Camera Shake or out-of-focus?

<p>^ EXIF reveals the shot above was pulled at: F/10 @ 1/200s @ ISO800 using a 33mm Lens on a K-5.</p>

<p> </p>

<p>I checked the EXIF data (see below) of the non-blur shot of the woman and noticed the typical signs of Pentax's over cranked incamera contrast rendering seen in the clipped highlight on her wrist and plugged up black shirt and hair shadows setting contrast to Normal.</p>

<p>I can't express it enough that if you're going to shoot jpeg instead of Raw, you'll at least get more dynamic range and exposure latitude if you reduce incamera contrast and saturation. I have contrast and saturation reduced to max on my K100D which has a far more over cranked look when set to Normal than your K-5. I thought Pentax fixed this after my K100D and did noticed they did on the K20D.</p>

<p>Do an exposure test on pale skin or similar pastel light tones by checking the histogram after adjusting exposure to make sure the spike representing the pale tone falls a bit left from the right side. For example if your histogram is divided into 5 sections you want the spike to be within the first section as far from right side of the histogram as possible.</p>

<p>If you have an HDR setting experiment with that as well. With all these extra rendering settings you'll basically be calibrating your camera for human perceptual rendering style or what I call WYSIWYG.</p><div></div>

<p>I post jpegs because I thought that's all we're allowed to post. I shoot RAW+. Interesting to know about the contrast setting.</p>

<p>Interesting thread. Of course, an APS-c image does represent more of a blowup under the same circumstances. It is essentially like using a longer tele lens, and the rule for tele vs shutter speed must be adjusted, even though it can be lower than standard when SR is employed. As to throwing out pixels, the fact that a 24mp FF model has less pixel density than does a 24mp APS-c model might also be considered.</p>

<p>So after addressing the camera/lens aspects, as others have stated, the main issue is that here a live subject creates a separate set of parameters compared to still subjects. Live subjects invariably require higher shutter speeds to assure sharp results. At least 1/125 sec. and better at 1/250 for a subject holding "still". Portraits are often shot at larger apertures to defocus and blur background to make the subject stand out from the background. The last example shot would have been better at a larger aperture instead of f/10, especially since the background contains nothing of interest to include in the scene, so would be better defocused with the subject standing out from it. The larger aperture would also allow a faster shutter speed at the same ISO.</p>

<p>Using a large aperture can also be overdone. In the case of a lens having f/1.4, using maximum aperture presents its own problems: The lens will not be all that sharp at such a wide aperture. At a portrait distance, focus would become extremely critical, especially with a live subject, since DOF is extremely shallow. The eye may be in focus, while the ear is not. Also, with the slightest movement, the subject could move out of focus between the time focus is locked and the shutter is tripped. f/2.8 would be better depending on the focal length used and the subject distance. DOF would be shallower at 200mm, but less so at 70mm. 70mm would be better for portrait because subject's features would be more natural, less flattened than at 200mm.</p>

<p>I am not sure about the multiplying of APS-c images to a smaller aperture to get the same DOF as a FF image. I may be misinformed, but my impression has been that APS-c models tend to have MORE DOF under the same conditions, while a FF model has more capability to reduce DOF. That is to say, the APS-c body with a lens of 135mm @ f/2.8 would produce about the same image as a FF body with a lens of 200mm@ f/2.8 but the FF model would have greater DOF. Or, when getting a portrait shot, the FF model could obtain more background blur with a lens using the same aperture and image framing.</p>

<blockquote>

<p>I post jpegs because I thought that's all we're allowed to post. I shoot RAW+. Interesting to know about the contrast setting.</p>

</blockquote>

<p>Since you shoot Raw the settings I indicated will still aid in using that histogram to maximize exposure for clean and full capture of data which is how I use it.</p>

<p>What you should also attempt to do now is see how dark you can make your Raw captures when shooting portraits of people moving by prioritizing shutter speed with ISO setting so that you can use wide open aperture settings if desired.</p>

<p>That demo image I posted of the paint brush shot at ISO 1600 should give you an idea of what I'm talking about. Test first to find the sweet spot for clean and detail rich captures and forget ETTR. Noise reduction software has improved immensely compared to incamera jpeg use of it.</p>

<p>Think of your camera as a data collector and throw out all that focal length/shutter speed complexity and all the rest of the preemptive optics and light science rules and judge with your eyes in post shooting Raw.</p>

<p><strong>ERROR:</strong><br>

My error above: I should have written <strong>OPEN UP</strong> the lens ... <br>

This paragraph should read: (italics and bold has the correction) <br>

***<br>

The reference to APS-C Cameras <em>"adding 1.5 stops of aperture"</em> is to do with the Subject of: EQUIVALENCE in PHOTOGRAPHY. An easy way to explain the meaning of the particular reference of <em>"adding 1.5 stops"</em> is by a practical example - if you FRAME a Subject exactly the same way (i.e. "FRAMING" means the Subject occupies exactly the same amount of space in the image), using the same lens on a 135 Format ("Full Frame") and then do the same using an APS-C Format Camera - <em><strong>then you will need to OPEN UP the lens 1.333 Stops</strong></em> when using the APS-C Camera <strong>to have the same Depth of Field</strong> in the image as what you'd have if you used the 135 Format Camera.<br>

***<br>

In other words - to get the same DoF using an APS-C Camera as you would for a 'full frame' camera - if the Subject appears the same size in the frame (i.e. FRAMING is the same) and you use the same Focal Length lens on both cameras:(for example) using F/2.4 on the APS-C Camera will give the just about same DoF as using F/4 on the 'full frame' camera.</p>

<p>Sorry for my error, thanks to Michael Kuhne's comments for alerting me to that error.<br>

</p>

<p>WW</p>

 

<blockquote>

<p>"I may be misinformed, but my impression has been that APS-c models tend to have MORE DOF under the same conditions, while a FF model has more capability to reduce DOF."</p>

</blockquote>

<p>You are informed. And you are correct. I had it correct in my head and typed it the wrong way around. Thanks for picking that up.</p>

<p>BTW - I think you got this bit backwards: <br /> "That is to say, the APS-c body with a lens of 135mm @ f/2.8 would produce about the same image as a FF body with a lens of 200mm@ f/2.8 <em>but the FF model would have <strong>greater</strong> DOF.</em>"<br /> I think you meant to write the FF model would have LESS DoF?</p>

<p><a href="/photo/13136700&size=lg">Here is an example using an 85mm Lens on APS-C Camera and a 135 Lens on a 'full frame' Camera. </a></p>

<p>WW</p>

<p>Now that everyone is thoroughly confused...</p>

<p>Michael said:</p>

 

<blockquote>

<p>Interesting thread. Of course, an APS-c image does represent more of a blowup under the same circumstances. I<strong>t is essentially like using a longer tele lens,</strong> and the rule for tele vs shutter speed must be adjusted, even though it can be lower than standard when SR is employed. As to throwing out pixels, the fact that a 24mp FF model has less pixel density than does a 24mp APS-c model might also be considered.</p>

</blockquote>

<p> <br>

You're not understanding, Michael. Using a an APS-C is not like using a longer telephoto lens, it's like taking a full-frame image and cropping it in post processing. It's not the same thing at all. A 500mm lens on a APS-C will still be 500mm. The sensor is smaller, so it doesn't use the full image circle that the lens produces.<br>

<br>

Now, my full-frame body and my APS-C sensor body have the same pixel density. Higher pixel density does allow you to put more pixels on the subject at a given focal length, but, once again, pixel-density does not change the focal length. Camera's with higher pixel-density have higher Resolution and can show more detail on the subject at a given focal length. However, pixel-density has nothing to do with full-frame vs. crop-sensor</p>

<p>Something else - I am pretty sure I know what you mean here - </p>

<blockquote>

<p>"f/2.8 would be better depending on the focal length used and the subject distance. <strong>DOF would be shallower at 200mm, but less so at 70mm.</strong> 70mm would be better for portrait <strong>because subject's features would be more natural, less flattened than at 200mm</strong>."</p>

 

</blockquote>

<p>But for clarity, here are two points to consider referencing the bold text in the portion I quoted -</p>

<p>1. On DoF:</p>

<p>It doesn't matter what Focal Length is used - provided the FRAMING of the Subject stays the same and the APERTURE stays the same and the Camera Format remains the same - THEN > the DoF will remain the same: (this is the Axiom of DoF and is true for 'Typical' portrait shooting distances - the Axiom begins to fail slightly as we begin to approach the Hyperfocal Distance).</p>

<p>This axiom allows (especially Portrait Photographers) to choose to learn a rote memory for DoF.</p>

<p>These are examples of my DoF Cheat Sheets for my 135 Format gear (aka "full Frame") - <a href="/photo/17875347">LANDSCAPE ORIENTATION</a> and <a href="/photo/11721919">PORTRAIT ORIENTATION. </a></p>

<p>I have other cheat sheets for 6x6 and 4x5 and APS-C Formats, but I know most of them by rote, it's not that difficult to remember because it is all about the FRAMING. </p>

<p>* </p>

<p>2. On PERSPECTIVE (the flattening of the facial features)</p>

<p>It is <strong>not</strong> the choice of the Focal Length of the Lens which produces the "flatness" or the "elongation" of the facial features - but it is the SUBJECT to CAMERA DISTANCE which determines that.</p>

<p>Certainly, if we keep the FRAMING the same for two portraits and we use a 70mm lens for one and a 200mm lens for then we will have to step further away to make the 200mm shot - it is the fact that we are further away that makes the Face appear more "flat".</p>

<p>Another way of looking at this matter is to make a Portrait using a 200mm Lens and then DO NOT MOVE and make another Portrait using a 70mm lens and then in post production crop the 70mm image to the same framing as the 200mm . . and there will be the same "flatness" of the facial features in both images - and that is because the PERSPECTIVE of the two shots is the same.</p>

<p>In this technical meaning, the <strong>Perspective</strong> of an image is determined by the Distance from the Camera to the Subject and the Elevation of the Camera, relative to the Subject. </p>

<p>The choice of Focal Length of the Lens used will determine the Field of View or the Framing - but it does not determine the Perspective of the Image - the 'flatness', 'elongation' and 'foreshortening' are results of where the camera is located relative to the Subject and not because of the choice of Focal Length of the lens used. </p>

<p>To be clear, I am not arguing whether a 70mm lens is better or not better than a 200mm lens but rather pointing out exactly what factors will effect which outcomes. </p>

<p>WW</p>

<p>William, you are right- what I meant to say is the FF image would have LESS DOF. Glad you spotted that. Yes, you are absolutely right. Subject distance relating to focal length and perspective are very important considerations. The distance of the background behind the subject is also important in determining what aperture is needed to blur the background.</p>

<p>David, apparently you are not understanding what I said. Of course, technically, a 500mm lens is a 500mm lens regardless of its being on a FF or APS-c body. But the practical result IS different.</p>

<p>Of course, technically, the APS-c image represents a cropped cut-out of the concurrent FF image. But that is an important difference.</p>

<p>I mentioned pixel density because someone brought up throwing out pixels. But here goes. Pixel density, of course, does not "change" the focal length of a lens. But differences in pixel density have a lot to do with cropping and the resulting image. I see that you know that higher pixel density means higher resolution, which can show more detail on a subject. Good thing to know. But at the same time, this has a LOT to do with FF vs. crop sensors.</p>

<p>Carrying around a 135mm f/2.8 lens on an APS-c body, in real-world use, for all practical purposes is akin to carrying around pre-cropped images similar to those taken with a 200mm f/2.8 lens on a FF body. In using the APS-c setup, one is saving weight, size, and expense compared to the FF setup for getting similar results.</p>

<p>24mp APS-c sensors and 24mp FF sensors do not have the same pixel density, because the 24mp's of the FF sensor are spread over a larger area. The 24mp's of the APS-c image are crammed into a smaller area, having greater density compared to the same amount of pixels in a FF image. Even when evaluating lens and camera testing, where resolution is given in lines of resolution, it is important to take this difference into consideration, because results are given in lines per picture height. The numbers like 3,000 lines look more impressive in a test using a FF model compared to less lines for an APS-c model, but the frame height with FF is 24mm, so the lines given are spread over a greater area compared to the 16mm frame height of APS-c models.<em><br /></em><br /> <br /> One has to have a FF body with a sensor of more than 24mp (I am not sure how much more) to equal the pixel density of the APS-c body having a 24mp sensor. But even then the weight and expense is still an issue. If one has a 24mp FF body, shooting with a 135mm f2.8 lens, then cropping to obtain a framed image comparable to that of the APS-c body using the same lens, pixels must be discarded in the process. No longer will it be a 24mp image, while that taken with the APS-c setup will remain a 24mp image. In only one respect could the FF setup prove to offer an advantage when shooting telephoto, as we are speaking of- that is noise occurring when facing low-light or higher ISO use, where FF sensors exhibit superior performance. But when speaking of wide-angle to normal focal lengths, the advantages are pretty much all for the FF format.</p>

<p> </p>

<p>Michael said:</p>

 

<blockquote>

<p>David, apparently you are not understanding what I said. Of course, technically, a 500mm lens is a 500mm lens regardless of its being on a FF or APS-c body. But the practical result IS different.</p>

</blockquote>

<p> <br>

I am understanding what you say and you are wrong. There is no difference between cropping a full-frame image to the size of a crop-sensor size than taking the same picture with a crop-sensor and the same lens. It DOES NOT have the same affect as adding FL with a longer lens, a tele-converter or some other physical device. An image taken with a full-frame lens on a crop sensor body uses only a portion of the image circle. It's the same as taking the image on a full-frame body and cropping it.</p>

<p>David, did I not just say basically the same thing?? What did I write in the last paragraph above regarding using a 135mm lens on both FF and APS-c bodies? "If one has a 24mp FF (Full Frame) body, shooting with a 135mm f/2.8 lens, then cropping to obtain an image comparable to that of the APS-c body using the same lens...." </p>

<p>Then you write saying "There is no difference between cropping a full-frame image to the size of a crop sensor size than taking the same picture with a crop sensor and the same lens." I wrote the same thing, except for using other wording: "cropping to obtain a comparable image to that of the APS-c body using the same lens..."</p>

<p>The caveats I brought up were regarding the pixel loss due to the cropping, in addition to carrying larger, heavier, more expensive equipment.</p>

<p>Michael, the only reason to carry a crop-sensor vs. a full-frame sensor is file size and resulting fps. Yes, you can take the same image with the same lens on the full-frame and crop it to end up with the same image as the crop-sensor with the same lens, but, with today's processing technology, the full-frame will be at around half the fps and the buffer will fill up twice as fast. The focal length DOES NOT change, technically or practically, except the camera with the smaller sensor can be more responsive.</p>

<p>I shoot almost everyday with both full-frame and crop-sensor, with one on my shoulder and the other around my neck. For BIF, I grab the one that gives me 10-fps; otherwise, I'm indifferent. (The full-frame is slightly newer sensor technology so it's slightly better at high-ISO, so I will use it when ISO exceeds 1600). These choices are made from looking at hundreds of thousands of shots at 100% and pixel level, using one of the very best lenses that we can buy.</p>

<p>My Wish List includes a full-frame EOS 1D X II. Its pixel-density is less than half my 7D2 and 5DSR, BUT it'll produce clean images at crazy-high ISOs and shoot 12 and 14-fps. For a BIF shooter, those two factors are hugely important. Also, the AF is blazingly fast, even at f/8, so I can combine my 500/f4 with my 2.0x TC-III and have a true, usable focal length of 1000mm. Some might say, my 500mm, plus my 1.4x TC is already "equivalent" to 1120mm. Wrong. Look at pixel level, particularly in lower light and it's not the same, or close.</p>

<p>David, I'm glad you're happy with your Canon gear. It is obviously doing the job you selected it for, and I agree you've made a very appropriate choice. Yes, DSLRs for shooting fast-moving subjects are generally of less MP design for faster write times and higher FPS with larger buffer capacity. That serves your needs very well. And yes, a 20MP or less FF model can deliver clean images. Likewise, I can even get clean images from my little general-use APS-c 12MP K-r. Yes, I know of cropping and its effects, which I now occasionally do digitally, but used to sometimes get done with reprints back in the film days.</p>

<p>The only rationale for higher MP designs these days is to obtain yet higher resolution and greater cropping capability- but greater pixel density does come at a cost to those factors above, which are valuable to your requirements, including higher ISO performance.</p>

<p>My position on advantages of both FF and of APS-c designs I have covered, so I'm done with that. Besides, I think we've gone way off the topic of this thread, which is sharp focus and background blur.</p>

<p> </p>

<p>Michael said:</p>

 

<blockquote>

<p>The only rationale for higher MP designs these days is to obtain yet higher resolution and greater cropping capability- but greater pixel density does come at a cost to those factors above, which are valuable to your requirements, including higher ISO performance.</p>

</blockquote>

<p> <br>

I thought we were ready to end and then you said that. Generally the reason for higher mp is to print in larger sizes at higher resolution. If we never printed beyond 8x10", then 12MP would be plenty, in either crop or full-frame. For those that print 50" and above, then both sensor size and pixel density become more and more important. Printing at 300-dpi in large sizes requires lots of pixels. Being able to crop in focal length limited situations, as you state, is a good use for high pixel density, but far from "the only rationale." </p>

<p>"topic of this thread, which is sharp focus and background blur."<br>

Michael, actually it was subject blur or out of focus. I WAS going for background blur in some shots but the problem was subject blur. I know, it's been a long thread.</p>

<p>Yes, David, I thoughtlessly omitted print size, sorry.</p>

<p>Right, Howard, subject sharpness was the main issue. I got off the main issue in the long thread process, but of course it was. I hope comments have been helpful.</p>

<p>They were. I will try them out the next time I go on a shoot. Thanks.</p>