View Full Version : Sweet Spot of Sharpness


Alain St-Amour
February 13th, 2009, 06:08 AM
I've come across a few posts that indicate there is an optimum f-stop range where the XH A1 performs best, where the picture is clearest and free of grainyness. I believe Marcel D. Van Someren had made a reference to it in the huge custom presets thread.

What is the optimum range?

Marcel D. Van Someren
February 13th, 2009, 07:25 AM
since my ears were burning here, my understanding is that f4 -f5.6 is the sweetspot for the A1.

Tom Hardwick
February 13th, 2009, 09:11 AM
All lenses have a sweet spot though being 'free of grainyness' isn't an optical term. Very nearly all lenses are designed to work at their best fully open. They don't for mechanical reasons of element centering and internal flare, so stopping down a couple of stops (to f/4 for an f/2 lens, say) will have effectively removed all the vignetting , not gone into noticeable diffraction and have cleaned up the internal flare problem.

tom.

Alain St-Amour
February 13th, 2009, 09:40 AM
I was hoping the range would be wider, but just being aware of this will help improve my video and I'm grateful for that. Thank you Marcel and Tom.

Steve Phillipps
February 13th, 2009, 03:26 PM
Very nearly all lenses are designed to work at their best fully open.
tom.

Not really, if you look at MTFs of most stills lenses for example wide open performance tends to be not great, improving 1 stop down then more for the middle aperture then tailing off slowly then rapidly beyond f16.

The main problem for really small sensors is diffraction, where a small lens opening scatters the light rather than focussing it (sort of!), the smaller the sensor/film the earlier this occurs. Can't remember the specifics but with 35mm film it's something like f8, 6x6 film it's f11 and 8x10 film it'll still be OK at f32. With 1/3" chips it's probably f5.6 at most, yet the lens won't be that great until stopped down maybe 1 1/2 stops so you're not left much room to maneouvre! I'd aim for f4-8 as a good compromise.
Steve

Tom Hardwick
February 14th, 2009, 02:14 AM
Not really, if you look at MTFs of most stills lenses for example wide open performance tends to be not great,

You didn't read my post Steve. I said they were designed to work at their best fully open, and I stand by that. The fact that mechanical imperfections (miss-centering of elements, internal flare from the black cylinder etc cause the lens to work better a stop or so closed down is just a fact of engineering life.

Diffraction is certainly evident on HD 1"/3 chips at f/5.6 (remember they're not as big as 1"/3 4:3 chips).

tom.

Denise Wall
February 14th, 2009, 08:01 AM
You didn't read my post Steve. I said they were designed to work at their best fully open, and I stand by that. The fact that mechanical imperfections (miss-centering of elements, internal flare from the black cylinder etc cause the lens to work better a stop or so closed down is just a fact of engineering life.

Diffraction is certainly evident on HD 1"/3 chips at f/5.6 (remember they're not as big as 1"/3 4:3 chips).

tom.

I don't get what you mean by "work at their best" then. Most still camera lenses are sharpest somewhere around two stops above their widest aperture. If the OP means sharpness by "sweet spot," what are you saying?

Steve Phillipps
February 14th, 2009, 08:58 AM
My question too Denise!
I'd say they're just designed to be as good as they can be in general, and optical laws seem to dictate this won't be at very wide apertures, particularly for zooms. The only excpetion seems to be wide aperture telephotos which often do perform best near wide open (but even these tend to improve when stopped down about 1/2 to 1 stop).
Steve

Tom Hardwick
February 15th, 2009, 03:05 PM
Spot on Steve - designed to be as good as they can be. Diffraction losses occur as soon as the lens is stopped down, but (if we take an f/2 lens, say), it will vignette at least until f/4 and internal flare will degrade the image at wide apertures too.

Zooms have more flare than primes because generally they have very many more elements and there are internal mechanical rods and so on (for the moving groups) that reflect and scatter light unevenly.

So we accept the diffraction losses at f/4 simply because the vignetting, miss-centering and flare losses are more harmful to the image quality wide open.

tom.