sharpness issues... Still!!!

<p>hi again..<br>

went out and took a few photos -- with/without tripod auto/manual focus --- these are straight out of the camera--no editing done at all--i also have RAW versions available---to me i don't notice any difference in quality--so this mean i can rule out my technique for handheld shots?...</p>

<p>http://www.photo.net/photo/16645872<br>

http://www.photo.net/photo/16645865<br>

http://www.photo.net/photo/16645860<br>

http://www.photo.net/photo/16645859</p>

<p>think i messed up my links there--try these ones!</p>

<p>http://www.photo.net/photo/16645919<br>

http://www.photo.net/photo/16645918<br>

http://www.photo.net/photo/16645917<br>

http://www.photo.net/photo/16645916</p>

<p>This is still the wrong way to evaluate. You need to shoot a test target, as Mark suggested, in a wind-free, evenly lit environment, on a tripod with a remote release or timer. You also need to specify your post-processing, including any sharpening and clarity settings. Until you do that, it's going to be very difficult to evaluate anything.</p>

<p>The problem with looking at what you posted is I/we have no idea what it is you were focusing on, with a target such as the one I provided a link to, there are gradient / reference lines and point to which you or the camera which ever the case focus to, the gradient lines provide a reference to what is actually in focus.</p>

<p>hi Jeff, hi Mark<br>

will print out that test sheet you provided and test it over the weekend</p>

<p>cheers!</p>

<p>Tom, your posted images shows what looks like signs of noise reduction from the looks of the alternating spots of sharpness and softness especially in the grass.</p>

<p>Even if you were shooting Raw some converters apply their own default noise filter that can't be turned off especially evident in early versions of Adobe Camera Raw starting from 4.6 and back. Not sure about other Raw converters.</p>

<p>And if these are jpegs straight out of the camera there are settings to turn on/off noise filtering in camera, so you might need to check that. Some in camera filters are more aggressive than others depending on camera brand and model.</p>

 

 

<p>hi Tim<br>

i have gone through the menu settings and the only options i can find that might be related to what you are saying are:<br>

long exp noise reduction--set to auto<br>

high iso speed noise reduction-set to off<br>

These won't be causing my problem surely?..<br>

regards</p>

 

 

 

 

<p>hi again<br>

so i went back and took a few more photos--indoors -- all at <strong>17mm f5.6</strong>---had to put the<strong> iso to 1600</strong> to get a decent shutter speed --here are the results-- </p>

<p>http://www.photo.net/photo/16661286 -IS off,tripod, manual focus 17mm, f5.6 , 1/50 s<br>

http://www.photo.net/photo/16661288 -IS off,tripod, auto focus 17mm, f5.6, 1/50s</p>

<p>http://www.photo.net/photo/16661293 -hand held,IS on, auto focus, 17mm, f5.6, 1/50s<br>

http://www.photo.net/photo/16661296 -hand held,IS on, manual focus, 17mm, f5.6, 1/50s</p>

<p><strong> i also put my 50mm lens on the lens and took the following photos:</strong><br>

http://www.photo.net/photo/16661313 --tripod,autofocus, f5.6, 1/30s<br>

http://www.photo.net/photo/16661315 -- tripod,manual focus,f5.6, 1/30s<br>

http://www.photo.net/photo/16661316 -- handheld, autofocus, f5.6, 1/30s</p>

 

<p>None of these help. Nobody knows where you focused. There is no way to separate an autofocus error from another error. You need to shoot a target that makes it clear where you focused.</p>

<p>hi Jeff<br>

all the focus points on the camera are selected so i presume it chooses a few of them?...<br>

should i set the camera so that only the central focus point is selected , maybe place a bottle on the table in the middle of the picture , focus on that and shoot at say f5.6??</p>

<p>i am not sure how exactly to set this up....</p>

<p> </p>

<p>Tom, yes that is exactly what you should do. I think by letting the camera choose your focus points for you you may be creating your sharpness issues. The camera can't really know where you want to focus so it will pick the closest point, or just some random point to focus on. This may or may not be what you want in focus. I always choose the focus point for landscape type shots like you posted earlier. Use the auto mode for moving subjects such as runners coming towards you or birds in flight and no subjects closer to you. </p>

<p>Also when you compare the shots between the two lenses set the zoom to 50mm as well. Focus on the same subject using the same focus point. Then compare the images at the same aperture. But shoot a full set of images at each aperture for each lens.</p>

<p>Tom, if you're needing to use ISO 1600 to get a reasonably fast shutter speed at f/4-f/5.6 with a slow, variable aperture zoom, chances are the lighting is too dim for the AF to function well.</p>

<p>Even the best AF systems struggle with lenses having a maximum aperture of f/5.6, and the AF systems in entry and mid-level dSLRs may struggle at f/4 to focus accurately in dim lighting. (Note: this does not apply to faster lenses stopped down, such as your 50mm f/1.8 lens at f/5.6 - only the maximum aperture matters to the camera's AF sensor).</p>

<p>However I'm not seeing any particular problems with motion blur due to camera shake (your IS zoom should help there anyway), or with focus. The netting on the playpen appears to be in focus and reasonably free of motion blur.</p>

<p>It's mostly a matter of understanding depth of field as it applies to your camera's sensor size and lens focal length. Beyond that, it's up to post processing to get the desired look.</p>

<p>The problem is that unless you process the photo through the manufacturer's software (at least this works with Canon), you have no way of knowing what focus point was selected. Also, a bottle is a round target. A lens focuses at exactly one point. Put a flat target on a wall and focus on it with the camera on a tripod, and using either a remote release or the timer. Then examine that photo.</p>

<p>hi again--thanks so much for the replies<br>

L G - so for landscape photos it's best to have just the central focus point activated????-- where is best then to focus--1/3 into the scene, keep shutter depressed , recompose and shoot??----use between f5.6 to 8???--so in the photo i provided maybe if i focus on the red band on the top of the play pen and shoot the photo at somewhere between f5.6 and 8?--by focussing 1/3 way in i read that this is best to get most of the scene in focus?----or should i repeat the test with one focus point shooting at say the saucepan on the cooker???---not sure which way is best--maybe try both?-will try take the shot earlier when more light coming into the room</p>

<p>Lex--i will try the test again with better light--if light was not an issue could i expect to get the cooker, saucepans etc in focus (they look slightly blurred to me) at f5.6-8 if i focus say on the red band on the top of the play pen( i read that f5.6-8 is the best aperture for sharpness for my lens..) --- if it was a landscape scene would f5.6-8 be sufficient to get the max amount of detail in focus? -- i know going over that f number introduces distortion and "cancels out" any improvement from increasing the f number </p>

<p>Jeff--maybe use central focus point only and focus on a postcard on the wall?<br>

can use that instead of saucepan i suggested<br>

I used timer for all tripod shots</p>

<p>cheers folks!</p>

<p> </p>

<p>Tom, sounds like you're referring more to depth of field rather than to focus.</p>

<p>Regarding "sharpness", don't get too hung up on notions of using a lens only in its sweet spot of f/8 or so. While diffraction does occur at smaller apertures, it's not necessarily an unacceptable compromise. Some lenses still produce very good results stopped down to f/11 or more. Whether the diffraction is noticeable depends on the subject matter, enlargement size, viewer's proximity to the print, etc.</p>

<p>If you need the depth of field, the choices are to stop down or use focus stacking. The best technique depends on how much time you have, whether the subject or object is stationary, how much the lighting is changing, etc.</p>

<p>Take a look at this depth of field chart. Plug in the data for your camera (to determine sensor size), lens, aperture and distance: http://dofmaster.com/dofjs.html</p>

<p>For example, in the room photos you attached, I'd guesstimate that with the lens set to 17mm, you'd need an aperture of f/11 or f/16 with the hyperfocal setting at around 3 feet to get the entire area in reasonably sharp focus.</p>

<p>But with my tiny sensor digicam set to the equivalent focal length (6mm actual; 28mm equivalent), f/4 would give me the same apparent DOF with the hyperfocal setting of around 4 or 5 feet.</p>

<p>Tom, you should be able to use any focus point, not just the center. But the center does tend to offer the best performance on many cameras.<br>

Lex covered your question about depth of field pretty well. However I don't usually pull out the charts while I'm out taking pictures. If there is a distinct subject I will usually focus on that and stop down enough to get the whole subject in focus. If it's large like a building I'l focus about 1/3 of the way into the subject. But if it's a portrait I focus just on the eyes and stop down enough to get most of the face covered by the depth of field. But these are generalities. Use the LiveView on your camera to check if the important features are in focus at the chosen aperture. If not adjust the focus point and the f-stop.<br>

I think the best thing for you to do is to experiment. Either use a single focus point or manually focus with LiveView and see if you can get various objects in focus at different distances and apertures. Shot wide open on the 50mm lens you should be easily able to see the falloff on a round item like a bottle at close distances. Stop down a bit and watch the writing towards the back come into focus. It's a simple experiment, but it will show you a lot.</p>

<p>Lex--thanks a million for that chart---it really shows what is going to be in focus and what's not---by not selecting the central focus point and focusing about 4.47ft into the shot(hyperfocal point according to chart) at f/11 then all from 2.3ft to the wall should be in focus---presuming a fast enough shutter speed and tripod i should be then able to tell if there are issues with the lens or issues with "my technique/knowledge of DOF".....i found an app for my phone to give my the hyperfocal distance for a certain focal length, aperture -- seems ok to use it for the test, but not for every photo i am going to take!----do you use such charts/apps for most photos or does in come with time and experience how much will be in focus----maybe my initial "sharpness" issues are more to do with my lack of understanding of this dept of field calculator?---for e,g for the kitchen photos i provided at 17mm and f5.6 if for eg i focused at a point 3 foot in that would give me a 2.25ft dept of field, meaning items at the back of the kitchen are out of focus ( and not just unSHARP as i was referring to)<br>

so i think i should focus at a point at the hyperfocal point in the photo and then check the results<br>

But What if all focus points are selected? how does this affect the dept of field calculator?--one focus point might be on 3 ft in and another on 5ft in--how is the dept of field calculated then--does it only "work" if one focus point is selected in the menu??...<br>

L G --say you had a landscape scene with nothing of particular note in the foreground, just maybe rolling hills stretching to the horizon and wanted as much as possible in focus---going by the chart would i be right in saying at 17mm with my eos 450d and f11, i should focus 4.47ft into the scene, use focus lock, recompose and shoot?--that means (according to chart) that everything from 2.29ft to infinity is in focus--but what if i only focus 4.2 ft in--that will only give me a DOF of 67.3ft meaning the mountains/trees on horizon are out of focus--is the "trick" to always make sure i focus at 4.47 or over, never less?--as this seems to keep the far level of acceptable sharpness at infinity -so for example look at the chart, see 4.47ft and focus at say 7 ft to be sure?---is that over-complicating it all?--what do you do?<br>

You mention LiveView--never used this function on my camera--thought it drained the battery...not sure if i actually have it-will look tomorrow when doing some more experiments..--but if it shows you how the DOF changes on the screen as you change the aperture then this is great!--i presume it would work great for my kitchen scene but maybe not so well for a landscape photo?</p>

<p>cheers again!</p>

<p> </p>

<blockquote>

<p>it really shows what is going to be in focus and what's not...ll from 2.3ft to the wall should be in focus</p>

</blockquote>

<p> <br>

This is a misunderstanding. Only one distance point is in focus. Everything in front of and behind that point is out of focus. What changing the aperture does is change how much out of focus other distances are, but they will never be in focus.</p>

 

<p>hi Jeff<br>

so for the kitchen photo if i took two photos , one at f11 one at f22 should i notice a big change how out of focus items are?--according to the chart at those settings the far limit of acceptable sharpness is at infinity for both?? (focusing at the hyperfocal point)</p>

 

<p>Tom, what Jeff is saying is there is only one plane where the lens is perfectly in focus. Everything else is to some degree blurred. But in reality with the size of the sensor elements (or film grain) at small apertures and the faults in the lens and diffraction much of the image will look to be just as in focus as the plane its focused on.<br>

As you get smaller apertures more of the image looks to be equally sharp. But the plane of focus at f/22 will look blurred relative to the lens at an optimal aperture. It just becomes a balancing game.<br>

As far as using hyperfocal distances, you need to keep in mind the intended size of the print. The charts are usually setup to make an 8x10 print and have it viewed at arms length. There are a lot of assumptions in there that don't apply to the way I make prints. So using a standard hyperfocal chart (the ones engraved on my manual focus lenses) I find I really need to stop down an extra two stops. This will result in more blur from diffraction, but that's frequently better than having an even blurrier subject.</p>

<p>In you example with nothing in the foreground I would probably try to focus more towards infinity and frame the image so nothing in it was closer than say 20 feet. Then I could use a larger aperture (lower f-stop) and get the main subject in sharper focus. Of course I would also start thinking that images with nothing of interest in the foreground are frequently boring, so I may look for another composition. But that's another discussion.</p>

<p>hi again....<br>

so i took some more photos this morning---just not enough light/lens not fast enough to hand hold the camera at the various aperatures--i took 3 photos with my 50mm lens--f5.6,8, 11 and 3 with my 17-85mm lens (set to 50mm)--f5.6,8, 11 -- here are the results--all taken with central focus point only activated--focusing just under the second dial from the left on the cooker--all on tripod, no IS--needed iso 1600 to get a half decent shutter speed<br>

To my untrained eye i don't notice a major difference between the lens--what do you guys think?</p>

<p>http://www.photo.net/photo/16664952 17-85 lens ,f 5.6-looks a bit sharper than below<br>

http://www.photo.net/photo/16664957 50mm lens ,f 5.6</p>

<p>http://www.photo.net/photo/16664953 17-85 lens ,f 8-don't see any difference in below<br>

http://www.photo.net/photo/16664956 50mm lens ,f 8</p>

<p>http://www.photo.net/photo/16664954 17-85 lens,f 11-don't see any difference in below<br>

http://www.photo.net/photo/16664955 50 mm lens ,f 11</p>

<p> </p>

Back in the old days before camera makers made things easier, photographers would look at their lens to see what was going on.<P>

<center><img src="http://jdainis.com/hassey2.jpg"></center>

 

<P>The lens on the left has been set to f/22 and focused to infinity. On the scale, 18 (feet) is above the 22 (f/22) on the left of the scale. At f/22 everything will be in acceptable focus from infinity to 18 feet. On the lens on the right, the the focus has been adjusted so the infinity symbol falls above 22 on the right of the scale and now 9 feet is above 22 on the left side of the lens. The hyperfocal distance has now been set by doing this. Everything will be in acceptable focus from 9 feet to infinity. <P>

 

Many people would zone focus at parties or dimly lit areas where it was hard to focus in the viewfinder (or now auto focus hunts). People are about 10 feet away? Move 10 to the center pointer, bring up the camera, compose and shoot. The lens markings would show the near to far acceptable focus for the aperture set. What could be easier or faster? <P>

 

Many people would use cameras such as Kodak Instamatics that were fixed focus. Nevertheless the images were in acceptable focus from 4 feet to infinity given the quality of the cheap lenses. One could call that fixed zone focus.

<p>Tom, 50mm is suppose to be very sharp lens. I took one of your pictures in photoshop, adjusted the level and curves and used smart sharpen. The result is below. It appears quite sharp to me. The problem with original pictures is they lack proper contrast. Try some post processing.</p><div></div>

<p>thanks Sandeep<br>

i have altered the internal settings in my camera to increase contrast , saturation, sharpness<br>

going on holidays soon so looking forward to checking the outcome<br>

but will also play around with photoshop to alter the contrast as needed<br>

currently saving up for the canon 17-55mm lens </p>

<p>Tom,<br>

Other thing to keep in mind is exposure. If your picture is overexposed by even 1/3 stop, you will lose contrast. In this scene there is lots of black so camera meter may be over exposing by a bit. Just like for snow you have to expose by almost +2 stops from what meter tells you, it is exact opposite for blacks (you need to under expose). You can try same scene again and try under exposure of 1/3, 1/6 or -1 to see if picture matches up with above. Just wanted to drop some hints without getting into too much details since this is beginner's forum.</p>