A 28mm lens remains a 28mm lens regardless of the format it is used on. What changes is the field of view.

As an example, a 24mm lens would be a standard on 110 format, a modest wide angle on a DSLR, an wide angle on 35mm and an incredible ultra wide angle on 6x6cm

Glad everyone is having a good time!

Just imagine how fantastic a design would be where the removable 645 insert was replaced by a digital module that included the CCD and its electronics...Then the module could be returned for upgrade as required, or even service-exchanged...

The 24-90mm also suffers at the wide end, but in practice, with careful composition, is useable for architecture. Mind you, there are degrees of requirement even there, depending upon the final use.

Unfortunately, shooting for Images of England, software solutions are not allowed, so the lens and the use of it become very important.

For other purposes, some defects can be made into advantages. For example, David Hamilton did very well out of poor lenses throughout the 1970s.

This thread has brought out the valid point that lenses do have defects and being aware of them enables us to make good choices.

I've just looked at the photodo review of the Sigma, and the first thing that struck me was that the barrel distortion at the wide end is atrocious - over 5% is just not acceptable for the sort of pictures I take.

The pentax 16-45mm is not perfect in its drawing either, but it is much more reasonable and quite useable for architecture.

Even the 12-24mm, where some distortion has to be accepted, is fine for architectural subjects.

For general pictorial use where the distortion can even be used to effect, it may be another matter.

This is where a site like photodo is so useful - we need to know the design philosophy behind a lens to know if it is what we are looking for.

I only have the 16-45mm so I can't comment on the comparison being made, but as a generalised comment I have noticed from published MTF graphs that Sigma lenses are geared very much towards high contrast at the expense of fine resolution. Higb contrast always enhances apparent sharpness, especially where there are edges. You can see this effect very clearly if you watch what happens at pixel level with Unsharp Mask.

What I am saying is that there is a lot more to lens design and quality than a casual assessment of overall "sharpness", the very word in itself being delightfully vague, certainly in scientific terms.

So after the initial sharpness, how is the bokeh, the resolution of fine detail, the pictorial quality in both low and high contrast subjects?

Both the lenses under discussion will have strengths and weaknesses, as no lens is perfect.

That assumes that backwars compatibility is a more expensive option, and it might not be. The K mount was a good modern design to begin with, has served well and has proved to be adaptable enough so far.

I've had a look at the pictures, but they don't prove anything unfortunately. It isn't a very scientific test, more a couple of casual shots.

If a lens test is being undertaken at least start with using a tripod!

Lens coating has been around since the 1940s, possibly earlier, but certainly post war lenses are often factory coated.

Multicoating was known before the SMC coating, but Pentax put it on the map with their new SMC 7 layer coating process. It is highly effective and a good reason to use Pentax lenses.

The most amazing coating I ever saw was on the Rolleiflex 2.8GX that I once had. You could literally not see the lens and I had to save it from many a poking finger that was headed for the front element with the words "Is there no lens in it?"

The natural plane of focus for a lens is curved, which can be clearly seen in many cheap cameras. As price increases steps are taken to make the focus point fall on the film or CCD at all points, but this will break down at wide apertures and close distances for many modern zooms.

Try photographing a sheet of newspaper with a cheap zoom at open aperture and see the edges go out of focus when the centre is sharp.

Small errors are covered by the "depth of focus" which allows the position of the CCD to be slightly in front or behind without losing too much sharpness (compare with depth of field).

A Macro lens will be designed with a flat field, making it more suitable for document copying and other critical uses.

For distant objects there is not so much of a problem design-wise.

I think it was Minolta who made a Variable Field lens which enabled alteration of where the plane of focus lay. Very useful in very obscure circumstances, and it never caught on.