F8 and subject sharpness

[davidclick]

<p>Good Morning from York UK..<br>

I wanted to resolve something once and for with a field experiment. Before I conducted the field experiment I had it in my head that all images are sharpest when shot at f8.<br>

In this filed experiment i wanted the best aperture to get the statues face sharpest.<br>

So I stood 2 metres away from a statue and here are the results. The first is at f2 the next f10 but i can spot no difference in image sharpness when looking at the face.<br>

<a href="http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_700_f10.jpg">http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_700_f10.jpg</a><br>

<a href="http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_f2_700.jpg">http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_f2_700.jpg</a><br>

So I wonder is the reason the face of the statue has no discernable difference in image sharpness is down to the fact that i was standing 2 meteres away from the subject. In other words if i stood alot closer i would see the difference?<br>

Thanks in advance :-)</p><div></div>

[newmanuk]

Hi David. I checked the 2 images. and there seems to be a background difference i.e the windows of the building at rear. was this a total manual shot? Time for others to chip in! regards miken

[davidclick]

<p>I can see the background difference but I'm not interested in the background. I am only interested in the statues face sharpness so I wonder aperture is a depth a field concern, the actual subject remains sharp at any aperture and there is no difference.</p>

[ted_raper1]

<p>The "all images are sharpest at F8" thing is a generalization, and will likely vary with the lens being used. Most lenses are sharpest two or three stops down from their maximum aperture, and depending on what that is, max sharpness could be F4, 5.6, or 8. I have a Panasonic micro 4/3 lens, a 20mm F1.7, that is actually sharpest around 2.8.</p>

<p>The only way to tell the sharpest aperture of a given lens, in my experience, is to test it yourself with a good steady tripod and controlled conditions.</p>

[kevin_delson]

<p>As already mentioned, you will need a more controlled study to determine the "Sweet Spot" for a particular lens.</p>

<p>The images posted are of little help, viewed at 100% and some pixel peeping is required.</p>

<p>You do however make a point that relates to image quality yet is NOT associated with optical quality of your lens...</p>

 

<blockquote>

<p>In other words if i stood alot closer i would see the difference?</p>

 

</blockquote>

<p>This relates to the angular image size projected upon the sensor; another topic entirely.</p>

<p> </p>

[paddler4]

<p>If what you are looking for is effects of diffraction, you may have found a not-very-well-kept secret: the practical effects of diffraction are often far smaller than postings would have you imagine. Unless you are blowing up to a huge size or pixel peeping, you are not likely to see much of an impact of small changes in aperture. If you search, you will find a number of other tests of this.</p>

[Dieter Schaefer]

<p>Not sure which 35mm lens you were using - but if you look at the MTF date of the Nikkor 35/2 AF-D<br>

http://www.photozone.de/nikon--nikkor-aps-c-lens-tests/214-nikkor-af-35mm-f2-d-review--test-report?start=1<br>

then you will see that the differences for the center of the image are small and that by the time the borders catch up at f/8, the center is already dropping a little bit. Only using well controlled setup and pixel peeping will you be able to see the differences in the center - it should be more obvious at the borders.<br>

For the Nikkor 35/1.8 AF-S, it's even more pronounced - and from the data, it appears that f/4 is "sharpest". http://www.photozone.de/nikon--nikkor-aps-c-lens-tests/422-nikkor_35_18g?start=1</p>

<p>The Zeiss Distagon 35/2 shows a characteristic more like the Nikkor 35/2 - albeit at a higher level. http://www.photozone.de/nikon--nikkor-aps-c-lens-tests/258-zeiss-distagon-zf-t-35mm-f2-review--test-report?start=1 f/2.8 would appear to be sharpest in the center, and the borders catch up at f/8.</p>

 

[starvy]

<p>David,<br>

Hello from South Wales! The 'f8 and be there' is perhaps something of a misnomer in this day and age. The lens construction is so diverse across the range of systems we use that some are perhaps sharper at different stops.<br>

With my old point and shoot Canon S30 from 2001 I find the sharpness to be equal from 2.8 to f8 (the full range) when the subject is close to a background. With my basic Sony A100 DSLR, when mounting the Minolta PF 58mm 1.4 lens (from 1963?!?), it is at its sharpest around f11 with a minimum f stop being f16. The Minolta AF 28mm 2.8 lens is very sharp at f16 with minimum being f22. However, this lens is acceptably sharp from f4.</p>

[andrewg_ny]

<p>Again, more generalizations but you'll probably also find that in the center of the frame most primes are already near peak sharpness at f/4 or faster and many zooms perhaps one stop later (by f/5.6)...but the edges & corners take a little longer to reach their peaks, often as late as f/11. The other concern of course is what are you shooting--is it flat like a brick wall--does it match the lens's plane of focus? (of course most lenses have some focus field curvature, so a flat surface may not be 100% ideal) If your scene is not, then depth of field likely comes into play and may play a more important role than absolute lens sharpness in real world images.</p>

<p>I don't think the old adage 'f/8 and be there' is necessarily about lens sharpness--it's more of a reminder that getting the shot is the most important thing--so if you don't have time, pick a middle-of-the-road aperture to provide hopefully adequate depth-of-field, get at least a usable if not optimum exposure.</p>

[davidclick]

<p>Thank you so much for your replies but I think Andrew has touched on something and I need to refine my question so here goes:<br>

If I pictured a brick (2 metres away) wall straight on with my 35mm prime DX lense on the end of my Nikon D200 in well lit conditions that did not vary between the shots would the bricks be any sharper at say at f2 & f11, would the grain of the bricks be any sharper between the two?<br>

Thanks in advance :-)</p>

 

[Dieter Schaefer]

<p>For the 35/1.8 at f/2 and f/8 there will be very little discernible difference to the eye - except in the corners where resolution will be lower at f/2. At f/11, due to diffraction effects, the image will have less resolution overall - the difference to the image at f/8 is still small though. You probably see the biggest difference if you compare to shots taken at f/4. Field curvature at large apertures could be a problem - but more likely any differences are either due to not having the camera sensor parallel to the wall.</p>

[jim_momary]

<p>35mm F1.4 AiS, at F8, 1/60th sec, lens on a Nikon D200, tripod, about 20 ft from a wall.<br>

Just more extraneous data.<br>

Jim</p>

[jim_momary]

<p>oops</p>

<div></div>

[kari v]

<p>If you can't spot any difference between f2 and f8 congrats you have a very good lens. :)</p>

<p>However, I think you'll find some differences when you shoot RAW and sharpen a bit more aggressively than camera default, especially when extracting detail by low radius sharpening (or "detail" slider in Lightroom). If your lens is wide open at f2 (or almost in the case of 35/1.8) you should see a difference in 100% view. Even if the lens retains very good detail at large apertures there's usually at least slight "mist" over the image compared to, say, f4 or 5.6. In web sized images you provided it's of course impossible to see anything, images in this size should be pretty much perfect anyway unless there's something really wrong.</p>

<p>As others mentioned, "sharpest at f8" is just for being sure. Only in the case of poor quality tele zooms (like the copious cheap 70-300 variants) and older slow kit zooms you should try f8-11 for best results. For primes middle apertures f4-5.6 are usually the sweet spot. For example my 5D + EF 50/1.4 produce absolutely stellar results there (so do my film bodies for that matter) even though the results at f2 are already very good and depending on the subject and light there might be little to choose between them.</p>

<p>Care to post 100% crops of the statue's head?</p>

[kari v]

<p>Here's a test chart for the $100 EF 50/1.8 on 5DmkII, take a look what happens at f4:<br /> <a href="http://www.photozone.de/canon_eos_ff/415-canon_50_18_ff?start=1">http://www.photozone.de/canon_eos_ff/415-canon_50_18_ff?start=1</a></p>

<p>Here's EF 50/1.4 on 15Mp APS-C:<br /> <a href="http://www.photozone.de/canon-eos/565-canon50f14apsc?start=1">http://www.photozone.de/canon-eos/565-canon50f14apsc?start=1</a><br /> Very little difference between f2.8 and f8 but in real world DoF will of course play an important role.</p>

<p>While I find that taking tests and charts too seriously is rather counter productive, here the findings agree with my eye very well. Basic primes tend to behave like this, that is, stop down one stop and you're good to go (most hazy aberrations are gone which makes the image appear a lot sharper, EF 50/1.4 is "dreamy" wide open, at f2 it looks like totally different lens), stop down to middle apertured and you're golden.<br>

But does the sweet spot matter in the real world? Depends. I have absolutely no problem shooting 50/1.4 wide open when I need it or want it. Sometimes sharpness is highly overrated and shooting wide open may produce desirable qualities in the image. Also, resolution and perceived sharpness are two very different things, 50/1.4 can look stupendously sharp wide open depending on how you use it.</p>

[kari v]

<p>I'm feeling a bit talkative today it seems...</p>

<p>I feel that statues as subjects need very little resolution the look good. As a natural formation rocks (and trees and grass) take happily all the resolution there is but statues are perceived a bit differently, their overall form overrides the details unless you really set your mind to look for softness etc. Or perhaps it's just me. :D More coffee...</p>

[davidclick]

<p>Thank you so much for your comprehensive replies :-) Jim the brick shot v help full & everyone else.<br>

Kari heres the statues head as big as i can get:<br /><a href="http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_f2_head-2.jpg">http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_f2_head-2.jpg</a><br>

<a href="http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_f10-head.jpg">http://i216.photobucket.com/albums/cc53/zymurgy_bucket/constantine_f10-head.jpg</a><br>

I think i get the whole sharpness issue but I think i need to go out into the filed again and see the results for my self. A nice old York gritty second century roman wall should do it :-)</p><div></div>

[kari v]

<p>That looks like more than 100% crop or resized from a small file. How are you editing these? You should open the full size jpg, zoom in to 100% (or 1:1), then crop a piece and save as...</p>

[scott_ferris]

<p>The reason you have absolutely no detail in either image is because you are shooting at 1600iso! With a 35mm lens a 1/50 sec is handholdable, take your shutter speed down from the very high 1/320 sec @ f10 or the crazy fast 1/6400 sec @ f2.</p>

<p>A normal setting for this kind of image would be f8, 1/50, 100iso. That will give you a huge amount more detail.</p>

[William Michael]

<p>Interesting thread. Expansions on three points:</p>

<p>F/8 as related to the (aperture) <strong><em>sweet spot of a lens:</em></strong><br />Is a generalization, but as generalization go - a reasonable one to suit most DSLR lenses – especially the lenses used with (bought with) prosumer or entry level DSLRs (kit zooms spring to mind).<br />Perhaps a better generalization is the (aperture) sweet spot: “is about three stops closed down”.<br />But note I use the modifier “aperture” for the sweet spot – remember that for zoom lenses there is usually also a zoom FL sweet spot.<br />And for Varying Maximum Aperture Zooms – the aperture sweet spot will likely change, as the FL changes.</p>

<p>***</p>

<p><strong><em>“F/8 and be there”:</em></strong><br />Has nothing to do with and Aperture sweet spot – at least that is what we teach in Photography History – maybe we are wrong?<br />This is a phrase originally used by Press Photographers and the origin of it is under dispute, but none the less the meaning is pretty much agreed – that it is related to <strong><em>Shooting Distance.</em></strong><br />It is widely believed that originally it related to using a Speed Graphic and Bulb Flash and computing the DISTANCE for Manual Bulb Flash - F/8 gave the “best” (most suitable) Shooting Distance for Press Photography. <br />Later, when I worked for a newspaper (circa 1975 - 8) - I was using 135 format SLR and newspaper hacks who shouted that phrase at me were meaning <strong>F/8 for use on a 35mm lens </strong>(most favoured by Press Photographers) - and agin this still related to SHOOTING DISTANCE (and if you will likewise DoF).<br />At F/8 (35mm lens) the Hyper-focal Distance is around 20ft. So setting the focussed distance (using the lens turret scale) at around 15ft one was assured of getting a good DoF and the subject in focus for doorstop shooting. Also F/8 was still useful – for that approximate 15ft SD, for <strong><em>manual</em></strong> Flash Fill – which is what we used on our 35mm cameras.<br />The point is the "best"shooting distance for Press Photography hasn't changed much - it is not that is it the "best" distance, but rather it is the distance one finds oneself at the most: at door stops; in the front row of a press conference: at the egde of a street rally; etc - about 15ft - plus or minus a bit - so hence "F/8 and be there" and it is all "in shot" with a 35mm lens on a 135 format camera - crop what is required later.<br />I still use “F/8 and be there” when using manual focus (widish) lenses especially for hip shooting.<br />Occasionally I use “F/8 and be there” with AF lenses, hip shooting – but I have pretty much nailed my technique to centre the subject from the hip, and use centre point AF when using a 24mm or 35mm lenses at reasonably close range – about 8ft to 15 ft.</p>

<p>*</p>

<p><strong>Re testing things: </strong><br />I acknowledge my good mate Scott mentions that Hand Holding @ 1/50s with a 35mm lens loaded is OK - and for a quick field test I agree. <br />But if you want to <strong>TEST</strong>, then you will use a tripod - and mirror up - and a remote release - and certainly not ISO1600.</p>

<p>WW</p>

<p> </p>