Which factors should dictate the choice of focal length?

<p>I've only recently learned that perspective is determined by the photographer's position, not the focal length.<br>

So, which factors should dictate the choice of focal length?<br>

Assuming you are not on a paid project in which the resolution is critical, my reaction is that go as wide as possible and then crop the picture in PP. It would give you more room for editing and also make wider aperture available, as most zooms allow wider aperture in the shorter end.<br>

I see two possible limitations: First, there is only so much you can crop in PP without sacrificing the image quality. So this is definitely one factor that should dictated the choice of focal length. (I think this restriction will be less onerous as sensors continue to improve.)<br>

The second factor would be the distance from your subject in a portrait shooting for example. You want to keep a good, interactive distance from your subject and still visualize the end-result in the viewfinder sufficiently. That may require the traditional portraiture focal length (short telephoto) rather than something wider, although, by the logic that I mentioned above, you can still shoot with a wider focal length and crop it to arrive at a comparable result in PP.<br>

Are there other factors that need to be considered when choosing the focal length?</p>

<p>Right, perspective is dictated by the camera's position. Angle of view is dictated by focal length. In other words, the only difference between one focal length and another is how large a frame it puts around the center of the image. (There are special-case lenses for which this is not quite true, such as fisheye lenses, but in general this is true enough for most purposes.)</p>

<p>In the abstract, as you say, you could simply shoot as wide as possible and then crop, but in practice this is limiting because you do not have infinite resolution available to you, and the more you crop, the more lens aberrations will become noticeable because the resulting tiny image has to be blown up more than an equivalent image shot at a longer focal length.</p>

<p>Here are a couple of pages you can look at to see the effect of shooting portraits at different focal lengths (and correspondingly different camera to subject distances):</p>

<p><a href="http://www.mcpactions.com/blog/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/">http://www.mcpactions.com/blog/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/</a><br>

<a href="http://stepheneastwood.com/tutorials/lensdistortion/index.htm">http://stepheneastwood.com/tutorials/lensdistortion/index.htm</a></p>

<p>Personally, I think the "shoot wide and crop later" attitude suggests a rather lazy desire to avoid thinking about what you're doing at the time of shooting. You should know what perspective and framing you want when you shoot, and you should choose a camera position and focal length to give you the desired result. Not only will this result in better image quality, but also, pre-visualizing your images leads to better photography because it requires you to focus your attention on your work and to be clear in your mind about what you are trying to accomplish.</p>

<blockquote>

<p>I see two possible limitations: First, there is only so much you can crop in PP without sacrificing the image quality. So this is definitely one factor that should dictated the choice of focal length. (I think this restriction will be less onerous as sensors continue to improve.)</p>

</blockquote>

<p>There is one other problem, you'll run out of lens resolution, same as you did and still do with film.</p>

<p> </p>

<blockquote>

<p>I've only recently learned that perspective is determined by the photographer's position, not the focal length. So, which factors should dictate the choice of focal length?</p>

</blockquote>

<p>Camera Viewpoint is chosen for the Perspective required in the Scene.<br /> Focal Length is then chosen to display the Main Subject (from Wide to Tight as desired), within that Scene.</p>

<blockquote>

<p>Assuming you are not on a paid project in which the resolution is critical, my reaction is that go as wide as possible and then crop the picture in PP. It would give you more room for editing and also make wider aperture available, as most zooms allow wider aperture in the shorter end.</p>

</blockquote>

<p>Not a good idea.<br /> The gain of Aperture is relatively minimal and in most circumstances not of concern nor use to you.<br /> The lost of quality by cropping is another issue.</p>

<p> </p>

<blockquote>

<p>Are there other factors that need to be considered when choosing the focal length?</p>

</blockquote>

<p>But my major concern is the lack of Vision (or effort applied to the Vision), going into the shoot.</p>

<p>Shooting a LITTLE wide to allow to crop to various PRINT RATIOS, should not be confused with Shooting Wide just because there is no Vision or Concept, before the shutter is released.</p>

 

<blockquote>

<p>With a wide angle, under my scenario, you wouldn't stand so close to the subject that distortion arises. You'd stand where you'd with a telephoto lens, shoot and crop later. There should be no distortion of the subject's features.</p>

</blockquote>

<p>Usually only true if the Subject is close to the centre of the Frame – at the edges it might be a different story.</p>

<p>WW</p>

<p> </p>

<blockquote>

<p>The gain of Aperture is relatively minimal and in most circumstances not of concern nor use to you.<br /> The lost of quality by cropping is another issue.</p>

</blockquote>

<p>I cannot agree to this. Depending on the lens, the gain of aperture IS substantial, sometimes as large as close to 3 stops. I've had to shoot wide and crop later just to make use of a wider aperture in low available light situations on a number of occasions.</p>

<p>Hi Ken,<br />Sure it depends upon the lens used, but if we are to dissect this issue:<br />What lenses with three stops?<br>

How often would we consider using those lenses at the extreme wide to crop to the extreme telephoto to make that 3 stop difference?<br>

If we are now discussing now hypothetical and theory – that’s OK I can do that ‘til Mrs Brown’s Cows come home.</p>

<p>But the original question reeked of the practical and the application of making quality photos - <br />Hence the response -<br />The gain of aperture “<strong><em>in most circumstances”</em></strong> . . . and etc</p>

<p>***</p>

<p>In any case, the collector of the titbit information which forms part of me, would like a few examples of Zoom Lenses which have close to three stops or more compass of Maximum Aperture across their Zoom Range, please.</p>

<p> Noted mentioned was: “<strong>close to three stops</strong>”<br>

A quick scan of my brain box and I come up with “super zooms” as one answer.</p>

<p>And (as one example) the Tamron 18<strong>-</strong>250mm<strong> </strong>F/3.5-6.3 AF Di-II LD Aspherical (IF).<br />(Which is 1⅔ stops?) – And that is NOT close to three stops.<br>

<br />I didn’t think too hard about it though – hence my curiosity about the lenses which meters close to three stops of aperture compass, across their zoom range.</p>

<p>***</p>

<p>But then, if we are limited to only these types of super zoom - and taking this TAMRON lens as an example and (reasonable to) assume that this lens is a F/6.3 at about FL = 150mm, and also that F/3.5 would be used up at about FL = 24mm –<br>

It would be a rare (practical) circumstance that we would encourage shooting at (for example) FL= 20mm and then to crop to the equivalent framing of FL = 150mm to gain Shutter Speed value of close to TWO stops – even if we were maxed out at ISO and using an old DSLR (for example) which was only useful to ISO1600.</p>

<p>It would be far more reasonable to suggest (if we are discussing the PRACTICAL):<br>

to move a tad closer and frame the original a bit tighter;<br>

and to crib perhaps ½Stop of Exposure and under expose;<br>

and to crop less tight and to leave the shot wider;<br>

and to carefully convert the image to Black and White, rather than use the Colour version.<br>

The COMBINATION of these re-works, would be a far better PRACTICAL solution to the problem – that is if we are having a discussion of the PRACTICAL.</p>

<p>***</p>

<p>However, if we are conversing about the Hypothetical and the world of Theory:<br>

Then let’s being with a few lenses which come close to three stops compass . . . and we can then theorize about the extent of the range of cropping which is required to make the framing.</p>

<p>WW</p>

<p> </p>

<blockquote>

<p>In any case, the collector of titbit information which forma part of me, would like a few examples of Zoom Lenses which have three stops or more compass of Maximum Aperture across their Zoom Range, please.</p>

</blockquote>

<p><br /> It's not really a "lens" in and of itself, but my Canon S95, a P&S that I use almost daily to take casual shots, has f/2.0-4.9 in its 28-105mm in 35mm equiv. Ok, it's 2.45 stops, so I exaggerated, but the point is that I've had to shoot wide to be able to choose a larger aperture.<br>

Same with the kit zoom lenses for my micro four thirds. The 14-42mm has f/3.5-5.6, and it's not unusual for me to shoot wide to gain as much exposure as possible.<br>

<br /> http://usa.canon.com/cusa/consumer/products/cameras/digital_cameras/powershot_s95#Specifications</p>

<p>I have a 12mp camera. I know from experience that if I do my job right when I take the picture (exposure, focus, no camera shake, etc) I can send the file to a lab and get a nice looking 16x20 print for my wall.<br>

But if I am framing a picture and I only care about the middle 1/3 (horizontally & vertically) of what my lens is delivering to the sensor, then my 12mp sensor is being reduced to a 1.3mp sensor (12/(3*3)). My chances of getting a nice 16x20 out of a 1.3mp image is not so great.<br>

But if I swap out my lens and replace it with one that is 3x the focal length, then I can have all 12mp dedicated to the picture I want on my wall.</p>

<p>If I had a lens with infinite resolving power and a sensor with infinite resolution, it would be fine for me to just crop after the fact. The camera might even allow me to trace a square on the lcd of the crop I have in mind when I take the picture.<br>

But since I have to take photos with the camera and lenses I own, I try to get the framing right "in camera". And sometimes it's not possible. Sometimes I have to use my longest lens, and then crop. Sometimes a low quality picture is better than no picture at all.</p>

 

<p>Yes it is true that the distance between the camera and subject determines perspective but it's perspective that dictates the choice of focal length. Simply if you stand far away to get a flat perspective the wide angle lens isn't get you the right framing. Same thing to get the pronounced perspective for being too close you would see nothing in the viewfinder with a long lens so you need a wide angle lens. So perspective dictates distance and distance dictates focal length so still it's perspective that dictates focal length.</p>

<p>A factor which immediately came to mind is how good are you with your legs? Second question ... In how many of your photographs does perspective matter that much?</p>

<blockquote>

<p>Yes it is true that the distance between the camera and subject determines perspective but it's perspective that dictates the choice of focal length.</p>

</blockquote>

<p>No, perspective plus desired framing dictates focal length. Perspective by itself isn't enough.</p>

<p>For me, how far away my subject is, dictates my focal length. This may differ in a studio environment, but out in the great out doors, I often need to get close to something not very close.</p>

<p> </p>

<p>If I may be so bold, I believe it was a post by me that has caused Ken to start this thread. Here is the image I used to prove position makes perspective, not focal length.</p><div></div>

<p>Obviously the 17mm image is severely cropped, but in the previous thread I didn't post the full 17mm image. So here it is. I would point out, for further thought, both images were shot at f8, though the 17mm image is back focused slightly there is a massive dof difference.</p><div></div>

 

 

<blockquote>

<p>Yes it is true that the distance between the camera and subject determines perspective but it's perspective that dictates the choice of focal length.<br>

No, perspective plus desired framing dictates focal length. Perspective by itself isn't enough.</p>

</blockquote>

 

Both of these are wrong. How far you are from the subject may have nothing to do with desired perspective. It may have to do with accessibility. It may have to do with minimum distances. It may have to deal with avoiding obstructions. Focal length is determined by the distance you are from the subject, regardless of desired perspective, and it's worth pointing out that perspective is just as influenced by relative angle of the camera to the subject, not just distance, and what framing is available. "Desired framing" is just as restrictive as "perspective." It's often simply about what is possible rather than desirable, and perspective for many photographers is just an implication of where they stand. Focal length takes care of getting the right stuff in the frame.

 

 

<p><em>"It would give you more room for editing and also make wider aperture available, as most zooms allow wider aperture in the shorter end."</em></p>

<p>From a selective depth of field point this doesn't help, for instance take an 18-55 f3.5-5.6, at a subject distance of three meters your dof at 18mm and f3.5 is 9.6 meters, with the zoom at 55mm and f5.6 the dof is under a meter. So you would get much deeper dof at the faster wide end, even thought the f stop is wider. It would help if you wanted everything in focus or if you needed a faster shutter speed.</p>

<p> </p>

<blockquote>

<p>Canon S95, a P&S that I use almost daily to take casual shots, has f/2.0-4.9 in its 28-105mm in 35mm equiv. . . The 14-42mm has f/3.5-5.6, and it's not unusual for me to shoot wide to gain as much exposure as possible.</p>

</blockquote>

<p>OK – thanks for the nuts and bolts answer. <br>

So it seems that we are discussing the Practical and not the Theoretical and Hypothetical –<br>

I’m experienced with similar cameras as I shoot with a couple of Canon Powershot Cameras, quite often; the lesser of the two (S5 IS) is not very capable at ISO1600 – and I shoot a lot of indoor stuff in available light, where the Subjects are in motion.<br>

I expect your P&S would be acceptable at ISO1600? <br>

So, OK, I can envisage SOME shooting scenarios where the Tv required would dictate prostituting other of the exposure parameters – but I would suggest, even for that list of shots where a faster Tv were required - that there would often be OTHER methods of getting around your problem – to either be used in CONJUNCTION with SOME shooting wide and a LITTLE cropping or INSTEAD OF shooting very wide (to get a very large aperture) and the cropping mercilessly in PP to get the framing you require.<br>

Four of those possibilities, I have listed above.</p>

<p>***</p>

<p>The mere 1⅓ Stop Difference of the 14 to 24 zoom would surely account for even fewer occurrences, where it were absolutely necessary to (for example) shoot at FL = 14mm and crop to the equivalent framing of FL = 30mm. </p>

<p>WW</p>

<p>I do not think that resolution (or shall we say the optical performance of the lens?) is important only in paid projects.</p>

<p>@ Scott,</p>

<blockquote>

<p>If I may be so bold, I believe it was a post by me that has caused Ken to start this thread. Here is the image I used to prove position makes perspective, not focal length.</p>

</blockquote>

<p>You are not being bold at all. You are absolutely right.<br>

The two pictures you've posted again lead me to wonder: so the two photos were shot at the same f/stop? I'd have thought that the two would result in the same depth of field if you crop a wider-angle shot and make it have the same frame of view as the tighter one. I'm surprised that the two were shot at the same f/stop and still exhibit significant difference in DOF even after cropping.</p>

<p>If a wide angle shot ends up with a larger DOF even after cropping, then DOF indeed would be a very important factor in choosing focal length.</p>

 

<p>"If a wide angle shot ends up with a larger DOF even after cropping..."</p>

<p>It is popular on the internet to say focal length doesn't affect DOF, but it isn't always made clear that they are assuming you are changing your position relative to the subject to maintain a similar composition. If you change focal length without changing position (focusing distance) then shorter focal length = more DOF, and longer = less, as the posted examples show.</p>

<p>shorter focal length, longer focusing distance, smaller aperture = more DOF<br /> longer focal length, shorter focusing distance, larger aperture = less DOF</p>

<p>Similar to exposure reciprocity it's possible to increase DOF with one while decreasing it with another, and end up with the same DOF. For instance short distance + short focal length can have the same DOF as long distance + long focal length.</p>

<p>I choose focal length based on my experience. Many factors may influence my decision, but thinking about it technically it's mainly a matter of efficient in camera cropping.</p>

<blockquote>

<p>Both of these are wrong. How far you are from the subject may have nothing to do with desired perspective. It may have to do with accessibility. It may have to do with minimum distances. It may have to deal with avoiding obstructions.</p>

</blockquote>

<p>No, I wasn't wrong. Your ability to follow logic is at fault. Yes, there may be other practical considerations that limit your choice of positions, but what that basically means is that some desired perspectives may not be achievable. This doesn't change anything, though. You choose your position (from those that are available to you) based on desired perspective, and you choose a focal length that gives you the framing you want. There are other assumptions embedded in this too; for example, if all you have with you is a small selection of prime lenses, then obviously you can't choose any arbitrary focal length you want; you may decide to alter your position somewhat to compromise between the ideal perspective and the ideal framing. This may be the practical reality at times, but it doesn't change the conceptual relationship between position and focal length in defining perspective and framing. Same for your objection that the photographer may not have total freedom to be just anywhere he wants to be, or may be limited by his lens's minimum focus distance (or still other considerations -- the range of his flash, for example).</p>

<blockquote>

<p>You choose your position (from those that are available to you) based on desired perspective</p>

</blockquote>

<p><br /><br>

I generally choose my position based on desired proximity to the subject. I then figure out my framing and see if the perspective works.</p>

<p>Ken,</p>

<p>Yes dof changes significantly with focal length, as the example I repost below shows.<br>

<em>" for instance take an 18-55 f3.5-5.6, at a subject distance of three meters your dof at 18mm and f3.5 is 9.6 meters, with the zoom at 55mm and f5.6 the dof is under a meter. So you would get much deeper dof at the faster wide end, even thought the f stop is wider."</em><br>

<em><br /></em><br>

My first two images were both shot at f8, 1/60 sec, 100 iso. The subject distance was 4.6 meters. Given that metadata we can calculate the following. dof at 200mm = 0.26m: dof at 17mm = 1m-∞, a huge difference.</p>

<p>There are many many working pros who don't get this stuff either. For instance, many, many people covert the "unmistakeable look" you get with the 85 f1.2 wide open. Well at 3 meters subject distance you do get a very narrow dof and a slightly compressed, flattened, natural look to the face. But step back one meter increasing your working distance and changing your perspective slightly, use a 135 f2 at f2 (a $1,000 lens as opposed to a $2,000 lens) and you have the same dof and a very very similar perspective. Put the two images next to each other and few could tell them apart. If you want narrow dof on the cheap go long.</p>

<p>All these compromises and calculations just happen in the minds of most photographers with little thought to "how". For instance, most people end up using one lens they own for one type of shot because that is what they have done before and liked the result. They created the perspective, focal length and dof previously, by accident or design, that then becomes part of their repertoire. You can start with any of the factors, but the sum of the image can only achieved by the combination of all three.</p>

 

<blockquote>

<p>" . . . many people covert the "unmistakeable look" you get with the 85 f1.2 wide open. Well at 3 meters subject distance you do get a very narrow dof and a slightly compressed, flattened, natural look to the face. But step back one meter increasing your working distance and changing your perspective slightly, use a 135 f2 at f2 (a $1,000 lens as opposed to a $2,000 lens) and you have the same dof and a very very similar perspective."</p>

</blockquote>

<p><a href="../photo/11961575&size=md">Yes</a> - even for the Tight Head Shot.</p>

<p>So in this aspect of the analysis / comparison, also - if we think of that as a zoom lens (85 to 135 F/1.2~2.0), <em>in most circumstances</em> the 1 stop difference is not that much difference or not that much use, either . . . ? </p>

<p>WW<br /> Hi Scott!</p>

<blockquote>

<p>"...But step back one meter increasing your working distance and changing your perspective slightly, use a 135 f2 at f2 (a $1,000 lens as opposed to a $2,000 lens) and you have the same dof and a very very similar perspective."</p>

</blockquote>

<p>Makes me wonder how I can best use my bag of well UNDER $1000 lenses. The big problem is I only have two , f2.0 or wider lenses. A 50mm and a 35mm. Neither are going to be as effective as the 85mm or 135mm. I could probably work out something with my 180mm f2.8, but it's not going to be in a normal sized room anymore. </p>