Why are my pictures not sharp? Part II

[dan_tripp]

<p>Everyone thanks for your posts last week. I did a few experiments with my camera on a tripod, and I don't think my K-7 or Pentax 50-135 f/2.8 has front or back focus problems, at least from wide at 50mm. I'll post a few test shots below. My question about sharpness is still there and I want to know if this is a Pentax issue or do all cameras lack sharpness when the subject is a few feet away. I seem to not have a problem when I do a full head shot, but when I do a body shot, the face ALWAYS looks out of focus, fuzzy, and no sharpness. I read from someone before that Pentax is inferior when it comes to details on people's faces, is this true?<br>

Here are some sample shots.</p><div></div>

[zane1664879013]

<p><em>...Pentax is inferior when it comes to details on people's faces, is this true?</em></p>

<p>No, it's not.</p>

[dan_tripp]

<p>Here is the full body shot.</p><div></div>

[dan_tripp]

<p>Another example. Note all these pictures were in RAW and I did no editing.</p><div></div>

[dan_tripp]

<p>Full body shot</p><div></div>

[dan_tripp]

<p>Experiment shots.</p><div></div>

[dan_tripp]

<p>Full experiment shot.</p><div></div>

[mountainvisions]

<p>Not really, I spend quite a bit of time removing details from faces. Often the individual hair tips on freshly shaven faces of mens portraits are frighteningly sharp. And crows feet on the eyes are also pretty awful. It's precisely why I don't like shooting portraits, a lot of work with masking and such to keep sharp areas sharp, and unflattering areas naturally soft. I suppose if I was lazy I could just apply Digital Gem Airbrush at full strength, but I never liked the overly soft plastic skin look.</p>

<p>I'm not sure where you are getting this info from, this reminds me of the 40D had a 3 stop advantage over the K10D, still makes me laugh! But it seems like you are looking to move on from Pentax. Nothing wrong with moving on at all but I'm still confused on why someone who seems to understand photography quite well is getting such poor image quality. If you are doing everything right, then you have an equipment issue, if you are not doing everything right than it might be an expensive way to find out your technique is lacking. Canon/Nikon/Leica won't fix technique if that is the issue (and it doesn't sound like it is).</p>

<p> </p>

[dan_tripp]

<p>As far as my experiment, I put my camera on a tripod and stuck tacks on my cards every 9 inches apart. I focused on the center card. I tried different f-stops and they all were the same results. I did NOT try different focal lengths. Could that make a difference? I did read somewhere that people have front/back focus problems when they are a few meters away. Like I said earlier, my close up shots are usually ok, except on my Tamron 28-75. For some reason those close ups always lack sharpness.</p>

[dan_tripp]

<p>So Justin, do you think this is equipment issue? I don't want to switch I really like Pentax. What is the next step to find out if this is equipment or user?<br>

Do the above close ups look normal for Pentax?</p>

<p>Thanks for your help?</p>

[jeremy_corbin]

<p>As a last ditch effort, maybe sending your 50-135 in to CRIS to get it calibrated might help. If one of the elements is off center or out of alignment (even slightly) that might explain things, but it does seem odd. The flowers in the full body shot of the ladies looks like the focus point, but it's hard to tell on my netbook whats what. At f/2.8 if the focus point was the flowers and not the face, than that would explain that shot's OOF area... but this is guesswork.</p>

[dan_tripp]

<p>Jeremiah, thanks, I ALWAYS focus on the face/eyes, unless it is action shots. I use center focus, wait for the beep, then compose my picture before I take the shot.</p>

<p>Here is an example of a closeup.</p><div></div>

[dan_tripp]

<p>Full head shot.</p><div></div>

[dan_tripp]

<p>Jeremiah, here is a closeup of the flowers, I think you may be right. What is CRIS and should there be a warranty on my lens? I purchased it from Amazon.</p><div></div>

[mountainvisions]

<p>Dan,</p>

<p>Those shots definitely don't look normal.</p>

<p>I'll look around my files and crop some facial closeups, but it ain't pretty so to speak. Much more detail than necessary most of the time. </p>

<p>Also reading your reply to Jeremy I see a (or the) problem.</p>

<p>You can't focus on the eye and recompose using a large aperture closeup.</p>

<p>The reason is that even a few inches will throw the zone of focus way off.</p>

<p>if you are doing this you need to use the focus points, or manually focus using the ground glass.</p>

<p>I prefer center point focus for speed, accuracy and simplicity, but this is one of those times that you need to use the 11 points or manually touch up the focus.</p>

<p>Hope that helps!</p>

<p> </p>

[mountainvisions]

<p>Just to add to what Jeremy and I wrote.</p>

<p>Jeremy noted the focus point appears to be on the flowers. Lets say you focused on the eyes via center point, than recomposed, just a slight difference might move you DOF a few inches front or back, making the faces just enough out of the zone of focus.</p>

<p>I'd recommend 1) going to a slightly smaller DOF (like f/4 or f/5.6) 2) using the focus points.</p>

<p>I will bet if what I think is going on is the case, you start cursing the sharpness as a negative.</p>

[pisq]

<p>I would say it was the lens. you are shooting wide open, the girl's blue eyes, and the business cards look sharp though. I missed if you had tried smaller aperture, say, f/5.6. a small amount of sharpening will help these images, although 100% crops will only look marginally different.</p>

[michael_elenko]

<p>Dan,<br>

This is some missing data that would help me help you. I'll repeat this from part two: was SR activated or not when you shot on the tripod? Did you use MLU? This may make a difference.</p>

<p>Secondly, I just wouldn't take those shots for a focus test shooting at f 2.8. The DOF is too thin. In the real world I won't shoot people shots at that aperture, especially in daylight unless I'm hindered. I think shooting at f 4.5 to 6.7 would make a significant difference on a few shots. So I challenge your test methodology.</p>

<p>Thirdly, it is almost useless to evaluate sharpness in a straight unedited and un-processed RAW file unless you have another lens setup exactly to compare it to. RAW files are soft by their very nature. And that's the only point I see demonstrated here.</p>

<p>Fourth, I would not leap to technical conclusions about the quality of Pentax or any other brand lenses here. There's no reasonable logic that would lead there; plus some of us do quite OK using the same equipment.</p>

<p>Fifth, how do you know that your lens doesn't require front or back focus adjustments? What formal diagnostic testing have you performed? You camera has some powerful and effective individual lens adjustment capabilities, but to do it correctly requires an accomplishable, but strict lens testing protocol. Basically you aim at a diagnostic test chart at a 45-degree angle and perform a series of exposures. Google is your friend here. I've performed this on 6 of my lenses and found it invaluable. Ended up doing serious tweaks to the Tamron 17-50mm, less serious ones to the FA 50mm f1.4, and a very minor one to the DA-50-135mm in question here.</p>

<p>Sixth, is it possible for you to provide us with a few of the RAW files to experiment with?</p>

<p>Thanks,</p>

<p>ME</p>

[dan_tripp]

<p>ME, thanks for your input. I'll try to answer some questions:<br>

1. the test shot with the cards the SR was off. Not sure what MLU is<br>

2. I'm not sure if I can agree with f/2.8 being too thin. I know many wedding photographers who shoot wide open. My friend shoots mostly primes and wide open most of the time. Her work is amazing. http://untamedheartphotography.com/blog/<br>

3. I'll attach the RAW file<br>

4. ?<br>

5. I don't know my lens does not need adjustment...I thought I did before this post. :-) So how do I make adjustments to my lens...that sounds exciting. I'll Google "diagnostic test chart" I hope I don't screw something up.<br>

6. I'm attaching, thanks.</p>

 

[dan_tripp]

<p>Here is the link to download the first RAW file. I can't thank you enough. </p>

<p><a href="https://docs.google.com/leaf?id=0B7ND8aBQ_BXGM2U5YWU5YTMtZWEyZi00MmY4LTliZDItOGQwNmFmNjg4YmQw&hl=en">https://docs.google.com/leaf?id=0B7ND8aBQ_BXGM2U5YWU5YTMtZWEyZi00MmY4LTliZDItOGQwNmFmNjg4YmQw&hl=en</a></p>

<p> </p>

[jeremy_corbin]

<p>Dan, <a href="https://www.criscam.com/newsite/index.php">CRIScam</a> is the Pentax Authorized Service Center in the US after Pentax quit handling the repairs in Golden, CO. I'm not sure if they are the only Authorized Center in the US or not. One of the other guys might know more.</p>

<p>When I have used center point focus can recomposed I've had mild success, but I do note a difference between what I think is in focus and what the camera believes is in focus when I use that method. I would try using the selective focus method to see if that changes your results.</p>

<p>I'll try to play with the raw file tomorrow when i crank the desktop back up. Its loud enough it keeps the wife awake at night....180mm fans do that....</p>

[michael_elenko]

<p>Dan,<br>

You are welcome.<br>

-MLU =Mirror LockUp and it is automatically activated on the 2-second delayed shutter function that is intended to reduce camera vibration. It also disables SR. Pretty clever.<br>

-Re: f2.8, let's get the context for these pictures well understood. You are testing for sharpness in your posted shots; your friend, who is indeed a primo shooter, is exploiting the dreaminess of a thin DOF for a romantic effect. The difference between the planes of sharpness is intentional and makes her shots. Is she also, by chance, using a camera with 135mm-sized ("FF") sensor like a Canon 5D or a Nikon D700? Those cameras produce better separation. Plus she's performed post-processing, probably lots of it, whereas you have posted unedited RAW images. There's a difference there beyond aperture.<br>

-Thanks for posting the RAW file; I'll try to find the time to explore and play.<br>

--Focus testing: Photo.net's Bob Atkins has a good <a href="../learn/focustest/">approach</a>; I have the <a href="http://jimdoty.com/Tips/Equipment/USAF_Test/usaf_test.html">USAF focus chart</a> from 1951 hanging on my wall. <a href="v">Pentax AF adjustment</a> chart here is useful, but I prefer the USAF chart. Our friend Yvon Bourque outlines the how to <a href="http://pentaxdslrs.blogspot.com/2008/06/part-1-autofocus-adjustment-for-pentax.html">here</a>.</p>

<p>ME</p>

[zane1664879013]

<p>Going back to ME's point, which was emphasized in Part I: RAW files need sharpening in post. This is true of all cameras, not just Pentax. What, if anything, have you done in the way of sharpening?</p>

[mountainvisions]

<p>The post above by ME is about all you need to know.</p>

<p>Just click the links and use the K-7s focus adjust menu to get everything squared up. I had to do this with my test model K-7, and while my personal K-7 is MUCH better out of the box, i have a feeling it could use a little adjustment.</p>

<p>Also, read the Bob Atkins article chiefly because it talks about the shortcomings of Canon AF, read the comments below it.</p>

<p>I think if nothing else this should at least give you some pause about going with a different system for better results.</p>

<p>There is a comment by Dennis Yee (clearly english isn't his first language) where he notes Minolta and Pentax cameras were not found to have worse focus accuracy in chinese test (and actually in Pop Photos accuracy test Pentax usually is tops). I'd love to know more though about his statement to the effect of "mid range nikons and canons have better AF than top end." if this is true I assume he means accuracy vs speed, and I'd love to see that data!Although the comments and article are from 2003 so take it all with a grain of salt.</p>

[sjgcit]

<p>I think your problem is very simple. Depth of field, focus selection and a high-ish ISO.</p>

<p>Using the DNG you uploaded I see that it is shot at 135mm at f4. At 8 feet that gives you a depth of field of about 2-inches. And that's one inch in front of and one inch behind focus.</p>

<p>If you look at the photo of the two ladies, the one in red ( the bridesmaid I presume :-) ) is in focus. You can clearly make out goose-pimples and her necklace is in focus.</p>

<p>The bride is standing at right angles to her, and hence her face is some distance in front of the bridesmaid. Your very small DOF means that she is out of focus.</p>

<p>In addition you shot at ISO 400. You will get noise at ISO 400. You can believe anything you like about this, but you will still get noise ( on any camera ). Noise means loss of detail, which mean loss of sharpness.</p>

<p>All of your examples posted are at ISO 400 or above. All have a small DOF. And they are focused on something you did not want.</p>

<p>You indicate you use centre focus and 'wait for the beep' to shoot. That's not a method I'd recommend. What you are doing is waiting for the camera to tell you that it thinks it has focus. That does NOT mean what YOU want is in focus. Those focus areas are never perfectly aligned ( on any camera ). The time between hearing the beep and reacting is enough for two subjects to e.g. sway a little out of focus.</p>

<p>I'd use a purely manual focus ( with a split focus screen if possible ) and rely on your eye-to-hand reaction. I'd also suggest shooting several shots - expect some out of focus and act accordingly. In case it's an issue I will mention that I wear reading glasses and asked my optician for advice on this and he recommended wearing the glasses ( and not using a diopter adjustment ) when shooting. That's worked well for me. People often forget that their own eyes can be a problem. Note that taking glasses on-and-ff when shooting manual focus is a potential issue, as your eyes need time to relax to the changes.</p>

<p>I saw nothing in your posted examples ( some at 2:1 for goodness sakes ! ), which was not exactly what I'd expect from your settings.</p>