Almost there now, just packaging everything up and testing it. Here's a taster...

An animation of the morph using a PP2 created from the model in the previous post (with a corrected knight), rendered in DAZ|Studio.

Thanks EB, that makes it so much easier :thumbupboth: !

(That's the big [only?] advantage of doing things a really stupid, long-winded way. Discovering the proper, sensible way is such a relief!)*

And here's all six pieces, created from the same*** twice-subdivided cube (with two extra horizontal edge loops added near the bottom of each piece to create the rounded base). The biggest problem this time was the king.

As always, constructive criticism is very much appreciated.

*the knight actually has an extra horizontal edge loop near the base that I forgot to remove.

Yes, a good lesson in  "less is more", or "Keep It Simple, Stupid!"

 
Regarding the MORPHING chess piece (which is still my ultimate goal here). I obviously can't use the knight texture map, so a single texture/bump/displacement map will have to work for ALL SIX chess pieces.  
 
But since I now have a very simple base mesh, a few extra faces to add detail for each piece seems a good solution. I'll post the results back here soon.

Well I'm amazed - I just tweaked the vertices, created a simple texture map, and voila!

Anyway, I've decided to use the 96-polygon cube as the basis for the morphing chess piece. The outline in the previous picture is too angular, so I need to round that off a bit. And I reckon that I need some more faces to define the ridgeline where the horses chin and lower jaw tuck into the neck (I think I've just realised one of the purposes/uses of edge loops, and I don't have one there).

Plus I need to bear in mind that I'm modelling a pawn, bishop, rook, queen and king from the same mesh, and each of them will need more detail but in different areas. This has turned into an intriguing exercise for me.

The upshot of all this is that the morphing chess piece will be delayed (i.e. not released by the end of September), but will hopefully be a better model.

Tantarus - thanks for the positive comment*. I think I'll do as you suggest, and develop the morphing version separately (but using this as the base).

( btw - it didn't take me 14 days just to do that! )

After many abortive attempts at modelling a knight from a twice-subdivided cube I came up with this. The shape seems recognizable enough to me, and loading the obj into DS or Poser I can render a recognisable knight (from a side view at least - the top and front views needed more work). I din't use the bottom face of the cube - I deleted it and simply extruded the bottom edges to give a starting point for the radially symetrical base.

odf - I like it. Fits with tantarus' "...Find some other theme to model and start with low poly..."  idea too. But I also find that deciding (and posting my intention) to release a model as a freebie does wonders for focussing my mind - and it does need focussing!

I think I'll split this project into two: a basic morphing chess piece for release as a freebie (using the two spheres, plus an extra cylinder primitive for the base) by the end of this month; and a "let's see what I can make out of a 96 face cube" project.

Well, the mesh is definitely far too dense! I tried continuing from what I had - I was quite pleased to start with, as it definitely began to look like a horse's head! But there are just too many damn faces, and they're all being squashed up under the neck. So I decided to do as I was told and start simple. A simple 128-face sphere is actually okay for the top of a pawn, so I started with that. And I can easily manipulate this sphere to get the cross for the king, the crown for the queen, or the rough shape for the bishop. The rook's castellations are just about acceptable but need more thought. As for the knight, the 128-face sphere can be twisted quite neatly to form the basic shape for the neck, but  I added a second 128-face sphere for the actual horses head. (I haven't incorporated this extra sphere into the other pieces yet, but it will give me enough faces to do the slot in the bishop and make the others a bit more fancy) For the details on the knight (i.e. eyes, mouth, nostrils and ears), maybe the easiest approach would be to extrude the appropriate faces to give me extra vertices in just the areas I need them? These extra vertices will simply flatten into the original faces for the other pieces.

Thanks again Tantarus/AirFlamesRed - this seems a much better approach. I have plenty more WIPs that I'll take a fresh look at now as well!

Tantarus - thanks again. I've read, and been given similar advice many times, but I seem to have a problem in actually following it! I still end up with overcomplicated meshes. You wouldn't believe how long it's taken me just to get this far, so I won't remodel from scratch - I think I'll follow your initial advice about tweaking the shape to match a reference image.

Airflamesred - "Is there any reason behind the morph route?" - it was just an idea I had. There are hundreds of free chess piece models available (far better than I could hope to make at present). But no morphing ones, at least none I could find. So I thought, "why not give it a shot?".

The real problem is therefore creating a single mesh that is capable of being transformed into any of the six chess pieces. The rook's castellations and the queen's crown effectively set the minimum number of vertices around the model's circumference, and I knew the knight was always going to be the biggest problem.

I suppose I could have started by modelling the knight and then trying to morph it to get the other pieces, but that seemed even more awkward.

(btw - my DAZ thread on the same issue if here http://forum.daz3d.com/viewtopic.php?p=2507837)

Thanks tantarus. So that's what the 'View>Background Image' is for!  I already had a reference photo, but hadn't realised I could actually set it to appear in the modeler - doh!

I'm very interested in your 'too dense a mesh' comment (please bear in mind that I am still a relative novice!). There are 1664 faces - each sphere for my starting point had 16 rings of vertices, with 32 vertices in each ring. This seemed to me, at that time, the minimum needed to get a good enough 'roundness' on the pawn, while allowing enough flexibility to manipulate the shape into the other pieces.

But looking at the result I see your point. If I was to remodel from scratch (maybe I will, maybe I won't!) what sort of face count (or number of rings/segments for the spheres) would you suggest?

Also, since the very top of the pawn became the tip of the knight's nose, the faces all converge, making the mesh even more dense here. Maybe I should slide some of the edge loops back from the nose area to the eye area? This would perhaps give me the extra faces I need to model the eye?

Interesting - I think I follow that one. Took me a while to spot that the triangle isn't actually a triangle.

I've also had some interesting replies to the same queries (mainly along the lines of Keep It Simple Stupid!) over on the DAZ forums here http://forum.daz3d.com/viewtopic.php?p=2342807

I like the idea of splitting it into separate objects and just stacking them - I'm trying that approach for my next attempt at a chunk of architecture.

I've been away for a few days resting my poor old brain, and just saw Midnightcarnival's question.

"sorry to high jack the thread" - not at all, the question's rather appropriate. And you seem to be getting some helpful answers - QED, posted in the right place!

"...having to get into making illustrations to demonstrate the problem..."*  - if you can explain a problem in words that's all you need, and there's plenty of really helpful people on these boards. (I post pictures because my words often puzzle people, myself included! ;o)

"...with freeware [clay-style] programs ... are primitive-based modeling programs still practical? ... what advantages are there in toughing it out with the more complicated to use programs?" - my advice (beginner to beginner) is to try to get to grips with the basics of as many different (free!) programs as you can, then they can all go in your toolkit. You'll inevitably end up preferring one or other, because they have the best fit to the way you think and work.

"I'm just doing this stuff for fun and learning"  - that's the acid test, when it stops being fun you know you're doing something wrong!