Done some searching and trawling, but I've not found anything that quite explains this adequately for my poor Monday brain...

This is probably quite basic stuff, and I know I should experiment more... sorry... but I am hoping someone might already know the answer and give me some guidance :)

I'm trying to make the effect of looking through the reverse side of a magnifying glass.

I'm using the Refract Node at present... is this correct, for starters?

What settings do I need to increase / decrease / invert to get the effect I'm after?

Many thanks for any advice or guidance you can give here...

p.s. I am trying to combine this with reflection ultimately, to get a reasonable glass like appearance, with real reflections. But let's start with the basic refraction effect I guess? I clearly need to understand this some more, relative to Poser's Refract node etc...

Quote - I'm trying to make the effect of looking through the reverse side of a magnifying glass.

Looking through either side of a magnifying glass is exactly the same. They have no reverse side.

Quote - > Quote - I'm trying to make the effect of looking through the reverse side of a magnifying glass.

Looking through either side of a magnifying glass is exactly the same. They have no reverse side.

He he :-) Doh! What am I thinking of then... something else clearly. Just another lens of some sort I guess!

Don't have a magnifying glass in the house, or else I might have realised that... ha ha!

Okay... that makes things simpler then!

Well, in that case, any pointers on how I would just increase the magnification effect?

Many thanks

Hmmm... I mentioned I was suffering from Monday-itis didn't I...

I guess to make the magnification stronger, it's just the geometry of my convex lens I need to alter... per the real world?

Guess I can make a morph for that...

;-)

Attached Link: http://en.wikipedia.org/wiki/Lens_(optics)

Ah, many thanks EnglishBob :)

I wondered if just upping the IOR would do it... thought that had to be the setting I needed to change... but it was the maths of that I was unsure about really I guess...

...e.g. if I had to increase the IOR, or invert it, or what.

I've made a nice morph to increase the thickness (or rather z-depth diameter) of the magnifying glass prop lens now too...

...so I will play around combining that "increase magnification" morph, with increasing the IOR.

Originally, I'd woken up with the notion that the lens of a magnifying glass was convex on one side and concave on the other! Which, of course, makes no sense, I don't think, in the real world.

It is Monday... this sort of thing seems to happen to my brain, on a Monday... ;)

To a first approximation, it's the radius of curvature which determines the strength of the lens, not its thickness - and Poser won't render a particularly accurate simulation anyway. If you do want to make a variable focal length, simply scaling the lens thickness would work fine since that will also increase the curvature / reduce the effective radius.

A lens with one convex and one concave face is called a meniscus lens - and it can still magnify or reduce depending on the relative radii. Contact lenses always have the inner face concave so as to fit the curvature of the cornea. Guess who's father used to be an optician? :)  

Thanks... yes, actually it was the radius of curvature I increased in the morph, er, I think ;)

Well... I made the lens more, er, "spherical"... if that's the best way to put it... in Hexagon and loaded that back in as a morph.

A meniscus lens... contact lens... yes, of course :)

Anyway... as Ashley9803 pointed out, my magnifying glass should only magnify... not de-magnify, whichever side of the lens I look though, he he...

Quote - ...my magnifying glass should only magnify... not de-magnify, whichever side of the lens I look though, he he...

Indeed it will.

Practical magnifying glasses are often made with assymetrical curvatures, which I seem to remember is something to do with reducing chromatic aberration. No need to worry about that, though; Firefly is nowhere near an accurate enough light simulator to have that problem.

On the other hand... Your magnifying glass effect might be enhanced if there was some visible aberration. I have no idea how to achieve that in the renderer, though, or even if it is possible at all; I suspect not. Postwork may be required. ;)  

Yes indeed, some postwork may be resorted to ;)

dunno if I'm losing me marbles, but remember when you were at school doing pencil drawings of a bottle for art class, there was always that little reflection bubble i used to call it that the teacher insisted you got into the drawing to simulate a real world reflection, maybe incorporating something like that as a layer might have the desired effect.

Ha ha... yes, I think I know exactly what you mean Mystic ;)

I used to always add something like that when I hand drew pen and ink comic strips as a kid...

I did one, a very long time ago, in Bryce. Pick the Looking Glass image from http://www.main.artbeeweb.nl/?p=3990 

Each one was made of two slices of a ball, glued together. Ball was made of simple glass, index 1.5 or so.

Each looking glass has a focal length. If you're inside that range, the image turns upside down. So if you set each magnifying glass withing the focal range of the next one, you'll get a series of flipping images. In the first, you can see the second upside down which shows the third one straight up. The renderer must do that automatically.

But do note that you have to crank up the max amount of ray bounces in the render settings, otherwise the inner ones will turn black. Each glass adds four bounces (ingoing, two halfay and outgoing) so this image required at least 20.

have fun.

Ah... wait... the flipped / compressed image is actually exactly what I was originally talking about / expecting to get...

...well, I'd forgotten the flipped part...

http://www.warren-wilson.edu/~physics/PhysPhotOfWeek/2007PPOW/20071130LensPPOW/index.html

...and this is exactly what I am now getting.

The magnifying glass prop in my scene is a bit of a distance from the camera.

It all makes much more sense now :)

If I remember right, Poser can do a magnifying glass or glasses with 5 or 6 raytrace bounces. You have to model both sides of the glass to get it to work.

I think BB did a writeup on it as well, not sure if it was here or at DNA thou.

Is this what you are looking for?

Best Wishes,

Rªnce

I take no credit for the material used here.  As I said, I think I copied notes posted by BagginsBill.  Most of the tricky stuff I do comes from that source.  I do not have a link to the original thread, but I can post a screen shot here of the settings used, or upload the MAT file and post a link here to it.

Maybe someone has a link to the thread.

Be Well,
Rªnce

Thanks Rance01.

Yes, that's it... although, in fact I'd initially expected, because my magnifying glass, in my scene, is a reasonable distance from the camera... and what is behind the glass is a reasonable distance away, behind the lens... to get the effect detailed in that physics blog link I'd posted above...

Basically the scene behind is actually shrunk, effectively, and also flipped upside down.

Least it looks like that could well happen in the real world, in the scenario I'm depicting.

What I've done so far today is now based on the shader from BionicRooster's Sherlock Holmes set, in Free Stuff here... which I discovered had a working magnifying glass shader on it already! So I've applied that to the magnifying glass prop I'm using (wanted a less ornate style of handle basically).

I need to get some reflection on it now I guess... so I expect a fresnel_blend node controlling the mix of reflection and refraction is in order... in Poser Pro 2012.

Apparently the IOR of Crown Glass (which I think is used in most reasonable grade magnifying glasses) is 1.523.

But not sure how that translates into Poser... I think I read previously that some heuristics are required?

http://www.renderosity.com/mod/forumpro/showthread.php?message_id=3817511&ebot_calc_page#message_3817511

Which is pretty much what I thought it was going to be actually... I've read dozens of similar threads. Just not in this exact context... and I hadn't initially figured out the correlation I guess.

I've subsituted a fresnel_blend node into the basic "Reflect and Refract nodes, blended off the alt_diffuse" setup... and I suspect that's really now doing the trick (at least as far as I need it to) in combination with the convex lens, within the magnifying glass prop I'm using.

Set the IOR of the fresnel_blend node to 1.523 and then adjusting the IOR of the refract node seems to work to increase or decrease the effect... as does adjusting the radius of curvature via the morph I added.

Now... about that "visible aberation" that EnglishBob mentioned... maybe just a displacement or bump using a clouds or spots node? Either that or I make another, assymetrical, morph I guess.

I'll maybe play with that next I guess... ;)

Quote - Now... about that "visible aberation" that EnglishBob mentioned...

I've been thinking about that. You don't mind me taking over your project, do you? :D

Chromatic aberration occurs because different colours are refracted by different amounts - so maybe you could fake it by doing three renders, one each for the R, G and B components of the lights used (i.e. turn the other two down to zero). Vary the IOR of your lens for each render, then re-combine. Only the light that's passed through the lens should be different, so the rest of the scene should recombine; if not, then a full-colour rendering will be required along with a little more compositing.  

Quote - > Quote - Now... about that "visible aberation" that EnglishBob mentioned...

I've been thinking about that. You don't mind me taking over your project, do you? :D

Chromatic aberration occurs because different colours are refracted by different amounts - so maybe you could fake it by doing three renders, one each for the R, G and B components of the lights used (i.e. turn the other two down to zero). Vary the IOR of your lens for each render, then re-combine. Only the light that's passed through the lens should be different, so the rest of the scene should recombine; if not, then a full-colour rendering will be required along with a little more compositing.

Ah, okay... I think I understand what chromatic aberation is as a phenomenon, from that description.

Thanks EnglishBob :)

Hmmm... it's a tenuous link... but the Hipgnosis, Dark Side of the Moon album cover connection, that has just formed in my mind, would go rather nicely in the scene I'm doing this for actually... if I could pull that in somehow... hmmm....

I wonder if that splitting of RGB and recombining could be done in one shader though?

e.g. color_math add the result of three different Refract nodes (with R, G or B bias set for each Refract node too)... in a single shader...

I wonder...

Bill's work is great.  I really appreciate everything he posts.

Be Well,
Rªnce

Thanks Rance01 :)

Yes, it's essentially the same shader, I think...

I've just used the new fresnel_blend node in place of the simulated fresnel equation that BB had in that pre-Poser Pro 2012 shader... as I understand it...

...and I used ks_microfacet, instead of blinn. That choice was just based on a crystal shader, posted by LaurieA, based on input from BB, in one of her threads.

Sorry was going to post a screenshot of the rainbow MAT, but I'd rather link to the original thread:

http://www.renderosity.com/mod/forumpro/showthread.php?message_id=1881864&ebot_calc_page#message_1881864

gmadone, back in August of 2004.  Renderosity is a great place!

...I've got renders running in both Poser and Vue currently so I can't test it... but here's my starter idea for simulating chromatic aberration in a single shader.

Will this fly, do you think? ;-)

EDIT: whoops, changed the math_function nodes to color_math ones...

Ah yea, just re-read your post.  Still using Poser 7 here.  LaurieA has some really nice shiny metal shaders too.  The magnifying glass is fun to play with -- pull Gosha's hand away from his face and at 1.4 IOR the distortion become really pronounced.  It's a nice effect and I'm glad I stumbled upon the thread with the MAT information.

Be Well,

Quote - Ah yea, just re-read your post.  Still using Poser 7 here.  LaurieA has some really nice shiny metal shaders too.  The magnifying glass is fun to play with -- pull Gosha's hand away from his face and at 1.4 IOR the distortion become really pronounced.  It's a nice effect and I'm glad I stumbled upon the thread with the MAT information.

Be Well,

I think the great thing about a lot of BB's shaders is they're backwards compatible to Poser 7... many of them... as I understand it.

I don't really know how much different the fresnel_blend node is to that set of nodes BB has in the shader you showed there?

It's simpler to use I guess, certainly...

Quote - Sorry was going to post a screenshot of the rainbow MAT, but I'd rather link to the original thread:

http://www.renderosity.com/mod/forumpro/showthread.php?message_id=1881864&ebot_calc_page#message_1881864

gmadone, back in August of 2004.  Renderosity is a great place!

Wow! That's awesome! Thanks for the link :)

Here's a resource for you.  Some of the goods ARE dated now - the page has been hanging for so long - but Acadia, Mistress of Links, gathered together many fine resources here:

http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2635126

Years ago, R'osity updated servers and many of my favorite links no longer worked.  I gathered together what I could find and hung that list online.  Acadia's list is much more detailed.

Renderosity itself - and this long running forum - is an amazing resource for learning Poser.  And the people here who have helped me in the past is manifold.

PS: my rainbow MAT has fewer nodes ... I wonder if I should re-visit that thread.

Cool - thanks for that thread link too :)

Actually, I think I have seen that thread of Acadia's links before ... but hadn't bookmarked it... it would have been nearer the point I'd just got Poser... and I think revisiting it now, the stuff will likely make way more sense to me.

Quote - Okay, on the chromatic aberration front... 

...I've got renders running in both Poser and Vue currently so I can't test it... but here's my starter idea for simulating chromatic aberration in a single shader.

Will this fly, do you think? ;-)

EDIT: whoops, changed the math_function nodes to color_math ones...

Ha! Well, from the prelimary test render, that seemed to work! :)

The difference between IORs in the three refract nodes needs to be a fair bit less though... the RGB offset was too much... but those values at least showed that it (appears, so far) to work.

I can try making some more test renders that I can share, tomorrow or so...

maybe somebody already mentioned it, however one might get reduced/reversed/flipped scene with poser hi-res sphere and thick glass shader.

Use two lenses. Changing the distance between them you change how much is the magnifying effect or its reverse. Maybe you get an inverted image. Add more lenses and you can have a non inverted image.

How physically correct is Poser I don't know, just experiment and see what happens.

Thanks Miss Nancy and Kawecki.

The shader I posted on the previous page is in fact just BB's thick glass shader, as far as I understand it... or a variant of it.

I've now used two color_math nodes to add together three HSV nodes, setup as RGB maximums of Red, Green and Blue. Each HSV node then has a separate refract node attached to its color channel.

For each refract node I've set a slightly higher IOR than the 1.53. of Crown Glass, and varied that by a fraction... e.g.

Red channel IOR=1.87, Green channel IOR=1.9 and Blue channel IOR=1.93.

This creates an approximation / representation of that chromatic aberration effect that EnglishBob explained. The extent of variation I'm just tweaking by eye, based on the scale of my render and how visible I want the effect to be.

The extent to which those Refract node IORs are set above the fresnel_blend node's IOR value, determines, in part, the extent of the magnification.

I also created a morph to vary how spherical the magnifying glass' lens is.

It seems to be working. Things close up behind the magnifying glass lens are magnified.

Place the glass at arms length from the camera with a more distant object behind it, and the glass produces a compressed and flipped view; a "real image", I understand the real world effect is called.

So far, just been testing this on my W.I.P. entry for the January Poser Forum Render Challenge ;)

Once I get that done I'll post some more test images of the results I've got so far with this, in case anyone's interested...

Good job on the CA. Just a note - you go about it with a bit more nodes than you need. Using an HSV to just do color chip multiplication, and not to do actually anything with H, S or V, is a bit expensive.

Put R, G, and B in your color_math adders instead and get rid of the HSV.

Note also: If you had put the words "chromatic aberration" into Rendo search engine, with me as author, you'd find stuff.

http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2754029&page=3#message_3350582

Also discussed spherical aberration in that thread.

Ah, of course... I don't need those HSV's now do I...

I started off thinking I needed to desaturate first... hence why I ended up down that route... and didn't think to back out of it. Cool... less nodes.

Thanks BB

I will go check out that link... and your previous notes on both chromatic and spherical aberration now too :)

Nice... that spherical aberration stuff from the "Artistic Lens" thread is cool :)

Uh oh... IDL overload. My render has a red rash ;)

All my tests so far have been just ray traced (no IDL or SSS).

I had IDL on, with intensity up at 0.5 though, for that last render, when I got the red spot attack... and it's quite a bright(ish) outdoor scene.

Trying again with IDL intensity down to 0.3... could go right down to that 1.5 mark I guess?

All the other mats in my scene will probably have 0.85 diffuse, I guess? e.g. BBGlossy2 or EZSkin2 based mostly.

Should I reduce the strength of the diffuse channel, for my magnifying glass lens shader to 0.85?

I guess I was assuming it would inherit that from the other shaders which are ultimately the source for its refraction and reflection.

But will those R, G  and B filters offset / reset that, if they're at 100% strength?

...also tried to separate out the variables.

The three maths nodes along the bottom row are (from left to right) meant to be controlling IOR-based "magnification boost factor" (just set to a factor of 1 there - no boost), material IOR (e.g. set to 1.523, Crown Glass), IOR variance amount (for the Chromatic Aberration - e.g. 5% variance either way, I think, in the example there).

Also added in a blender node to control the diffuse value numerically.

Guess I could just plug it all into Diffuse master channel, instead of Alt_Diffuse, and use the built in Diffuse Value... what's the difference or benefit with using the Alt_Diffuse, over doing that?

I posted some notes and images with the test renders I did.  That initial node set DOES invert - flip - and mirror what is behind it.  I haven't done anything with the color math nodes.

Playing with the magnifying glass was fun though.

http://www.ranquist.net/tutorial/MGlass.htm

Cool page :)

Yes, it's a lot of fun this subject... maybe in part because it's helping me figure out some of the fundamentals and some more mat room stuff.

But also, I guess I enjoyed playing with magnifying glasses as a kid ;-)

Next test render's looking better with the lower IDL Intensity (0.3). about a third done and no red rash so far...

...dunno whether I need that master diffuse value reduced or not. I had it on for the current render... but it may be that just reducing the IDL intensity was enough.

I think, from what I can see so far, that lowering the diffuse is just basically tinting the lens.

You should post a sample or two here.  We'd like to see what you are working on.

Quote - You should post a sample or two here.  We'd like to see what you are working on.

Yeah, I know, I'd better post some renders... been testing it so far on my forum comp entry scene. Keeping that slightly under wraps till it's finished ;-)

I'll put that aside though and do some other test renders to post here shortly...

Here's a hi-res test of the last version of the shader I posted.

Rendered with IDL on... render settings in next post... click on the image to zoom in my Gallery.

When you zoom you'll see there are some red poly edge artifacts appearing on the lens. So, my focus now is to see if I can get rid of them I guess...

...I'm running a render at same res, but without IDL no, for comparison...

Ah... looks like the render without IDL also has those poly edge artifacts...

Is this an Irradiance Cache issue?

Lowered the Min Shading Rate in render settings from my stock 0.4, down to 0.2 for that IDL render above, and still got the artifacts.

There's a shadow blur of 2.0 and shading rate of 0.2 set on the single, ray-traced, infinite light, that's providing illumination, alongside the IDL.

I've deliberately exaggerated the chromatic aberration effect there, by the way.

It was set at (what I think equates to) 3% variance, either side of the (amplified) IOR of the glass.

(the Refract node base IOR is also slightly amplified, to increase the lens strength... 120% of the IOR value being used by the fresnel_blend node)

Should I maybe be thinking in terms of increasing / changing the post filter settings to get shot of those artifacts?

e.g. I realised I've just got Box = 1.

Would a different filter type, with a slightly higher setting help?

Well, you are beyond my skills set now, my son.  I can no longer help you.  Go forth and render ....

Seriously, nice job, but I'm a Poser 7 user and have enough trouble with the more limited render settings of that program.  Someone else needs to step up with further advise on the artifacts front.

I'll keep an eye open for further examples of your work though.  Again, very nice job.

Rªnce