Canon 5D vs Nikon D5100 vs Pentax K5...Please help!

[acroreef]

Dear folks:
Now I'm very confused. I was debating between a Nikon D5100 and a Pentax K5. But I spoke with a very informed gentleman from Adorama who suggested I consider a full frame Canon 5D, which can be purchased used for approximately $800. My main focus is landscape and architecture photography and my first lens will be a wide-angle. I will be using a tripod and taking multiple exposures to produce images for HDR in post processing. My main concern is image quality. I was also told that I could print an image up to approximately 24" by 36" and still retain excellent image quality with the 5D.

Your suggestions are greatly appreciated.

Dave

[William W]

On the face of it appears the salesman knows his stuff.
Did you ask for him to give detailed and specific rationale for his suggestion?

Certainly a FF Camera allows much more scope with (inexpensive) good quality Wide Angle Lenses, for example, in the Canon range the 24/2.8. . . . To get to FL = 15 mm for an APS-C camera and to hold the same quality, you will be buying a much more expensive zoom lens.

And FF when compared to APS-C arguably offers a much better range of W/A lenses, generally – perhaps the EF 16 to 35F/2.8 L MkII as one example where there is not an equivalent.

Also, for Architecture the Tilt / Shift lenses are the most flexible (for Landscape use these lenses are also applicable) and to allow the full range of movements a FF camera is required. The TS-E 24 has recently has had the MkII version released: the original is a very good lens and you might look at one as a second hand purchase.

A 5D will have no problem making excellent display prints to 36 x 24 inches, using HDRI.
A 5DMkII would be arguably better. You are looking at a K5 new?, yes? A second hand 5DMkII would be about the same price as the Pentax, I expect.

There are similar Camera and Lens examples one could make the Nikon Range.

But alas, as Pentax doesn’t seem to be moving with any haste toward a digital 135 format camera (aka “Full Frame”) then all those very nice manual Pentax W/A lenses can stay in the Lens Cupboard.

WW

[Colin Southern]

Hi Dave,

At the end of the day, the camera really doesn't make any difference.

A full frame camera like the 5D will allow you to get a wider field of view with an EF lens (Canon), but you can get the same FoV with lenses designed for crop-factor cameras (although the same lenses can't be used on FF cameras). Generally, one needs wider-angle lenses for architecture photography (interior anyway). There is a popular misconception that "the wider the better" for landscape, but if used incorrectly (as most manage to do) you end up with a bland shot that takes in a huge field of view, but offers little by way of detail or interesting features. I do a LOT of landscape shooting and these days I'm using my 70-200mm lens more and more.

HDR is an "art" in itself. Many times folks confuse HDR with the over-saturated - over-clarified "mush" that we seem to see a lot of, whereas in reality, these "looks" aren't HDR at all - they're simply processing choices. For the most part, if you're finding you have the "need" for HDR for architecture photography then you're probably not lighting the scene correctly, or you're not making full use of the dynamic range of the camera (most capture around 12 stops - and most monitors only display around 6 - so one typically has another 6 stops that isn't being used ... that SHOULD be used). So in most cases, HDR just isn't needed.

I'm often asked "what's the biggest you can print with a xMP camera" - but it's an invalid question. In reality you can print as big as you like ... the real question is "for a given size print, how close will I be able to get before I notice the image degradation". In reality, the bigger the print, the further back people view it from. I have 22 x 44" prints hanging on my gallery walls that were shot with an 8MP camera and with a 21MP camera - and from a normal viewing distance you can't tell them apart. Also - from up close - it's STILL difficult to tell them apart. People get far too carried away with "pixel peeping"; if I were to come into your home to check how clean your carpets were, would you expect me to get down on my hands and knees with a magnifying glass, or just stand at the door and look? Same for prints.

So in summary ...

- Doesn't really matter what camera you get

- Lens selection is important

- Lighting is important

- Post-processing is important

- MP aren't important.

[rpcrowe]

I would tend to agree with the Adorama salesperson...

$800 USD for a Canon 5D classic is a pretty fair price.
http://www.ebay.com/sch/Digital-Came..._dmd=1&_ipg=50

The D5100 is a bit less expensive used...
http://www.ebay.com/sch/Digital-Came..._dmd=1&_ipg=50

A used K5 seems to run somewhere in the same range as the 5D Classic...
http://www.ebay.com/sch/Digital-Came..._dmd=1&_ipg=50

This snapshop.com comparison is actually not valid as a comparison becuase it is comparing the new cost of a D5100 with the new cost of a Canon 5D. Obviously, $800 for a used 5D is nowhere near the original cost of a new 5D. It also does places a lot of emphasis on video, and live-view.
http://snapsort.com/compare/Canon_EOS_5D-vs-Nikon-D5100
I don't necessarily agree with the results of the snapshot survey but, snapshot.com can provide a comparison of two cameras...

The Canon 5D is a full frame pro camera while the D5100 is an entry level crop camera. I certainly won't say that you CANNOT shoot landscape and architectural photography with a crop camera (I do it all the time). However, my crop cameras are general purpose cameras and not aimed only at landscape and architecture.

I am very happy with my Canon 1.6x crop cameras but, none of them are ENTRY LEVEL and the 7D is presently the flagship of Canon crop DSLR cameras. I certainly would not choose the D5100 over the Canon 5D Classic for landscape and architectural work despite the slight differences between their used prices.

The strengths of the 5D seemed perfect for architectural and landscape work while the weaknesses (primarily slower AF) would not be a weakness when shooting landscape.

The snapshot states that there are MORE lenses available for the D5100. However, I wonder if there is a lens as good as the 17-40mm f/4L at the price that the f/4L sells for used?

As far as the K-5, I know absolutely nothing of that camera.

[Manfred M]

Tough call on this one.

The 5D is rather long in the tooth (introduced in 2005), so it is three generation old technology (the 5D Mk II came out in 2008) and the 5D Mk III has just started shipping in the past month or so. The full-frame sensor on the 5D is the reason this recommendation is being made and if you are looking at printing to the 24" x 36" size frequently, that could be a consideration. On the other hand, I do a lot of printing to close that size (lots of 17" x 22" and some 22" x 34") with shots from a 12 MP crop-frame camera, and as long as I don't crop too severely, the quality is great.

The other consideration that might drive you to full frame is if you plan to shoot ultra-wide angle, rather than just wide angle. There is a limit to how wide you can go on a crop frame; the widest (not counting a fisheye) is around a 10 or 11mm lens; which translates to a 15 or 16mm lens full-frame equivalent. You can certainly get down wider than than in full-frame lenses. Grant you, lenses that are this wide angle in full-frame are extremely expensive.

Downside of a full frame is of course lens costs; you cannot buy the less expensive crop frame. As an example, my highly regarded f/2.8 11-16mm APS-C Tokina lens cost me about one third as much as my full-frame f/2.8 14-24mm Nikkor.

My own personal bias would drive me to either the D5100 or the K5, just because I have a lot of problems putting my money into a camera body that is more than one generation out of date. The sensor in the 5D really does not score all that well when compared to more modern cameras. On the well regarded DxO test, the 5D scores 72, the D5100 scores 80 and the K5 scores 82.

http://www.dxomark.com/index.php/Cam...%28metrics%29/

I'm not sure if this helps or just confuses matters a bit more....

[Snarkbyte]

Out of the 3 candidates you mentioned, I would have to go with the 5D for landscape and architecture shooting. While all 3 of your candidates are capable of the job in terms of IQ, there are some other practical matters worth considering.

First, it's worth noting that the Nikon D5100 isn't weather-sealed, and dust, wind, and rain are definitely concerns for a landscape shooter. The Canon 5D, while an older model, is a higher-end prosumer camera, built to higher standards and should withstand conditions better than either the Nikon or the Pentax. Also, I think the full-frame 5D would be more suitable for architecture. As others have mentioned, APS-C lenses can obtain the same FoV for the crop-sensor bodies, but buying these lenses could limit you to crop-body cameras in the future, because of lens investment.

[Manfred M]

Perhaps I know a bit too much about how seals work and how easily a bit of sand or grit can ruin an o-ring or a seal (unfortunate personal experience speaking here; though not on a still camera), I've never let a bit of rain (torrential downpour or just a bit of drizzle) interfere with my photography. I use a rain hood with my high-end body (D800) and sealed pro lenses and with my less robust amateur equipment (D90). I look at weather seals more as an insurance policy and hope that with my work methods, I never actually have to test them.

No, I would not worry about sealed cameras or not. I've got proper rain hoods and at times have shot just using a plastic shopping bag with my D90.



This shot was one of hundreds taken over three days in a temperate rain forest (it rained pretty well constantly) using a rain hood, D90 and unsealed Nikkor 80-400mm lens. You can see the streaks of rain in the image.

[Scott Stephen]

The "very" wide angle lenses tend also to be "very" expensive and "very" distorted. Depending on how much very wide shooting (as compared to merely "wide" shooting), you can get less distorted results, using lenses you already have, by stiching 2 images together in photoshop. That gives you a wider view than even the most expensive wide angle lens could do in a single shot. You can knit 4 or 5 together and make a panorama if you like.

Note there is a trade-off in going Full Frame. You do gain on the wide angle, but you lose out on the free "bonus" length. Buying long lenses (with good quality) is the most expensive thing in photography. Price an f/2.8 long lens (300mm and 400mm) or a decent really long one (500 or 600). The Canon and Nikon crop-sensors give you a "free" 1.6x (or 1.5x) extension on the length of all your lenses. If you ever shoot your kids playing baseball, or a bird, or anything other than landscapes and close poirtraits, then you may find you miss the free length more than you would appreciate the wide angle of FF.

By "free" length, I mean that adding a teleconverter (1.4X) to your setup loses you a stop of light and slows your autofocus. A 2x teleconverter loses you even more light, plus it tends to rob sharpness badly, plus probably eliminates autofocus on all but the very brightest lenses. You don't lose any of these things with a crop sensor Canon (or Nikon).

[Manfred M]

The Canon and Nikon crop-sensors give you a "free" 1.6x (or 1.5x) extension on the length of all your lenses.

I know that this is commonly believed out there, but unfortunately, it's really not what happens. Take your full frame image and crop out a portion of the image that is the same size as a crop frame sensor, and you will find both images are identical in magnification. A specific focal length lens will alway project the same size image, regardless of the sensor (or film) size.

[Colin Southern]

I know that this is commonly believed out there, but unfortunately, it's really not what happens. Take your full frame image and crop out a portion of the image that is the same size as a crop frame sensor, and you will find both images are identical in magnification. A specific focal length lens will alway project the same size image, regardless of the sensor (or film) size.

Ultimately it comes down to the MP involved. My 21MP 1Ds3 (FF) will capture more information than a 10MP 1D3 (1.3x) when cropped to the same field of view - but if the two cameras have identical pixel counts then the crop-factor camera does indeed give you a psudo focal length increase. Sure, the crop factor camera is discarding a portion of the image circle - BUT - it's also limiting it's light recording ability to the same reduced image circle area - whereas a FF camera image - cropped to the same field of view - would be discarding many pixels (and thus be at a disadvantage to the crop-factor camera with regards to resolving power in that situation).

[Manfred M]

Ultimately it comes down to the MP involved. My 21MP 1Ds3 (FF) will capture more information than a 10MP 1D3 (1.3x) when cropped to the same field of view - but if the two cameras have identical pixel counts then the crop-factor camera does indeed give you a psudo focal length increase. Sure, the crop factor camera is discarding a portion of the image circle - BUT - it's also limiting it's light recording ability to the same reduced image circle area - whereas a FF camera image - cropped to the same field of view - would be discarding many pixels (and thus be at a disadvantage to the crop-factor camera with regards to resolving power in that situation).

On the other hand, my crop factor 12MP D90 has a lot fewer pixels and my 36 MP full-frame D800, so I have a lot more room to maneuver on the full-frame image. In photography, like anything else, you have to understand that we are managing tradeoffs, and the photographer has to understand these various relationships in order to make the best call for his or her needs.

[Glenn NK]

Dear folks:
Now I'm very confused. I was debating between a Nikon D5100 and a Pentax K5. But I spoke with a very informed gentleman from Adorama who suggested I consider a full frame Canon 5D, which can be purchased used for approximately $800. My main focus is landscape and architecture photography and my first lens will be a wide-angle. I will be using a tripod and taking multiple exposures to produce images for HDR in post processing. My main concern is image quality. I was also told that I could print an image up to approximately 24" by 36" and still retain excellent image quality with the 5D.

Your suggestions are greatly appreciated.

Dave

I can't help you with the Nikon or Pentax cameras, but am familiar with the 5D series (more so the 5D and the 5DII which I have).

The full frame sensor should be a better choice for landscape work, but there is one factor that many of us forget or don't think of - lenses.

Although quite expensive, the Canon TSE lenses are a landscape and architecture dream come true - in particular the Canon TSE 24 Mark II which I have. The optics of this lens are excellent, and the tilt and shift controls greatly facilitate both landscape and architectural photography. For example, when I shoot a "seascape", I usually shoot with the lens wide open (f/3.5), because the resolution wide open is as good as (actually better than) stopped down to f/8. With a flat scene, using the tilt function, I can achieve focus from the tripod feet to infinity - dof is no longer a problem. Using the shift function, buildings no longer are tapered.

I mention this because in time, no matter what body you choose, it will be replaced. Lenses on the other hand, last much longer, so wiser photographers build their kit around the lenses, not the bodies.

Glenn

[Hazeb1]

Aloha Dave,

I really enjoy my D5100. I believe the image quality is great. (same sensor and processor as a D7000)
One thing you might note however, is that being an entry-level DSLR, Nikon decided to leave off features like high speed sync for flash. Other features may be buried in menus. You may not need FP sync, but there may be other "pro" features you may need, that may not be available or a pain to access on the D5100. If you can live with those issues, the D5100 is definitely worth considering.

Good luck!!

[Colin Southern]

On the other hand, my crop factor 12MP D90 has a lot fewer pixels and my 36 MP full-frame D800, so I have a lot more room to maneuver on the full-frame image.

If you (with your D800) stands alongside someone with a D90 - with the same lens - and you subsequently crop your image to the same field of view, you'll end up with roughly the same amount of pixels - so in terms of cropping from a FF camera -v- using a crop factor camera to give you a pseudo focal length increase, then no - it doesn't give you "a lot more room to manoeuvre" (it does of course give you more options when cropping, but that's not really an issue or the issue).

Point I'm trying to make is that when comparing "like for like" megapixels on a FF -v- crop-factor camera, the crop factor does indeed give one a pseudo focal length increase, and that may very well equate to a big difference in lens cost (eg the cost difference between a 400mm lens that the crop-factor shooter is able to use, -v- the 600mm lens that the FF shooter must buy to get the same FoV).

So when someone says "The Canon and Nikon crop-sensors give you a "free" 1.6x (or 1.5x) extension on the length of all your lenses" then (assuming cameras of equal MP) then IMO for all intents and purposes they're quite right. Yes - we both no that the actual focal length / magnification isn't changing, but the NET RESULT is equivalent to a focal length increase none-the-less.

[Manfred M]

Colin - I quite agree that from a perceptive standpoint, if one compares frame to frame this is exactly what happens. In my case the D90 has 12MP image and the same crop on the D800 would have 24MP (36MP / 1.5), so in theory, I will have twice the resolution using the same lens, but I do have to spend the extra money on a full-frame lens to get that resolution and of course the camera body cost three times as much as well as I paid for the D90 as well.

I remember have a similar discussion with a superzoom user who suggested I was silly wasting my money, as my longest lens gave me 400mm, while for about 1/10 of the money he had the equivilent of an 700mm lens on his cheap camera.

I also shot a fair bit of video using a pro HD MFT video camera, and of course my 100 - 300mm long lens on the video camera gives me the same framing and perspective as the 400mm on the D90; 600mm equivilent. In video (and if you are not planning to crop), this is important. And you are right, the MFT lens cost about 1/3 as much as the full-frame Nkkor 80-400mm lens, so that is a consideration.

The main reason for my making the point is that in photography, as in life, there is "no free lunch" and people should understand what the various tradeoffs mean.

[Colin Southern]

In my case the D90 has 12MP image and the same crop on the D800 would have 24MP (36MP / 1.5)

Um, no. We're dealing with area, so you have to square it -- leaving you with about 16MP after cropping.

35.9mm x 24mm = 861.6 sq mm (D800)

23.6mm x 15.8mm = 372.88sq mm (D90)

-> 372.88 / 861.6 = 0.43 (ratio of sensor areas)

-> 36 x 0.43 = 15.58MP

15.58 is only 20.77% more than 12.9 - so only 20.77% more pixels compared to a D90, not 86% more.

[rpcrowe]

I really think that the amount of pixels is a bit overhyped! Cameras with lower pixel counts can still acomplish a lot of photography and achieve quite decent results...

After all, I have had no trouble making 8x10 inch to 11x14 inch prints out of my cameras which have 8mp to 10mp. In fact, I have some excellent 8x10 inch prints with files from my 6.3mp Canon 10D. Sure, if I consistently needed prints in excess of those sizes, I would choose a full frame camera with more pixels.

I use top-line lenses and probably 90% of my imagery is done with either the 70-200mm f/4L IS or 17-55mm f/2.8 IS lenses shooting on a pair of cameras. Having a focal range from 17mm to 200mm (I don't really miss the 55-70mm gap) at my finger tips allows me to frame my images almost exactly how I want then framed. I need to crop very slightly. Therefore, the ability to drastically crop, is not necessary to my way of shooting.

When bombs "Little Boy" and "Fat Man" were dropped over Hiroshima and Nagasaki respectively, they produced equivalent 15 and 22 kiloton yields. These are tiny yields in comparison to the latest thermo-nuclear devices however, they were quite sufficient do do the intended job. I think that this can also be true with digital cameras which have lower pixel counts. The suffice for most work if the shooter is competent, good glass is used and proper post production procedures are followed.

[Manfred M]

Yes Colin - you're right. I know that but somehow my mind was asleep while writing my response; regardless the answer is still the same, you get a virtual improvement in size that can be matched by a crop of the image.

From a real and practical standpoint, I find I actually get larger images as I try to get closer to the subject matter in order to fill the frame. This is not always possible, of course. A bird up a tree, ducks on a lake, etc. limit how close you can get

[Colin Southern]

Yes Colin - you're right. I know that but somehow my mind was asleep while writing my response; regardless the answer is still the same, you get a virtual improvement in size that can be matched by a crop of the image.

From a real and practical standpoint, I find I actually get larger images as I try to get closer to the subject matter in order to fill the frame. This is not always possible, of course. A bird up a tree, ducks on a lake, etc. limit how close you can get

Manfred,

I'm sorry, but I'm starting to lose track of what you're getting at here. Are you saying that a FF camera can take as good a photo as a crop factor camera with the same lens provided you walk closer to the subject to achieve the same field of view? If so then of course -- but that was never what we were talking about. In Post #8 Scott said "The Canon and Nikon crop-sensors give you a "free" 1.6x (or 1.5x) extension on the length of all your lenses." - in Post #9 you replied with "I know that this is commonly believed out there, but unfortunately, it's really not what happens." - but in terms of real world results (again, assuming a comparison between cameras of similar MP counts) then this is - indeed - EXACTLY - what happens. Yes - I'm sure we can all appreciate that an image from a FF camera can be cropped to the same FoV - but that FF camera will need to have well over twice the number of pixels of the crop-factor camera to capture as much information for that same FoV.

One needs to remember that although the focal length of the lens isn't actually changing - and that crop-factor cameras do in fact discard about 57% of the image circle - they also typically have much higher pixel densities than FF cameras (when comparing similar generation cameras), which more than compensates.

[acroreef]

Dear Folks:
I decided to purchase the Canon 5D, used, in pretty good condition. I also ordered a Sigma 17-35, 2.8 - 4 lens. (I'm a bit concerned about this lens because the reviews are quite mixed, but that is all the budget will support now, in the wide realm). Thanks again to all for the helpful and informative information. I look forward to learn and grow in this very stimulating art form.

Dave