3D - Rejection - skewed & overfiltered?

Hi, I've just had a 2nd rejection on this set of orthographic computing diagram images. First time around it was due to the included clipping paths being too rough. The images are still exactly the same except rearranged & with better jpeg clipping paths.



1st image



Reasons for rejection:

*Screens / objects are all skewed upwards, top left of the object

*This image appears to be over-filtered/over-processed which has affected the image quality





2nd image



Reasons for rejection:

*Objects are all skewed upwards, top left of the object, they look twisted

*This image appears to be over-filtered/over-processed which has affected the image quality






I don't understand the skewed/twisted criticism. They're rendered in an orthographic projection, vital for making tidy network diagram layouts (otherwise you've got a mess of different perspective viewpoints that don't match).

As for overfiltering, I've added a small amount of contrast (+3%) & very small amount of noise (0.7% uniform noise in photoshop) to prevent jpeg artifacting on gradients, which is exactly the same post processing that I've used on most of my other accepted images.

Does anyone think these are worth resubmitting? They're made from laborious multipass renders so I really can't be bothered redoing them all with a perspective projection (which negates their usefulness as I've mentioned).

Thanks

(Edited on 2010-01-25 08:03:23 by mbortolino)

Hopefully Mark will be able to set that inspector straight. Those look fine to me, and are hardly "overfiltered". I'm impressed with the nice black outlines you're getting.

(I still think you've got too many in one image )

(Edited on 2010-01-25 08:03:27 by sjlocke)

I would agree with the objects looking skewed. The angels don't seem right. I can really see this in the laptop.
I disagree with them being less useful but I do know what you mean. Personally I'd re-render them as I think your images look really good and make for great stock. It might be worth your time.

The perspective just isn't right on these visually. I don't understand how any diagram requires making objects look as if they're on stretchy latex being pulled upwards at the top left corner.


I don't understand the overfiltering unless it's the slight jaggies on the diagonals that look (perhaps) like over sharpening. I don't see any banding or posterization anywhere.

These are orthographic renders. They are not meant to represent a true perspective vision. In fact, perspective (a foreshortening based on distance) is removed from the equation.

See below. The top is a camera with a 25mm lens. The bottom is the same camera changed to remove perspective.

I guess it's really just the desired effect you're looking for. I personally think the true perspective is a better option for those images rather than the orthographic versions. But that's just my opinion.

Here's a few quick layouts that show how the orthographic projections help make diagrams look neater:








In contrast, here's the vista icon set (drawn roughly from the same viewpoint but using perpective):





Doesn't look too bad I suppose, although a client that I do this kind of illustration for always prefers the non-perspective diagrams.

I see what you mean. I think they both look great. The perspectives also might be more sellable as you could use them stand alone (not in a diagram) and they would look natural.

Once you put the plane (via the dotted line) into the diagram the monitors no longer look tipped over. When you can't see that plane, they look wrong.


Is there some way you can include in your iStock submission the dotted lines (or something else that puts the plane in there) - not touching anything so it's easy to remove the individual pieces - to make the intended use more obvious?


And although I could imagine a problem with a "too similar" rejection, it would seem to me there's value in the collection having both versions. Perhaps mevans would provide some guidance on what the rules are with things like that?

Isometric projections are indeed a valid drawing method and are useful in diagrams and technical layouts.  Marcello, I sent you a sitemail.