This is faar from finished, the skin does not have any base specular, and the scatter colour is too bright.  I also need to play with the scatter colour node setup some more.  What I have done is use a marble node with scale controlled by an fbm node to create not only variation in the scatter color, but also vein like lines.  This node setup is giving the skin alot more variation than is immediately evident but needs quite a bit of tweaking.

still, I think it looks quite good, and should look much more realistic once a 3d environment surrounds her to provide more realistic gi.

i'm loving that skin btw templargfx

Quote - Hard to define "better." Pagan is clearly beloved of his pixie, but she's not very realistic.  She's got pointed ears, fer crissakes.

 

People are still not getting the difference between a real face as asked by the OP and realisism, which was not asked for by the OP.

Pixie's face is a real looking one, yep she has pointy ears, that's why she is a "Pixie", but that does not detract away from her natural looking face.

If the OP wanted realistic faces, then they would have put it in the first message. It's others misconception of a real looking face to a realistic face that has taken people away from the original message.

All the best.

LROG

I think the problem with the OP's question is that it's an ambiguous one. We have to define what 'proper' and 'real' are referring to. I've tried to define it at one end of the scale with the concept of a physically accurate translation of a real world face via 3d scanning. Others are looking at it in terms of all the right elements (eyes, nose, mouth ears etc) in the right places within a certain set of boundaries that the mind considers acceptibly accurate without direct reference to a RW face (I'd argue that this is relative to the individual observer, and that without direct reference our minds tend to stretch those boundaries, be it in relation to form, surface quality or placement within a scene). That's fine, but there is a difference.

The term 'character' is also used, which to me is more open-ended to interpretation - i.e. anything that isn't default V4 - on that point, any morph on V4 would fit the description, even a cartoon character.

I know, it just seems like I'm being pedantic.

Glad to see the untextured renders - thanks for doing them. I do find it useful to see the mesh shapes in isolation like this, although it's probably more useful if there are shadows cast in the renders. So often you see major flaws before textures get applied and you get a sense of how well light might fall across those shapes.

I'm not wholly convinced by either Reby Sky or Anna Marie Goddard, but they do have some interesting shapes. Ironically I find myself more drawn toward the Anna Marie Goddard morph (seeing past the eyelashes and brow geometry, and the oddness happening around the lips), particularly looking at bobbeschs render. I might have to pick that one up.

Totally agree. To others it may be different but to myself, I look at the eyes first of all.

As with RL faces, everyone sees something different in them, that's human nature.

I have lots of different characters for V4, but I always come back to Pixie as I know what lights etc work best with her. I'm not too bothered about ultra realisim, but I know what I like and Pixie is that character.

I also agree with characters of distinction, by that I mean not perfect models, give me something that looks as if it has had a life. Crows feet, wrinkles etc. These are what make the difference when I look at characters.

All the best.

LROG

Richard, frankly - faces like the one on your pixie are a dime a dozen in the MP. I don't see how your pixie proves anything as far as proper faces...lol. It's the typical Vicky stuff ya see all the time. But I do agree with you about details. Details are everything and what makes a face look believeable or ultra fake ;).

Laurie

Bring a render to the thread Laurie.

All the best.

LROG

Quote - Bring a render to the thread Laurie.

All the best.

LROG

Didn't realize this was a render competition....

Laurie

Show a render of your version of a real face.

All the best.

LROG

no need for anyone to fall out, it is an open thread and anyone is welcome to post but let's play nicely guys, I'm enjoying the discussion in this thread, I'd hate it to get locked as we seem to be progressing..pretty please with icing on........I'm just running some more renders,

Obviously couldn't read the word render. Also, not of V4.

All the best.

LROG

Post another fairy Richard.

I have what you have...a lot of smooth textures with no details.

Laurie

No need to, I made my point, it was on the topic of the OP and it was of V4.

All the best.

LROG

I suppose I'm missing something...but whatever.

Have fun.

plonk

Laurie

Yes you did.

Yes I will.

All the best.

LROG

I don't think the original thread starter has ever come back, so we are now continually derailing the thread and people are trying to push other people's buttons. :o)  Exactly what I am going to do...

Take a look at women depicted in gallery images and they all have the same blank, emotionless and expressionless face. Few show any signs of realism. Most are stiff looking in forced poses.  Why is that? It isn't the fault of the model used, it's the one pushing the mouse around.  There are plenty of dials to play with yet nobody bothers to use them.   Put some life into your "lonely" V4 or whatever characters and you might see some realism.  :o)

Everyday I walk around the city streets where I live and you know what I see?

Blank, emotionless and expressionless faces;)

Those faces are still real, if a little dull.

But I know what you mean.

I think this one is pretty convincing. Is this a proper face? I don't know since I have no idea what a proper face should be.

Quote -There are many factor which make a face believable. Not only a good morph and expresiom, but the texture, the actual shading, the lighting and maybe the render capabilities all come into play.

I think this one is pretty convincing. Is this a proper face? I don't know since I have no idea what a proper face should be.

 

Very nice render, and a believable face. 

Ive started sculpting a face using some references i took today. Boy is it slow going. 

winvdb - that looks really good, the face shape is great. Unfortuntely the lighting on the skin does not match the lighting on the hair (getting decent looking hair that matches your render is hard!) and this really detracts from the realism.  Also the eyes dont look quite right.

mysticeagly - its looking really good, I would do something with that hair or remove it, as it is really throwing off the image.  Otherwise skin colour and tones look really good.

I wont double-post my render here, but I have posted another render of my previous one with some tweaks for more realism in my skin shader thread if you want to see.  I will post the final image here once I have created an environment for her.

Quote - winvdb - that looks really good, the face shape is great. Unfortuntely the lighting on the skin does not match the lighting on the hair (getting decent looking hair that matches your render is hard!) and this really detracts from the realism.  Also the eyes dont look quite right.

I am curious here. The eyes, yes maybe the cornea blocks the light a bit too much.

The hair I would like to understand. I am using an IBL image as light source which has multiple light sources. The main light source comes from one direction and lits the skin and hair equally. What specifically make you say that they don't match? Not wanting to argue, but wanting to understand what should be different

I did not mention this earlier, but it is an octane render, so the lighting is different as in firefly. This to avoid misunderstanding

When I view this image I immediately feel that the skin is a little darker than it should be (according to the highlights on the hair) and there seems to be a little too much green colouring in the skin.  The lighting on the hair suggests a yellowish/bright light, but the skin looks more like a heavily diffused nature type lighting.

It is still a fantastic render though

Quote - When I view this image I immediately feel that the skin is a little darker than it should be (according to the highlights on the hair) and there seems to be a little too much green colouring in the skin.  The lighting on the hair suggests a yellowish/bright light, but the skin looks more like a heavily diffused nature type lighting.

 

It is still a fantastic render though

Ok, that makes sense. But there is little what we can do about these things. Each of the textures has been made under different light condition and will react differently on a new light source. This means that you have to modify skin or hair to match each other.

that is what I have been working on in my skin shader thread.  I am trying to remove the skin tone and colouring from the diffuse texture, then having SSS add the tone in through its calculations.  The image I rendered before looks quite horrific if the diffuse layer is removed, revealing the full power of the SSS layer :P

Would have been good if the OP had come back to clarify whether they meant realistic faces or just taking V4 away from that default look. Quite different things - Jessica Rabbit is a character!

wimvdb - love that render - is that V4 in Octane?

oops - xpost! I see that it is ....

that is really impressive.  If only you could combine the skin softness of the left image, with a little of the contrast of the right.

impressive stuff!

Quote - Some good points and excellent examples. My contribution - V4, Poser render and finished image (cropped from original). Hopefully she's got a bit of character ...

Would have been good if the OP had come back to clarify whether they meant realistic faces or just taking V4 away from that default look. Quite different things - Jessica Rabbit is a character!

wimvdb - love that render - is that V4 in Octane?

Yes, it was a quick setup and render (the body was still in the zero position). I added the morph and added an expression. In the plugin I changed the skin to my SSS version, then changed the eyes texture to get one without burnt-in reflection, changed the eyesurface transparency a bit to allow more light on the eye. I used an HDR as lightsource and rendered. Render took about 10 minutes to get at this quality.

I like the face in yours wim, but the green tinge that the skin has for me anyway, sorta ruins it. not criticising just stating personal preference...........

i do love the witchy caisson, tbh one of my faves so far

The main problem with that hair is two fold (in my honest opinion)

Firstly, the shading of the hair seems entirely too uniform.  the side of the hair in the direct light seems no different to the hair in shadow.  I find I have this problem with most default shaders that come with hair unfortunately.  I usually end up making my own.

Secondly, the amount of actual hair seems too little.  all those bits that are sticking up at the top seem out of place due to the number of visible hairs.

I think hair is the one thing in Poser that is seriously behind the times.  Content artists seem to produce the same type of hair using the same methods, and have been for years.  the textures are almost always straight strands of hair with little to no overlapping, and the size of the hair strips are usually quite large and too few in number.

mysticeagle - much prefer the second render, except that I find the reflection in the eyes a bit too strong. Maybe slightly more defined creases as well, especially the laugh lines? I like the character though!

wimvdb - very impressive quality - can I ask what card you're using? Also, is it possible to alter the RGB values for each texture in Octane? To me the hair has a reddish cast to it which looks mismatched to the scene, I'm wondering if it's possible to make it more neutral. I think I read that Octane has a white balance option but not sure if that's implemented in the plug-in yet - then again, I haven't calibrated my monitor in a while!

Quote - wimvdb - very impressive quality - can I ask what card you're using? Also, is it possible to alter the RGB values for each texture in Octane? To me the hair has a reddish cast to it which looks mismatched to the scene, I'm wondering if it's possible to make it more neutral. I think I read that Octane has a white balance option but not sure if that's implemented in the plug-in yet - then again, I haven't calibrated my monitor in a while!

I am using the Nvidia GTX680 (main reason is because of VRAM and texture slots)

You can cascade/mix nodes to manipulate textures. There are some limitations compared to firefly, but you can do most things you'd want to do. And there is a white balance option. Besides that, you can set ISO, Gamma, Saturation and camera color preferences which can make quite a difference.

The plugin closely follows the development of the Octane team (1.02 support for the plugin was withing 24 hours, it just needed a recompile)

I use a high-end calibratied monitor (is calibrated every week). I don't see the color deviations which were mentioned, but I do realize that it is also a matter of personal preferences.

The texture which you are using is really an important factor. You can change the color, but not the details. In this picture I was using Danae's Rio texture set. Danae's textures are excellent and work under most light conditions. There are several other popular texture sets which are more difficult, you really have to add texture details to make them interesting at closeup.

Wow! That is an amazing render Laurie! One thing I noticed when I was trying to create a Megan Fox morph-(see earlier render in this thread) is that V4's face is quite a bit wider than average. Even the face below- the distance between the cheekbones is quite a bit less than with V4.

[quote

]

Quote - Wow! That is an amazing render Laurie! One thing I noticed when I was trying to create a Megan Fox morph-(see earlier render in this thread) is that V4's face is quite a bit wider than average. Even the face below- the distance between the cheekbones is quite a bit less than with V4.

 

I agree entirely, her cheekbones are huge and jutting. Same with the jawline, the back of the jaw seems really wide, especially when comparing to celebrities

Caisson

There you are! My favorite displacement map maker. She does indeed have character!

Quote - Some good points and excellent examples. My contribution - V4, Poser render and finished image (cropped from original). Hopefully she's got a bit of character ...

Would have been good if the OP had come back to clarify whether they meant realistic faces or just taking V4 away from that default look. Quite different things - Jessica Rabbit is a character!

wimvdb - love that render - is that V4 in Octane?

 

oops - xpost! I see that it is ....

Reminds me of Comedian Phyliss Diller in the 1970's!

Quote - i'm still trying, i know this has still a long way to go but i think its sorta getting there, yes its too pretty, the skin is still too perfect but i think the slightly aged look has a greater rather than insignificant effect, comments?

I'm still waiting for someone to state the definition of a "proper" face to use in poser renders.  The question of why so few can't be answered until "Proper" has a proper definition.

I've seen a lot of faces in my years, and have yet to see one that I'd say is improper.  Some were pretty homely though. 

Doric.

I can see the original question as meaning two things :

How many V4 renders have characters applied to them (ie are not default V4)

or

how many V4 renders have faces with realistic non-uniformity and shape

I think the OP was going for the first one.  This thread has long since stopped being about the original topic and is now more about what people feel looks realistic when using v4 as your base.

OT : Dont you hate it when Windows thinks poser has crashed because its taking so long to a single bucket?  The window goes all white and you cant see the render.  Stupid windows.

I don't see anything wrong with using the default V4 face, just that it's not very pretty and the definition of features isn't as good as it could be.  Carry on.

Quote - OT : Dont you hate it when Windows thinks poser has crashed because its taking so long to a single bucket?  The window goes all white and you cant see the render.  Stupid windows.

Isn't it nice to know that with all of the advancements in computers, something has remained the same over all the years?

Doric.

By definition, any of the celebrity morphs are ‘real’ faces, and I’ve seen some excellent ones. That would seem to indicate that the base figures are capable of it. Now most of them are ‘pretty’ as well, but if you can duplicate Fox or Alba, you should be able to duplicate Jane Doe.

Let’s step back from accurate facial structure for a moment and consider something else, age. Why are there so few over 40, some would say even over 30 characters. That certainly doesn’t reflect reality. My hunch is that it doesn’t even reflect the user demographic – at least here in the forum. One could say the same thing about the relative dearth of children or non-Caucasian characters, and of course, the gender ratio.

So Poser Planet is apparently located on the campus of an exclusive women’s college somewhere in Scandinavia maybe. Faces are not the half of it. It is not that there aren’t any diverse products out there. Surely this market is not a dead zone where the rules of conventional market economics are null and void – if they were popular, people would make more of them. Instead, the market for non young, White, twenty something, attractive females is so great that well, as Yogi said, ‘It’s so crowded, nobody goes there anymore.’

NB: Dina Marie Vannoni was the fitness model 3D scanned for the Dina figure. You can look her up and judge the fidelity of the process. Some images may be NSFW.

IMO the OP's question can be reduced to a simpler one -

How do you take a digital model like V4 and give it the appearance of a dynamic, believable, life-like figure?

It doesn't matter whether the goal is realism or illustration, or even what the media is, the problem/goal is exactly the same. When I look at an image with V4 I and start thinking things like 'wow, nice morph' or 'that temple looks chilly, she'll catch cold dressed like that' then it may be a technically good image but it's missing something, and that's where I'd place everything I've done with Poser. So the question that I should really ask would be -

How do you take [polygons on a screen/lines on a page/paint on a palette - delete as appropriate] and arrange them in such a way as to evoke an emotional response in a viewer?

For me the image here that gets closest is wimvdb's. My initial reaction was along the lines of 'what's she thinking' followed quickly by something like 'holy cr*p that can't be V4' .... 

Quote - It doesn't matter whether the goal is realism or illustration, or even what the media is, the problem/goal is exactly the same. When I look at an image with V4 I and start thinking things like 'wow, nice morph' or 'that temple looks chilly, she'll catch cold dressed like that' then it may be a technically good image but it's missing something, and that's where I'd place everything I've done with Poser. So the question that I should really ask would be -

How do you take [polygons on a screen/lines on a page/paint on a palette - delete as appropriate] and arrange them in such a way as to evoke an emotional response in a viewer?

I understand where you are coming from and it is always my intention to evoke a response with my renders, something I strive for and rarely achieve.  On the other hand the OP referred to 97% of renders and from that I assume, rightly or wrongly, that he is referring to those seen in galleries.  I doubt from what have seen that this was the goal for a large number of renders.   For a start a large proportion could be classed as 'pin-ups' which from their very nature are not often aiming for reality.

This is where I have a problem with the whole reality goal.  Since reading this thread I have been paying extra attention to news readers and the like and looking to see where their faces are not symetrical.  With very few exceptions, I can't see this, that is not to say that their faces are symetrical and if I got close enough I could propably tell.  In the real world we do not get close enough, or study faces to the degree that this can be seen.  So it is not only a question of defining reality but also a question of how much is percieved reality.

After spending days getting V4 to look as realistic as possible, I finally set off the render for the final gallery image....
4 and a bit hours later I decide to jump online, so I plug my phone in and turn on tethering. BAM!  my computer crashes!

Although I have the scene saved, I did make changes to the teeth and eyes before rendering, and those changes are lost so I cant just continue the render :'(

I think symmetry is just part of the whole realism genre(if you can pigeon hole it as such) and also I think that very often people in the media, ie newsreaders, especially in recent years, are chosen as much for their looks as for their journalistic or presentation skills.

How much of the average Hollywood blockbuster depicts average people is another matter, so i think we are looking in the wrong places for comparisons to "real" if we stay with film stars, TV presenters etc. Yes there is no doubt they are REAL PEOPLE, but as i said, more often than not their personal appearance ie, their attractivity plays a larger part in their selection.

My personal defintion of real or real life is one that is slightly out of the main stream media star spectrum and would center on the average joe, the plain jane, maybe personally I am looking at the opposite end of the beauty scale as a true representation of average. I juist think that if you randomly invited 1000 people into a room, you could easily pick out the pretty/handsome specimens and several hours later have forgotten the faces of the others. Not that the pretty ones don't deserve some attention or that there is an obvious demand and desire to render/create pretty, but that I think the ugly'/plain/forgettable people need not be forgotten ....If this makes any sense whatsoever.......

Attached Link: My render

took 9 hours to render!!  

unfortunately it still looks unrealistic, but its the best i can do!