we have metallic atoms in an orthogonal network where each atom’s six points are connected to adjacent atoms, and all atoms are oriented vertically.  

How do we interpret "metallic solids may involve as many as 14 bonds per atom"? Hook positions can hold valance electrons, but they don’t form channels. Are any additional channels possible here? Can the orthogonal oriented stacked alphas in column and carousal positions form 6 additional parallel charge channels? Does that sound plausible?
.

This part has me confused:

if we consider the next nearest neighbors contribution of the body-centered cubic configuration as significant

That seems pretty strange to me. What does the next nearest neighbor have to do with bonding? They are trying to make something significant that is not. At least as far as bonding is concerned.

I can see 12 bonds being possible if the central atom is very large and the bonding atoms are very small, in comparison. That may be possible since 2 small atoms could bond to the same location on the large atom. I have used this arrangement to explain some explosives like RDX. Not sure if I have posted that work here or not. In that model of RDX I had a ring configuration for the central atoms (RDX is a complex molecule, not a single central atom bonding with other atoms) which had 3 arms coming off of it. Each arm had 2 Nitrogen atoms bonded to the same location on the end.

avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 16, 2017 6:23 pm

I may as well explain that RDX molecule while I am at it.

RDX was invented because the previous explosives were too volatile. Nitroglycerin is well known for its volatility and TNT was created to replace it. TNT is better and can be handled fairly easily but it still had issues with transport. RDX was developed to be more stable but also more potent.

The potency comes from the Oxygen atoms in the core ring structure. These allow more charge to be stored. We also have 2 Nitrogens plugging up each arm which is what holds that charge inside of the molecule and also provides an easy way to break it, leading to an explosion as all of that charge is released.

The volatility problem is solved by the central ring structure which I also found in TNT where-as Nitroglycerin had a Y configuration with 3 arms joined at the same location at the center of the molecule. Thus explaining why Nitroglycerin is so volatile. TNT had a core Benzene ring but still had 3 arms coming off of that so it was more stable, but still volatile. RDX keeps the stable ring core but has shorter arms which helps it reduce that instability.

Nitroglycerin:



Tri-Nitro-Toluene:

avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 16, 2017 6:24 pm

Aaaaand now I'm probably on some list somewhere for talking about explosives.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Fri Mar 17, 2017 12:10 am

.
Beautiful and dangerous structures Nevyn. I admit I have a feminine side too.

Nevyn wrote: What does the next nearest neighbor have to do with bonding? They are trying to make something significant that is not. At least as far as bonding is concerned.

Airman. Thanks, good observation. I haven’t seen Hume-Rothery Cool, and don’t wish to join ebooks for the privilege. Before reading Miles, and seeing Atomic Viewer, I suppose I thought of atoms more as spheres, with all kinds of possible atomic metal lattice configurations based roughly on the number of electrons and sphere packing. Now that I know about our six charge channels, my recollections make no sense at all.

Given our previous discussions, I assume our main axis’ hook positions would likely spin about the main axis, independent of the carousal. In a metal lattice, the main axis and carousal are locked into fixed positions, but the hook positions are still free to rotate. The Tech Note mentioned how metal samples can be exposed to electrical currents at various angles to enable precisely oriented measurements. It occurs to me that when a fixed metal atom is being examined under such conditions, the hook positions will always reorient themselves to maximize current intake. To that extent, definite and measurable hook charge channels will exist. Does that sound correct?

We must look at crystals some time, any chance AV might extend to lattices? Two or more?  
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Fri Mar 17, 2017 1:20 am

Whenever I look at that TNT molecule, I think it looks mean. Like a suit of armour with spikes all over it. RDX almost looks like a fluffy bunny in comparison. So smooth and rounded.

I agree about those hooks always finding the charge. I'm not sure how that happens at the moment, but they should be able to rotate about the axis coming out of the atom at their location. That is, the north hook can rotate about a north axis without affecting its relationship with the rest of the atom. They can even slope to the side some-what (think Methane) given the right conditions.

What I am unsure of is how the proton stack, and therefore the protons in that stack, actually get turned. It seems to me that the BPhoton of the proton must be affected directly in order to turn it. However, atoms are connected by charge streams rather than direct contact like that so it would require a very strong charge stream/field to reach all the way down into the proton, beneath its own charge field, in order to affect it.

I have thought about crystals and lattices a few times over the years, but never too far. They always felt like the next layer of things to study as I was bogged down in stacked spins and nuclear models. I kind of saw them as structured molecules and therefore not quite as exciting as other molecules like Hydrocarbons, Acids and Bases, and, of course, explosives.

I did play with some Carbon lattices in the really early days of desktop AV. I still have images I generated from it and they are so old that the different proton stacks aren't even colored. I don't agree with the way I structured the Carbon lattice anymore, but it was interesting to fit the atoms together.

I would like AV to be able to show lattices. I could do it now as I already have an atom positioning system in place but I really want it to work out how far apart things should be, given the atoms or molecules being used. It actually seems like this should be fairly easy, now that I think about it. I hadn't realised that my positioning system could be used for this purpose.

I will have a look into it when I get the new code fleshed out a bit. I'm still in the very early design phase as I am trying to restructure the code around a building block model and a charge calculation system. But I can see a page specifically for lattices, where I can use lesser graphics constructs, being quite useful. It is closely related to molecules too, so it may kind of fall out of that work which I am really keen to get to.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Fri Mar 17, 2017 1:52 pm

.
I'll throw out a couple of clay pigeons.

When I owned a Mac 512K I spent considerable amounts of effort designing, among other things, screen saver background tiles. I was most inspired by a popular screen saver that gave the illusion of flying straight through a starfield. Real simple, black background, expanding dots, roughly radiating from the center of the horizon, yet plenty of action with near misses and close shaves. Those stars could just as easily be interpreted as photons. What if you could provide an optional background choice to ‘view the charge field’. Your current black void may then be replaced with the random points or tracers of passing photons. Of course, to display lightspeeds accurately we would need to know what time scale we were at, the proton establishes the size scale. Limit the view of passing photons to briefly visible white dots that aren't visible when they are too far away, nor should they approach closely enough to reveal spin details, or not.

If you can create a lattice, you should allow someone to move their vantage points. Passing by orchards or stakes in a field, one observes a familiar optical effect. Various sized alleys extending beyond the field, opening and closing, flickering as you drive by. The atoms and their associated charge streams passing in a slow change of perspective would be wonderful. Still, hopefully, pretty much just a CAD background may do.

Or aqueous solution, or suspended gases. Gas and water can penetrate the outer layers of a metal crystal. Sorry, going overboard again, talk is cheap. I would be coding too if I could keep up.
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Fri Mar 17, 2017 10:50 pm

I started to plan out an ambient charge field shader the other night. I fear it will interfere with the existing shaders and look a bit weird since the ambient field will not be the same charge that goes into and out of the particles. But it is still a useful shader to have in my toolbox so I will see what it looks like.

I'm starting to get the hang of these shaders now, but I still find myself wanting to do things that you just can't, or shouldn't, do in a shader. It is a very restricted environment which is a great pain in the butt but I also find myself liking it. Having to work around the restrictions and get what I want out of it is a great source of accomplishment.

I have a lot to learn though and I have started to think about trying to use shaders in other ways, with different primitives. I have only used points so far, since I was modeling charge, but I am interested in working out how to get a more gaseous or foggy appearance. I think that might work better as an ambient field than point particles.

Imagine a 3D grid of cubes and each cube has a charge density which is used to apply a transparency across that cube (may be rendered as a sphere or other primitive). I can include a direction vector or apply some other dispersion mechanism to distribute the density to the surrounding cells. I might use the color to represent photon wavelength/frequency.

If the particles in an atom could change the field around them, then we could see the charge profile of that atom. No need to render the particles themselves. I think that is an interesting way to look at atoms and molecules.

As you say though, talk is cheap. I like to float ideas around for a while and I'm also trying to figure out what sorts of pages I want in the new AV and this is a good candidate.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Sat Mar 18, 2017 6:35 pm

Nevyn wrote, Imagine a 3D grid of cubes and each cube has a charge density which is used to apply a transparency across that cube (may be rendered as a sphere or other primitive). I can include a direction vector or apply some other dispersion mechanism to distribute the density to the surrounding cells. I might use the color to represent photon wavelength/frequency.
Nevyn, Your patience is appreciated. I’ll try to describe something to program by, as simple as possible; don’t know if I’m violating shader use rules or not.

Ambient Charge Field. Put yourself anywhere in the solar system *. Even the most remote locations receive a constant stream of photons from the sun and all other planetary bodies. All charge source bodies are lamps, illuminating the system with their own unique blend of photon emission spin frequencies and energies.  

Each charge source in the system will be found at specific directions and distances to our viewing location, such that a perpendicular cross sectional area at the viewing location directed to any of those sources will receive about the predicted W/m^2 contribution from that source. Those photons will arrive in that cross sectional area at a fairly uniform distribution; they will all share the same linear vector.

In effect, beams of photons will fall through the viewing space like rain from each and every one of our sources. If one were to look directly at any source, one would see the photons forming the ‘flight-through-the-starfield’ effect.

We can likewise simulate the planet’s surface charge emissions. At AV scale, a dense mix of photons rise vertically upward. We would still receive photons from all our celestial partners, although at reduced numbers, provided they were on this side of the Earth. More random photons would need to be included.

* We start with your Solar System model of course, just add charge.
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Sat Mar 18, 2017 8:03 pm

Don't worry about shader rules, that's my problem. Sometimes you can create the effect you want purely in the shader. That is the ideal solution because it is fast but you can't always get what you want that way. You have to manage part of it in the shader and part of it in the host application. It is best to minimize the host parts which usually means trying to only change things that apply to all particles rather than on a per particle basis.

An example might help to understand it. In my charge emission shaders, I have N particles and each particle represent a charge photon. I could implement that purely on the host and I would have to change each photons position every frame. In fact that is exactly what I did in the desktop AV and it did work fine because there aren't a prohibitive number of particles to deal with.

In browser AV, I used shaders which made me rethink how I go about it to get better performance. At initialization time, I create a direction vector for each particle and I also set a time offset value. Inside the shader, it works out where each photon is by using the direction vector and the current time plus time offset to determine how far along that vector the particle is. This works such that when the particle reaches a certain distance, it will go back to the start position. This allows me to only change the current time, which is a single floating point value shared among all particles.

Now, I have actually found that using the current time can cause problems when the app has been running for a long time. Essentially, the precision of the time value gets worse and worse which shows up as lines in the charge emission. Not a big deal but it would be nice to fix it, I just don't know how to at the moment. But I haven't put any time into it either.

Recently, I tried to make my charge emission shaders work without being tied to the position of the proton. This caused all sorts of problems because each photon needed to know where it started and where it was going. When tied to the proton, it got its base location from the proton (well, it is really just where the shader geometry is in the scene but that is the same as the proton location because it moves with it). I still haven't figured out how to solve that one.

What you have described above is what I call the Probe approach. That is, the viewer becomes a probe in the field and measures from the location of that probe. That is how we actually exist in the universe. We are all little probes measuring the photons that come our way. I think that can work in a solar system type of app but not so much in AV.

It could be interesting but it becomes a pain for the user because they have to move themselves, or the probe, around the scene to get a holistic view of the atom or molecule. It is interesting because it is very close to how our measuring devices see things and may actually become an important link between what the mainstream measure and Miles theory. Worth thinking about but you may need to remind me about it later as there is a lot of work to do before I can reach that.

Eventually, I'd like AV to have as many of these 'views' as possible. Different ways to look at the atoms and molecules. So far I have focused on a conceptual view while attempting to be as realistic as possible. It is conceptual, sort of, because it gives you the complete picture without taking the viewers position into account. For example, if the viewer was looking directly at the center of an element, it might not be able to see the charge being emitted from the north hook stack. But AV still shows you that stack because it is trying to give you a complete picture of the atom. The goal is to give the user understanding, not really to show it how it would look to a measuring device. Once that is in place though, we can look for other ways to view it to give further, deeper understanding. So keep on thinking about these sorts of ideas and don't get discouraged if I don't get to them straight away.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Mon Mar 20, 2017 12:40 am

.
Conceptually, at any scale, solar system or AV, we must assume the majority of photons received will be emitted by known charge sources. The atomic element is vertically upright, as Miles and AV describe, in order to maximize receipt of the ambient charge field. There is an implied field primarily comprised of upward traveling photons, followed by a somewhat less dense field of antiphotons traveling downward. We have our known sources, why not show those photon currents, along with some random additions? From one of our earlier discussions, I believe you expect the ambient field to be far more random than I’ve described here.

I’ve got to share a vision with you. I think back to The Matrix, I haven’t seen it since its release almost 20 years ago, but one scene seems quite vivid to me now. Our hero is facing two agents in a hallway, a showdown, there are bullets in flight. At that moment Neo’s perception becomes transcendent; He can see matter is comprised of vertical photon fields. He can see each photons’ path, they aren’t straight lines, they follow the photon stack spin motion. The hallway walls appear more like dense collections of delicate lines of looping, upward and downward Chinese script. That's it. I hope we agree enough to see some extra photon motion soon, even though it may be a bit of a shock.  
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Mon Mar 20, 2017 1:51 am

I find it hard to define the term 'ambient field'. I think I have attached some other concepts to the term 'ambient' which are skewing my usage of it. I just looked up the definition of ambient and it is the immediate surroundings of something. So, in that vein, you are correct to use the ambient field to represent the Earth's charge field when the focus of the discussion is something within the Earth's field.

I was taking it as more of a background, a source of randomness, which is correct in some situations but not in others. However, I feel that it is more productive to explicitly state the surroundings when they are known. In the case of the Earth, I don't think the term ambient captures the directional nature of its charge field. Maybe I've just been burnt too many times with vague words so I like to be more explicit.

The problem I have with trying to implement a background field is how it will interact with the elements themselves. The cheap way will not have any interaction but I am not happy with that. The expensive way will mean that each and every charge photon is an object in its own right. Not some shader effect but a fully fledged entity that has its own position, velocity and it can collide with things. That's a full simulator. I know I want it, and have even put some work into it, but it is still a bit out of reach.

How about this: I implement a shader that moves charge from one side of a volume to the other and repeats. Think of two plates with some distance between them and charge moves from one to the other. I place 2 of these into the scene. One moves charge from the bottom to the top and the other from the top to the bottom. I hook the density of these charge fields up to the existing controls that spin the various parts of the atom (Effects Settings -> Charge Field). That will give a visual representation of those controls, other than the spin of the north/south arms and the carousel level.

That seems quite feasible to do now. In fact, it is exactly like my existing charge emission shader but I change the way the geometry is setup. The actual shader will be the same and only the init code will change. When you have the right tools in your toolbox, the work becomes easy.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Mon Mar 20, 2017 3:35 am

Well I have the shader part working (I did have to change the shader code a little bit) and it does look pretty cool. Definitely a Matrix feel to it, mostly because you have 2 streams moving in opposite directions. I can set the densities in the code, but not through the controls yet, and a 1:2 ratio for N:S (ie top-down:bottom-up) looks good.

It isn't up on the web yet, as I have a few things to work out before I can update again.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Mon Mar 20, 2017 2:56 pm

.
AV is complete as it stands. Yes, it could go to the next level if the elements were displayed in a charge field with active photons. We’d all like to see it, but it's too early for that. AV is at a scale in which photon/proton interaction is at its most revealing; including our ignorance. For that reason, photons in AV must remain a background. In fact, AV may be unsuitable as a photon development platform; you might consider creating a plain old vanilla photon collider instead. Maybe learn a few things and relieve you of the onus of providing a fully integrated AV product.
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Mon Mar 20, 2017 6:03 pm

That is pretty much my plan. AV was to focus on demonstrating protons and neutrons forming atoms, and later, atoms forming molecules. I started a completely separate project that I soon realised could end up as a full fledged simulator. That project was written in OpenGL with a Java host since I wanted to make use of the GPU for particle processing and ThreeJS/WebGL didn't allow that in the browser. I have also found out that there is another kind of shader, called a geometry shader, that I want to play with as it allows me to create geometry on-the-fly but again, WebGL doesn't implement it (most people don't use geometry shaders so they left it out of the spec).

However, things change as time rolls on and new ideas start to form. Sometimes I get lost in the details and need you guys to point out things from a higher perspective. It is easy to get held back by what you can do instead of thinking about what you might be able to do.

With respect to my latest efforts at an ambient field shader, I made all of the charge photons transparent so that they don't get in the way of the actual atom being looked at. If you zoom out they do still get in the way a bit (depending on density) but you can still see the atom inside of all that charge. I'll hook that up to the controls tonight and look into getting my site updated. A while ago I restructured the directory structure of my apps a bit and that hasn't been uploaded yet so I need to change a few other pages and make sure everything is still hooked up and working.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Tue Mar 21, 2017 5:40 pm

I played with the new ambient shader a bit last night and got it hooked up to the ambient charge field controls. I'm thinking of changing those controls since I don't really need one for charge and another for anti-charge when I can just use one slider to represent the ratio between them.

It did get me thinking about charge vs anti-charge and if our assumptions are correct or not. I'm fine with the carousel level spinning and that this is related to the ambient charge field but not so convinced of the north and south arms spinning as well. I'm tempted to disconnect them and maybe introduce a new checkbox to enable/disable it with the default being disabled.

I didn't look into updating the site though and I probably won't have time tonight either. Shouldn't be too far away though.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 23, 2017 3:19 am

It has been some time, but finally, AV has been updated!

The new ambient field shaders are working and can be turned on in the Effect Settings -> Ambient Field -> Enabled checkbox. The Charge density and Anti-charge density sliders can be used to set the respective density.

I renamed the Effect Settings -> Charge Field group to Effect Settings -> Ambient Field to be a bit more descriptive. Other than that it is the same as before.

The positions of electrons has changed to reflect the discussion Airman and I had about that a few months ago. Previously the electrons were inside of an alpha but they are now above and below it, close in to the proton. They actually should be a bit further in towards the center but I have kept them out so that they are slightly outside of the through-charge streams. This just keeps them visible and not swamped by the charge shaders when they are enabled.

Neutrons in blocking positions now block the input charge shaders in that location.

As it has been so long since I touched AV, there are possibly other changes that I can't remember and didn't take note of.

The URL to AV has also changed. It is now www.nevyns-lab.com/mathis/app/AtomicViewer. If you go through the main page it will take you to the right spot but the old version is still up and working. I expect to take that down soon though.


Last edited by Nevyn on Thu Mar 23, 2017 3:34 am; edited 1 time in total (Reason for editing : Added changes to neutrons and charge shaders)
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 23, 2017 6:39 am

I have made a few minor adjustments.

Fixed a bug in Graphics Settings -> Performance when set to Maximum.
Adjusted how charge shaders are applied by the Performance setting.
Set charge density based on Performance setting.
Removed charge shader density controls from the View Settings -> Proton Stack group.
Removed charge speed control from the View Settings -> Proton Stack group.

This is the start of a move to using the Performance setting to control rendering choices in lieu of having many controls that are confusing. A great many controls will disappear in the near future.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 23, 2017 6:42 am

If you set the Graphics Settings -> Performance dropdown to Maximum, all charge shaders will be enabled, including the ambient field shaders and all associated rotation.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Thu Mar 23, 2017 5:33 pm

.

Atomic Viewer 0.9

The new photon background works well. It ties together and completes your previous shaders, now they all fit together as a single charge field.

One quibble. I don’t see any ‘random’ photons. Imo, a small set of randoms would improve overall naturalness. Only if it’s quick and easy though. Please don’t cause any seizures.
 
I found an empty place and looked up and down; pleased to see a new twist, the flight-through-the-starfield simulation works coming and going. My one dimensional thinking is expanded in two directions at the same time! Wow.

Your original shaders are more ‘intense’ than the new up or down fields; maybe you can dial the originals back to match the new fields, then limit control to just your two: Charge density and Anti-charge sliders.

I see the shaders are tied to the atom itself. Here’s an artifact;

Two Helium atoms from above or below (the problem exists for front/back too) we have some sort of Casimir effect (just kidding). If you cross your eyes and try bring the two images together for a 3D effect, it doesn't work and you may get a headache. This is not a problem since I use AV to view single atoms. I’ve never built or seen a molecules in AV, except for your pictures.

Good Stuff Nevyn
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Thu Mar 23, 2017 7:04 pm

LongtimeAirman wrote:
One quibble. I don’t see any ‘random’ photons. Imo, a small set of randoms would improve overall naturalness. Only if it’s quick and easy though. Please don’t cause any seizures.

I'm not sure what you mean by 'random' here. All of those charge photons are random, or at least as random as a computer can generate. Do you mean it should have areas of higher and lower density?

I probably should explain how that randomness is applied to give you a better picture of what is going on. All charge shaders create their paths randomly. However, they only create these paths when they are initialized, not while they are running as that would be expensive and defeat the performance gains of using shaders. So the same path is used, over and over again, as the shader runs. Each photon moves along its own path and when it reaches a set distance, it goes back to the start and traverses that path again. This creates a pretty good random effect, without actually being that random! Even at low densities, it still looks good.

LongtimeAirman wrote: 
Your original shaders are more ‘intense’ than the new up or down fields; maybe you can dial the originals back to match the new fields, then limit control to just your two: Charge density and Anti-charge sliders.

I spent a bit of time looking at the differences between the shader densities and tried to find a nice balance, but I did want the old shaders to be a bit more visible than the background. This difference may be increased because the ambient shaders use a transparency of 0.5 but all of the other shaders are opaque (except the ends of though-charge shaders as they taper off). I think the intake shaders look quite well balanced in your first image. The through-charge shaders are a bit too dense. I could definitely dial them back a bit. I was actually surprised how low they could go and still look good.

Now that I have removed the charge shader density sliders, you can only change these densities by going to a higher/lower graphics performance setting. As you go higher, the emission shader density starts to climb higher than the other two. You can adjust the ambient field shader densities to find a better balance. If you find something you like, post the values and I'll take a look.

LongtimeAirman wrote:
I see the shaders are tied to the atom itself.

Two Helium atoms from above or below (the problem exists for front/back too) we have some sort of Casimir effect (just kidding). If you cross your eyes and try bring the two images together for a 3D effect, it doesn't work and you may get a headache. This is not a problem since I use AV to view single atoms. I’ve never built or seen a molecules in AV, except for your pictures.

Yes, the ambient shaders are attached to an atom. All atoms have their own versions. At first, I did this because it was easier while I developed the shaders themselves. I planned to make them global but when I tried, I found there were problems with that approach. The biggest problem is that the user can move around the scene. So if the shaders are global, then the user could move outside of their range. The larger the area these shaders cover, the more charge photons they require to keep the same density and that leads to performance issues on low spec'd systems.  So I thought I would attach them to the camera, therefore the user could never move outside of their range, but this means that you lose the 3D effect of these shaders because they will move with you. So I left them attached to the atom. I've been thinking of limiting AV to a single atom at a time anyway. I'll create a whole new page that allows you to view 2 atoms side-by-side for comparing them.

That gap is a pain in the butt! I noticed it as soon as I put up two atoms at the same time. There was this small gap between them and I thought that the width of the ambient shaders just happened to make them that close. So I made them a bit wider but the gap remained. I'll take another look to see if I can fix it as I had other problems to fix at that time.

Overall, I'm pretty happy with them. Thanks for pushing me into this. Not only do we now have ambient field shaders, but it has also got me back inside of the AV code and thinking about where I want to take it.

I had some thoughts last night about building molecules. I want to create a cut-down visual representation of atoms. No electrons or neutrons (maybe blocking neutrons), just protons stacks. I need to figure out how to join the atoms together in code. When I did this in my desktop AV, I had to manually move hook protons over so that the two atoms joined correctly. This was very fiddly and time consuming, although I did get pretty good at it after building quite a few molecules. The main problem is that you have coordinate systems inside of coordinate systems inside of coordinate systems. So when you want to move a stack over a bit, you have to figure out the correct direction to move it. What you see on screen may not be the right direction that it needs to go because you are moving it in the lowest coordinate system but looking at it in the highest coordinate system. This seems a bit tricky to do in code, but I won't know until I dive in and see what I can do. At least in the code, it is easier for me to work in those higher coordinate systems and then remove them to find the actual change required. That is about all that is standing in my way of building molecules. Of course, there is also creating a UI for it, but I'm sure that will be easy enough, once I start thinking about it.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Fri Mar 24, 2017 6:39 pm

I have turned off the north/south arm rotation by default. You can enable it in the Effect Settings -> Ambient Field -> Rotate Axis checkbox.

I had a look into the gap between ambient charge shaders and have found the problem, but it isn't so easy to fix. Essentially, the code that places an atom in the scene uses the bounds of that atom to determine where to place it with respect to the other atoms. The ambient charge shaders are part of those bounds but I really don't want them to be. Unfortunately, the code responsible for that is quite deep and will affect nearly every object in the atom. Meaning I have to change the way it works and then try to find everywhere that uses that object and make it use a different one. Very error prone.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Fri Mar 24, 2017 9:19 pm

I have fixed the ambient shader gap, but it has consequences. The ambient fields of each atom overlap and cause a doubling of the density in that area. It seems I just can't win with this one.

I also uploaded the backend database that can be used to retrieve element data. This does cause the page to load a bit slower as it retrieves all of the element definitions. I'll try to fix that soon.
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by Nevyn on Sun Mar 26, 2017 4:14 am

I've been thinking about charge densities and how that affects the input of an atom. Not all charge intakes are the same, as is currently modeled in AV. Let me show you a picture of what I mean.



That is a Silver atom with arrows to show the charge inputs and the color of each arrow gives a rough density reading.

Red = dense charge
Green = medium charge
Blue = light charge

The Red arrows are the Earth's charge field and the Green arrows are the charge that directly impacts the Earth, mostly the Sun. The Blue charge is random charge in the atmosphere or surrounding environment. I wonder how low that blue charge is compared to the Earth's charge?

I was thinking about those north and south hook stacks. They don't seem to be getting much charge. This means that they rely more on the through-charge of the atom. Could this reduction in charge strength help with bonding? The outside carousel stacks are certainly getting fed plenty of charge and this suggests a reason why they are not used for bonding as much as the north/south axis.

What do you think? Could the side-ways charge be more dense than I am imagining?
avatar
Nevyn
Admin

Posts : 808
Join date : 2014-09-11

View user profile http://www.nevyns-lab.com

Back to top Go down

Re: Atomic Model Editor

Post by LongtimeAirman on Sun Mar 26, 2017 11:18 am

.
Nevyn wrote. What do you think? Could the side-ways charge be more dense than I am imagining?

We know that the atom aligns itself to maximize ambient charge input. In your silver diagram, the main charge received is the earth’s emission field upward. We may safely assume the charge source is constant and steady.
 
You’ve indicated the sun’s emission field as downward green arrows; that is true only at high noon on the equator. Most of the time the sun will not be vertically upward. During the night, earth blocks solar emissions from reaching the silver atom; the atom will operate in its minimum energy state with little downward or horizontal charge input. During the early morning or late afternoon, horizontal charge from the sun is at its maximum. The atom will tilt slightly toward the sun to increase the total horizontal and downward charge received while still primarily pointed to the earth. A video might show all surface atoms tilting to track the sun like sunflowers.
.

LongtimeAirman
Admin

Posts : 612
Join date : 2014-08-10

View user profile

Back to top Go down

Page 6 of 7 Previous  1, 2, 3, 4, 5, 6, 7  Next

View previous topic View next topic Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum