steve_dunn2

<p>As a test, try using the Digital Lens Optimizer feature in DPP. If that fixes the CA, then your lens pretty much matches what Canon thinks it should be, which suggests it isn't a bad copy of the lens and this level of CA is normal for this lens. If a substantial portion of the CA remains, however, then you probably do have a bad copy.</p>

<p>I may be wrong here, but I believe I've read that the industry standard is that the actual focal length of a lens should be within ±5% of the advertised focal length.</p> <p>Also, keep in mind that focal length is measured with the lens focused at infinity. The effective focal length can change (significantly in some cases) as you focus on subjects at closer distances.</p>

<p>Reading the Photonotes article cited above would be very wise indeed.</p> <p>Quick summary of your situation: in P mode, in "subdued" lighting, the camera assumes that the flash is intended to be the main light source. If you set FEC to -1, then, you've told the camera to underexpose your main light source by one stop. You had success with this in bright light because in that situation the camera assumes that ambient is the main light source and the flash is fill.</p> <p>In Av (assuming you haven't adjusted the custom function for this), Tv, or M, the camera always assumes that ambient is the main light source and the flash is for fill. So take the camera out of P and you'll get results more like what you're looking for.</p>

<cite>However, I've almost never seen a so-called TIFF file that doesn't contain extensions, which is still TIFF, because the rules allow for extensions. Whether those extensions to the format will be readable is anyone's guess.</cite> <p>This is a good point. For instance, there are several different compression methods you can use in TIFFs (mostly lossless, but lossy JPEG compression is actually one of the options within a TIFF), and while it's likely that just about any program that supports TIFF can read an uncompressed TIFF, it's probably not as good a bet that all your TIFF readers can read every different type of compressed TIFF. The same probably applies to other optional extensions.</p> <p>So make your TIFFs as vanilla as possible if you want to have the best chance of being able to read them on whatever software you may happen to use years or decades from now.</p>

<p>First, a simplified explanation of how the sensor works. Digital camera sensors are actually mostly analog. Light striking a pixel causes a change in the electrical charge in that pixel; that charge is then fed through an amplifier and into an analog-to-digital converter. ISO is implemented by varying the gain in the amplifier. The same scene exposed at (say) 200 will result in only half as many photons hitting the sensor as if it were exposed at 100, so the amplifier has to provide twice the gain in order to put the same signal into the A/D converter and get the same digital value.</p> <p>I haven't seen a list lately of which models do this (and so I can't speak specifically to the 5D3), but at least as of a few years ago, most EOS DSLRs were designed to control the amplifier gain in one-stop increments. Intermediate values were done as Mr. Avis suggests: the scene is metered at the ISO you set but the amplifier gain is set to the nearest full-stop value, resulting in an over- or under-exposure by a fraction of a stop. The resulting digital values are then adjusted to compensate. If you're shooting at an ISO that results in a slight underexposure, you'd expect slightly more noise (but also slightly more highlight detail since you'd have 1/3 of a stop more room before it clips); if you're shooting at an ISO that results in a slight overexposure, you'd get the opposite.</p> <p>A pro may have a specific reason why (say) they must shoot at 160 rather than 200. I'm not a pro and I do not have any such reason. Heck, back in the film days, almost all the films I used were in full-stop increments anyway (mostly 100 and 400). So personally, I have no need for 1/3-stop ISO increments. But that's just me.</p>

<cite>I use the lens correction feature in Lightroom. Is that what you mean by DLO?</cite> <p>I'm not familiar with Lightroom so I don't know if that feature is comparable, but my guess is they're two different things. DLO doesn't merely correct for things like vignetting and chromatic aberration; it can in many cases undo softness inherent in the lens. There's a bit more info on it, including a quote from Canon with a sort-of explanation of what it does, on <a href="http://www.p4pictures.com/2012/05/improving-image-quality-with-digital-lens-optimiser-in-dpp/" target="_blank">this page</a>.</p> <p>If you haven't tried it, it might be worth your time to give it a shot on a few of your images. Process the image in Lightroom as you usually do, then process the same RAW file in DPP using DLO. It's probably best to avoid doing much in the way of tonal manipulation because it will be difficult for someone who's familiar with the controls of one program to achieve comparable results in a different, unfamiliar program. But look at the parts of the image where you find the lens didn't perform well and see how DPP's version with DLO compares to Lightroom's.</p>

<p>I haven't used the 18-135 you mention so I can't directly compare the two. But I believe you'll find that in general the 17-55 is better. It's a very good lens, with optics much closer to professional quality than to the quality you'd expect from an all-in-one zoom like the 18-135. In their overlapping range, there's not a lot of difference in the pictures I get with this lens versus my 24-105/4L IS USM. That's the general answer; now, for your specific questions.</p> <p>You mention a couple of specific cases, and I'm not sure there would be a huge difference in those, to be honest. For landscapes at f/8-f/11, your 18-135 is already stopped down a couple of stops, so it's probably around its sweet spot (most lenses are at their best when they're stopped down a couple of stops). The 17-55 is good at those apertures, too, and I'd expect it to be better - but not necessarily dramatically better. I don't use mine for close-up work, so while I can say you will get more background blur at f/2.8 than at the apertures your 18-135 offers, I don't know how sharp the lens is when used with particularly close subjects. I probably wouldn't shoot night skies wide open on just about any lens unless there was a compelling reason to do so (e.g. if things in the scene are moving and your choice of shutter speed for the moving objects requires a wide-open lens).</p>

<p>I'm not 100% sure, but I think the C500 is the only camera in the EOS family that can do that. Most if not all of the normal EOS SLRs (the ones that are basically still cameras with video capabilities, as opposed to the cinema cameras like the C500) cannot do this, nor would it normally be desirable because of the massive amounts of data generated. For instance, 4k is about 8 megapixels, times three colour channels, and at 60 frames per second, that's over a gigabyte of data per second.</p>

<p>The body and the lens both play a role in how quickly autofocus works.</p> <p>The body's AF system has to determine what needs to be done to bring the subject into focus. As a general guideline, newer and higher-end bodies tend to do this more quickly than older and lower-end bodies do. Other factors such as lighting come into play, too, and may not affect each body equally (one body, for instance, might lose more of its AF speed in poor light than another body does).</p> <p>Once the body gives the lens its command, then the speed is up to the lens, and again various factors come into play. The type of AF motor and the quantity of glass that needs to be moved obviously both play a role. The CPU in the lens (each lens has one) and the firmware it's running also can have an impact. And the lenses that are compatible with teleconverters intentionally slow down AF drive when a teleconverter is used, trading some performance for improved accuracy.</p>

<p>I don't think I've seen a solid reference for how much light the pop-up flash emits when it's being used to control external flash units, but I have seen some comments that in some circumstances it can have a noticeable effect on the scene. Obviously if you're shooting at ISO 100 in bright day light, it's not going to cast shadows; at most you might see a catchlight in your subject's eyes. But if you're (say) shooting in a darkened room at high ISO, perhaps it would be enough to be noticeable in the final picture.</p> <p>I know you say it's adding "a lot" of light, and it would be hard to imagine a situation in which merely using the pop-up flash to control a slave would get that high, but I thought I'd throw it out there just in case. A sample photo or two might help.</p>

<p>In general, here are things that lithium-based rechargeable batteries don't like:</p> <ul> <li>Heat <li>Being charged all the way (or nearly all the way) <li>Being allowed to run down all the way (including being put into storage with a charge that's low enough that they'll self-discharge while stored) </ul> <p>Ever wondered why your two-year-old cell phone's battery can barely make it through a day any more, yet a several-year-old hybrid car still has a usable battery? It's because you plug your cell phone in every night, charging it to 100%, then holding it at 100% for several hours. The car's energy management system, on the other hand, does its darndest to avoid letting the battery's charge get too high or too low. That minimizes stress on the battery.</p> <p>For more info, have a look at these links.</p> <ul> <li><a href="http://batteryuniversity.com/learn/article/how_to_charge_when_to_charge_table" target="_blank">How and when to charge various types of batteries</a> <li><a href="http://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries" target="_blank">A somewhat more technical discussion of how a lithium battery's lifetime is affected by how it's used</a> </ul>

<p>The speed at which the camera can access the card really only matters if you fill the buffer or when you're chimping. If you're shooting, but not quickly enough to fill the buffer, the camera is always ready for the next shot, and the only difference between a faster card and a slower one is how long it takes for the little "card access" light to turn off.</p> <p>For what it's worth, you have to scroll a long way down Rob Galbraith's <a href="http://robgalbraith.com/camera_wb_multi_page7de5.html?cid=6007-12452" target="_blank">performance test of various cards with the 5D III</a> to find the first SD card; the top few dozen cards are all CF, with even the fastest SD being only about a quarter as fast as the fastest CF and none of the fastest SD cards coming anywhere near the speed their manufacturers claim. So there is probably something to what you've read about this camera being unable to use SD cards at full speed.</p>

<p>I can't help with much of this but I did want to make sure you're aware that if you use the 1.4x along with the 7D II and either of the two Canon lenses you mentioned, you will only have autofocus with the centre autofocus point. All the other points require a lens that's f/5.6 or faster, and while both lenses are f/5.6 or faster on their own, they're not when you include the 1.4x.</p>

<p>That little microswitch has generated its fair share of issues in this forum over the years, that's for sure. Some other possibilities to consider, not necessarily in any particular order:</p> <ul> <li>Does the flash fit as securely in the 7D's hot shoe as in the 7D2's? Since the history of this 7D is unknown, it's possible that it's had a rough life and perhaps the shoe has been bent a bit, and the flash is not being held as securely as it should, and therefore not making as good contact as it should. <li>Does orientation affect it? If it works in landscape mode (where the weight of the flash unit tends to push it down onto the contacts) but not in portrait mode (where the weight of the flash would tend to pull one side of the shoe away from the contacts), that would also suggest an issue with physical fit. <li>I suspect you're right that it's the 7D, not the flash, that's the problem. Do you have access to another flash unit (perhaps a friend's) that you could use with both bodies to try to duplicate the problem? If a different flash has the same behaviour, then that would confirm that it's the body. <li>Have you cleaned the contacts on the 7D's hot shoe? Maybe they're a bit grimy. </ul>

<p>The amount of light coming in does indeed affect the AF system, but that's not where the f/5.6 (for most bodies) limit comes from. If it was, then an autofocus system that would work at f/5.6 at dusk would work with even ridiculously slow lenses as long as there was bright sunlight - but if you try it, you'll find it doesn't work. You might get an extra stop or two, but after that, you can add all the light you want and the AF system simply will not work.</p> <p>The reason is that the light being sent to the two halves of the AF sensor comes from the outer parts of the cone of light that the lens sends into the body. Have a look at <a href="http://www.clarkvision.com/articles/understanding.autofocus/" target="_blank">this page</a> for an explanation, complete with diagrams, some of which illustrate what happens if the aperture is smaller than what the engineers set as a requirement. You can make the AF system more accurate by sampling light from farther apart in the cone (which is what the high-precision f/2.8 sensors do), but that requires a wider cone of light, which is only possible with a faster lens. Likewise, you could make an AF system that will work with as slow a lens as you like (as long as the scene is well illuminated) but accuracy will suffer.</p> <p>Some manufacturers, such as Canon, choose to turn off the AF system if the effective aperture of the lens (including any TC) is slower than a certain limit; others choose to allow the AF system to try to do its thing, even if that may be slow or inaccurate. The advantage of Canon's choice is that it (as you put it) prevents newbs from returning their supposedly defective cameras. The advantage of doing it the other way is that you don't arbitrarily lose AF in situations in which it might actually work.</p>

<p>The pins actually tell the <em>lens</em> that there's a TC attached, and which magnification it is (1.4x or 2x; the life-size converter for the 50mm macro also uses these pins). The lens then communicates this information to the body.</p> <p>Obviously, the end results include the body knowing the effective focal length and aperture range of the lens+TC, and as mentioned above, AF can slow down for improved accuracy. Another effect, which can be particularly useful in these days of being able to correct for things like distortion and vignetting, is that the body can record the actual lens used in the EXIF data, so (for instance) while a 200/2 + 2x is effectively the same speed and focal length as a 400/4, the body can tell the two apart, and so can your editing software.</p> <p>And just in case anyone isn't clear on why the AF system doesn't work with lenses slower than f/5.6 or f/8 (depending on the body), it's a physical limitation of the way phase-detection AF works. You can engineer a phase-detection AF system to work at slower apertures, but there are tradeoffs in speed and accuracy (which is why bodies that have "high-precision" AF points need faster lenses to activate those points than they do for the rest of the AF points). Canon has chosen f/5.6 for most AF points in most bodies. That doesn't mean that (say) an f/6.3 lens cannot work; there are plenty of third-party zooms that are f/6.3 at their long ends, yet they work, so Canon has engineered a safety margin into the system. And that's what allows some body+lens+TC combinations to AF if you tape the pins: you've gone beyond the official limit but are still (barely) within the limits of what the hardware can do.</p> <p>(So you're probably asking "Why doesn't the body disable AF with a slow third-party lens?" The answer: third-party lenses lie about this and tell the body they're actually f/5.6. It doesn't throw off metering; the body basically thinks the scene is about half a stop darker than it really is, and when the body tells the lens to stop down, the lens compensates for the difference.)</p>

<p>The EOS-1N is indeed a Type B body; it supports A-TTL and TTL but not E-TTL. Any genuine Canon flash with an E in its model name should work in at least TTL mode, with EZ units providing A-TTL in applicable situations.</p> <p>Having no experience with Yongnuo equipment, I can't tell you why it isn't working. Maybe since it's been (rough guess here) close to two decades since the last time Canon released a new Type B body, Yongnuo didn't put much effort into testing such a configuration?</p>

<ol> <li>Both flash metering systems are built into the camera, and in the case of E-TTL (which is what you have with that flash and either body), they both use the evaluative metering sensor. Basically, the camera meters ambient exposure, then sees what changes when the pre-flash is fired. The intent is that ambient exposure is set using the areas that don't change (much) from the pre-flash (e.g. the background that's too far away to get much light from the flash) while the flash exposure is set to bring the subject up to a proper exposure. <li>Since the camera's evaluative metering sensor is used for flash metering, it can't effectively meter anything that doesn't fill at least one of the evaluative metering zones. Those two bodies have different metering sensors; you can look in their manuals for more information on the sizes and positions of the metering zones. <li>This is basically the same as question 2. Av uses the metering system to set exposure, so if you're shooting something that doesn't lend itself to metering, you'll get a wonky result (or you'll have to set exposure compensation manually). M uses the meter to give you a hint, but the actual aperture and shutter speed are set by you, so if you can figure out the right settings, you can shoot anything in this mode. <li>I believe it should, since the metering sensor is still reading the light that comes in through the lens. But I've never used them so I can't say from personal experience. <li>You'll need to keep the shutter speed no higher than the body's maximum sync speed (unless you're using high-speed sync mode, which is supported with your equipment). Other than that, you can pick any shutter speed and aperture you like. </ol>

<p>One of the Mark II extenders has the same optical formula as the equivalent Mark I, but with weather sealing (and I think additional blackening inside to help tame stray light); the other one has a different optical formula. I think it was the 2x that got the new formula but I'm not 100% sure; that was quite a while ago.</p> <p>Both Mark III extenders are different from the previous models.</p>

<p>Canon doesn't keep parts forever for discontinued lenses; there are lots of older models that Canon can no longer repair because they don't have or make parts for them any more. If you're speaking of a genuine Canon repair centre in Singapore, then the fact that they can't get the part probably means Canon no longer makes this part, and if any such part exists, it would be sitting on a shelf somewhere at a repair centre that hasn't yet run out of stock.</p> <p>If that's the case, one other option is to search for another copy of this lens being sold "as-is" due to some other part failing (e.g. a dead focus motor), and use that lens as a parts donor.</p>

<p>Given the choice of only taking one of those two lenses, I'd take the 40. If you have the 40 and you need the 85, you can always crop; you lose pixels but you still get the shot. If you have the 85 and you need the 40, you don't get the shot.</p>

<p>Photo editing, for the most part, needs CPU power and RAM. I didn't see you mentioning what software you use, but most photo editing software these days can make good use of 4 or more cores*. Don't skimp on RAM because photo editing is a memory hog and things will slow way down if the computer has to keep swapping things in and out of RAM. RAM isn't all that expensive these days anyway; 16 GB should be good for most photo editing purposes.</p> <p>Before deciding on what graphics card to use, have a look at what (if anything) your software can offload to the GPU. I use Photoshop Elements, which does not use the GPU, so I could use anything from the integrated graphics in my CPU to a $1k+ graphics card and it would make no difference to performance (ditto for Canon's DPP RAW converter). The full Photoshop product can use the GPU for a limited subset of operations, and whether a high-end GPU will improve performance for a Photoshop user is going to depend largely on what operations they do.</p> <p>I'd suggest an SSD for the operating system and applications, and mechanical hard drive(s) for data. Mechanical drives are a better fit for large data files like images as they cost way less per unit of storage and give their best performance on sequential access, which is what you're doing when you open or save an image. Whether you go with a single hard drive or a bunch of them in RAID, or maybe a NAS, is probably going to be more a matter of personal preference in your application.</p> <p>*: How does hyperthreading alter this? All else being equal, hyperthreading is better than no hyperthreading but not as good as more cores. So (again, all else being equal) my 4-core hyperthreaded CPU will outperform a 4-core non-hyperthreaded CPU, but would lag behind a 6-core non-hyperthreaded CPU.</p>

<p>If you have the opportunity to rent one of these, that would definitely be a good idea. It's a relatively inexpensive way to find out if the lens fits your needs.</p> <p>Another thing to consider is buying used, particularly if there's a reputable photo equipment store in your area that sells used equipment and has some sort of guarantee (e.g. a policy that lets you return it for a full refund within a certain time period if you're not satisfied). I've bought a couple of used lenses over the years, including a 300/4L IS USM that would have been too expensive for me to buy new, and I've been happy with both.</p> <p>Speaking of the 300/4, that gives me an idea. I don't know how much extra reach you need beyond 300, or whether you need a zoom. The 300/4 is faster and much sharper than your 75-300, and with a good 1.4x teleconverter it's a 420/5.6 that's quite good (particularly if stopped down a bit) and retains autofocus. If 420 is enough, or close to enough, this might be an option. Canon also has a 400/5.6, though it lacks IS and you'll lose autofocus if you use a teleconverter. Both the 300/4 and 400/5.6 are (by the standards of professional telephoto lenses) relatively affordable, lightweight, and compact. Just a thought.</p>

<p>I switched from DPP3 to DPP4 when they added support for my 7D (mark I) and didn't find it to be a difficult transition to make. As someone else suggested, you can have both versions installed on the same PC (I do), so while it's not ideal to have to use the older version for your older files, there's no technical reason why you can't.</p> <p>If I was in your shoes, I'd probably give DPP4 a try. It's a free download and the worst that will happen is that you try it and find it doesn't do what you need, and then you go off and buy (or at least download a free trial of) LR. But since you say that DPP3 already has the capabilities you need, then I think DPP4 will also do what you need.</p>

<p>Puppy's recollection is correct: the EOS 3 and the 550EX were released together and are perfectly compatible with each other.</p> <p>The knob or lever (depending on flash model) used to use friction to hold the flash in the shoe, but at some point* Canon added this locking pin to make the attachment more secure. The pin fits into a hole in the shoe, and that hole is purely passive, so barring an unlikely manufacturing defect, the only thing that is likely to go wrong with the hole is that it might get plugged with debris. The pin should extend when you use the locking mechanism, though it should also be spring-loaded so that it remains compatible with older bodies that lack this hole.</p> <p>I don't know why your flash can use this hole on one body but not on the other; sorry. Hopefully something I've written will trigger a "hmmm" moment that will help you figure it out. In any case, since this flash is compatible with older bodies, including ones that lack the pin/hole mechanism, it <em>should</em> be able to hold itself in the shoe by friction even if the pin doesn't mate with the hole.</p> <p>*: One of the images in my old copy of the EOS 3 brochure shows a hole, so presumably the locking pin was already a Canon feature by the time of this body and flash.</p>