12mp vs 24mp

<p>I currently have a Nikon D90 and love it. I use a AF-S Nikkor 70-300mm 4.5-5.6 VR SD lens for bird/wildlife. I can get just close enough to subjects to produce adequate shots. I want to step up the IQ of my shots and am considering a Nikon D5300 hoping that doubling the mp's will go a long way toward my goal. However, will the sharpness of the 70-300 lens be up to the task? In other words just how much can I expect in increased IQ with my current lens and a D5300 body or am I better to spring for a longer lens and stick with my D90 (at considerably more cost). Thanks for any advice.</p>

<p>The 70-300 is an FX lens and for DX bodies you are only using the central, optically better, part of the lens. So it will be optically up to the job. This is one of the best lens bargains that Nikon makes and it works well on my D600 FX.<br>

However, a D5300 body is a downgrade from a D90 in terms of features and you might be better to consider the D7100 instead. The D90 was upgraded to the D7000 then to the D7100.</p>

<p>Keep in mind that while you are doubling the megapixel count, you are not doubling the resolution. While you will see a small difference if you are making huge posters or need a bit more cropping ability as compared to upsizing your files, for most typically sized prints you likely won't see any difference. Is it still worth the upgrade? Yes - you will likely enjoy the D5300 over the D90.</p>

<p>To see a significant improvement you need both upgrades. I would start with the lens. The D90 has a focus motor, and any Nikon body upgrade I would consider would also have a focus motor. Therefore you have the option of buying a used non-AF-S lens. The next best affordable lens would be the Nikon 300mm f4 ED AF. keh.com has one for $465. The AF-S versions are 2 to 3 times more. The AF-S are faster to focus in in-flight situations but until you can make that upgrade good technique can go a long way. </p>

<p>With the 300/4 you can consider adding a TC-14e giving you a 400/5.6 lens (again used is great and the first version is still excellent.) </p>

<p>I have bought mainy used lenses over the past 20 years with great success. I have bought at least three lenses from keh.com that I have been very happy with. </p>

The first hing you can do to help your technical quality of your photos is to use (if your camera has it built-in) the camera's

autofocus micro-adjustment setting to tune the camera body's Autofocus system and lens to each other. I use the

LensAlign Mark II target and accompanying FocusTune software with all of my cameras and lenses.

Http://www.lensalign.com . You don't need to the FocusTune software but it greatly simplifies the process of finding the

best AF micro-adjust setting for that unique lens and body combination.

 

There are a couple of other targets available, and I've tried them, and you can even do it without a target but the

LensAlign + FocusTune just works best for me. It will take you maybe a half-hour to set up and tune a lens and camera

combination but after that, tuning the same camera for another lens takes less time.

<p>For your shooting, do you have enough light to stop your 70-300 down a little when at the 300mm end and keeping enough shutter speed and keep the ISO down in the quality range? The 70-300 my friend has is a pretty good lens, but not that great fully open at 300mm.</p>

<p>Before you spend money, maybe you should rent a 300/f4 to compare to your current lens. I have had both AF and AFS versions, and agree with John above that the older 300/4AF is a bargain at half the price of the AFS version with most of the performance. There is a good bit of difference between the 300/4 I have and my friend's 70-300AFS VR.</p>

<p>The Nikon TC-14 would not provide AF (or mount without modification) with the early 300/4AF, but the less expensive Kenko Pro 300 does work well with the lens.</p>

<p>And yes, focus fine tune can and typically does make a big difference as Ellis mentions. Not sure if the D90 or D5300 cameras have that feature. I do find that the higher MP cameras seem to need focus tuning to get the best IQ, the 12MP bodies seemed to be a little more forgiving.</p>

<blockquote>

<p>Keep in mind that while you are doubling the megapixel count, you are not doubling the resolution.</p>

 

</blockquote>

<p>I think you should explain this more. I believe he is doing exactly that unless you wish to consider lens quality or airy disk issues. <br>

He wishes to crop more. The 24MP camera will do that better than his D90. A lot better. Then there is the issue of overall better electronics. The D5300 would be a significant improvement. There is the issue of more than a stop better dynamic range and greater color depth.</p>

<p>I agree that the D7100 would be a better choice and that the 300 F4 is a nice lens. </p>

<p> </p>

<blockquote>

<p>I want to step up the IQ of my shots and am considering a Nikon D5300 hoping that doubling the mp's will go a long way toward my goal.</p>

</blockquote>

<p>It isn't a simple a statement as may seem. Yes, possibly the D5200 or D5300 could be the answer, but it depends a lot on what exactly you find lacking now. Cropping may well not be the thing you actually need - it will not only reveal any flaw in the 70-300VR more, but also in your technique. Crop aggressively and the advantages in better high ISO performance are easily lost to the cropping as noise becomes more apparent at larger magnifications. Maybe a longer lens, or a better tripod/monopod are more effective ways to get what you want, or even better post-processing. It really depends on what is lacjing today, and nice as it sounds "a step up the IQ" is pretty meaningless (sharpness, contrast? the lack of both can have a multitude of reasons).<br>

I'm not saying to dismiss the idea of a D5300 (or D5200 which is a pretty sweet choice too), ideally a D7100 to keep roughly the same handling as your D90 has. But something like a Sigma 50-500 or the new Tamron 150-600 could also be worth considering. The 300 f/4 with a TC14 is also a very solid choice and a seriously good lens.</p>

<p>I'll assume you have the budget for either one of these steps; if you can do both, fine, solved :-) If not, post an example of your current images and describe more precisely in which ways you want the image quality to be better - that could help a lot more in deciding whether a body or a lens would make most sense.</p>

<p>While the 70-300mm AF-S VR is a bargain, it is not at its best at 300mm. I have used that lens on the 16MP D7000, and even 16MP seems to out-resolve it at 300mm, f5.6. In other words, putting a 24MP DX body behind it is kind of meaningless as far as resolution is concerned. You can put 1000MP there but you won't get any more effective resolution than 12MP, 16MP since the lens is not up to it.</p>

<p>If you are into wildlife photography, especially birds, you need longer, better lenses. There are not a whole lot of short cuts unless you are casual about quality.</p>

<p>If the OP is going to upgrade the lens, I would stick with AF-S lenses. For one thing AF tends to be faster. There is also no Nikon teleconverter that can maintain AF with screwdriver AF, AF-D lenses.</p>

<p>Does <a href="http://www.dxomark.com/Lenses/Compare/Side-by-side/AF-S-VR-Zoom-Nikkor-70-

300mm-f-4.5-5.6G-IF-ED-on-Nikon-D90-versus-AF-S-VR-Zoom-Nikkor-70-300mm-f-4.5-5.6G-IF-ED-on-

Nikon-D5300___262_439_262_919">this</a> help? Sadly, I don't think DxO have tested the 300 f/4, though

I can vouch for its goodness.</p>

<p>Thanks for all of the responses, this is the first time out for me in asking for photo advice on a forum. It is very pleasant to have such knowledgeable folks with whom to converse who also know how to write well and communicate respectfully. You (collectively) have given me so much to think over. I will consider your advice and re-frame any unanswered questions in a more precise way.<br>

Again, thanks<br>

Doug</p>

<p>1. @Rick M. :the linear pixel sharpness is only sqrt(2)=1.4 because the double population # of pixels is half vertical, half horizontal. It's a math thing...However, the lack of AA filter will also add an edge (noticeable =?)<br>

2. I'm in a position similar to the OP, in considering getting D3300 (cheapest 24 mp sensor) in addition to my D300, and, like the OP, wonder about my lenses being capable of delivering the goods.<br>

2.1 = 10-20 sigma(early) - not noted as being top of the line resolution....<br>

2.2 = 18-70 nikon - never seen evidence of anything else with this range (which suits me) being sharper on Dx<br>

2.3 = 50 mm f1.8 D - would have to manual focus on D3300, or upgrade to bigger heavier G<br>

2.4 = 70-200 sigma (latest non-macro) +(1.4 tc)- currently happy with it - usually use it f4 & beyond<br>

2.5 = 100mm sigma macro - good performer<br>

3. noise: newer sensor tech would give less high iso noise - 1 stop? 3 stops??<br>

4. dynamic range : expect better with newer tech, intertwined with pt 4 above<br>

5. ergonomics: D300 wins - even the higher weight offers more mass for steadiness.+popup flash as commander +AF fine tune, etc<br>

6. color discrimination - ??<br>

7. movie capability - not interested -yabut if it was there, I'ld probably try it... and???</p>

 

If you give me a 12 ounce can of beer, and then you replace that with a 24 ounce can of beer, the second bottle has twice as much beer.

It might not be twice the height of the first bottle, nor may it have twice the diameter, but it has twice as much beer.

 

A camera with 24 megapixels has twices as many pixels as a camera with 12 MP. You have doubled the number of number of pixels in

the image rectangle. That is most definitely twice the resolution by every imaginable standard.

 

That doesn't mean that you double the number of pixels on each side of the rectangle. That would require four times the resolution, I.e.

four times as many pixels (12 to 48). And it would require lenses with infinite resolution, which don't exist anyway.

 

So, yes, be assured that 24 MP captures twice as much information as 12 MP, regardless of what some helpful poster suggests.

 

---

 

I'm certain that people will argue with the above, and they are welcome to state their case. I'll answer the op's question in my next reply.

Suggestions for getting the most out of a high-resolution sensor:

 

A. Use the best lens that you have available and don't put a cheap filter over the front element

 

 

B. Limit the impact of camera shake by using one or more of the following

 

1. A high quality tripod, head, and camera plate

 

2. VR (hand held)

 

3. Fast shutter speeds (1/250 s and faster - especially when shooting hand held)

 

4. Make flash your dominant light source

 

 

C. Ensure focus accuracy using a number of techniques.

 

1. Use Contrast detection autofocus In Live View when shooting subjects that don't move

 

2. Use live view with magnification when you need to use manual focus

 

3. Learn how to micro fine tune the AF of your lenses on your camera. It's not rocket science, but you'll need to invest a

bit of time to ensure good results.

 

 

D. Use the lowest ISO that is practical for each shooting situation in order to control noise.

 

 

E. Expose correctly.

 

 

F. Select subjects, compositions, and lighting that are worth shooting at 24 MP.

<p>Doubling the number of pixels gives you 1.41 times the number of pixels horizontally, and 1.41 vertically, as stated above. If resolution is defined as the number of lines per millimeter that can be distinguished, does 1.41 times the number of pixels along a given axis resolve twice as many lines per millimeter? (Assuming of course, a perfect lens.) Just asking -- the math for that calculation is beyond me.</p>

<p>The ideal upgrade would be the D7100. I plan on getting one soon. I got my wife a D3200 for her birthday this past October and the images from her 24mp sensor just blows the images from my 12mp sensor D300s away. The better image quality is the reason for my upgrade.</p>

<p><em>A camera with 24 megapixels has twices as many pixels as a camera with 12 MP. You have doubled the number of number of pixels in the image rectangle. That is most definitely twice the resolution by every imaginable standard.</em></p>

<p>That's not how resolution is defined in imaging science. It is a linear measure, x line pairs (or cycles) per mm for example, or in the case of sensors another definition is line pairs (cycles) per image height. It corresponds the highest spatial frequency which can just be detected. Another definition of resolution (one that is common in medical imaging) is the resolution is the shortest distance between two objects in the image which can be reliably seen as separate entities. In any of these definitions resolution is a linear, not area measure.</p>

<p>I recommend the OP to consider the AF-S Nikkor 300mm f/4D, which is a very high quality lens and not unreasonably expensive. I think it would contribute more to the image quality than the move from 12MP DX to 24MP DX, though the latter can't hurt and the newer cameras have other advantages, such as in autofocus (especially in the D7100 which has Nikon's state of the art AF system). However, the total cost of the purchases does go up. And it is possible that within a few years there will be a new VR version of the 300/4, but in my experience with high resolution camera bodies, VR isn't all that useful (the use of a fast shutter speed and/or tripod are more effective in countering blur, as VR has limited precision and can be unpredictable).</p>

<p>I had a Nikon 300mm AF while it is a great lens,it is not a great lens for birding unless they are not moving...</p>

>> That's not how resolution is defined in imaging science. <<

 

Double? Quadruple? Is 8x10 double the resolution of 4x5 despite having four times the film surface area?

 

It's a pedantic argument about words. Anyone who has passed seventh grade mathematics (or understands the concept

of f-stops) realizes that doubling a linear relationships increases the squared relationship by the square root of two (1.4).

 

The point is that 24 MP is a heck of a lot more resolution than 12 MP, and regardless of contentious comments, the

photographer has to be more disciplined if they hope to get the most out of twice as many pixels?

 

(May I use the phrase "twice as many pixels", or is that a flagrant violation of convention in the science of optics?)

<p>Dan, I'm only trying to help. You made a statement which is incorrect and yet you underlined it with the words "by every imaginable standard" i.e. basically you said that there can be no other way of thinking about it yet everyone who has had formal education in the matter would likely disagree with you. To clarify the matter I wrote about the established use of the word "resolution" in imaging to help avoid future errors and misunderstandings. I'm fully aware that many photographers often equate pixel count and resolution (they are related, of course, just not linearly). When making quantitative comparisons, it's best to resort to correct use of physical quantities since otherwise the communication will easily be misunderstood by others.</p>

<p>In practice the difference between VR 70-300 at 300mm on a 12MP DX vs. 24MP DX camera is not as large as the difference in pixel counts of the two sensors might suggest. Resolution of the imaging system does not increase linearly by increasing the resolution of one component (the sensor) alone, especially if the other component (the lens) is not very good. With a better lens (such as an AF-S 300mm prime), the effect of increasing the sensor resolution on the total system resolution would be more pronounced.</p>

 

<p>I'm with Dan. Our OP did not ask a physics question, he asked a photography question. Is his question really about resolution? Not really. HE wants to know what the difference between the two sensors are as a practical matter. As a practical matter the 24 MP sensor will allow him to crop twice as much as the 12 MP sensor. Why do we have to make this stuff so hard.</p>

<p>I remain unconvinced that a 24 MP camera approaches in any real sense the capabilities of any good modern Nikon lens. There are FAR more important factors in that illusive thing called 'sharpness' than can be discovered in MTF tables. Dan's comments are absolutely spot on. Technique is infinitely more important than any differences between lenses. </p>

<p>I know that this is heretical on a gearhead forum but Dan's is the very best advice to give our OP.<br>

So the take away for Doug is that the D7100 will be a far better camera for him than his current camera. It will give him better picture quality and a much better ability to crop than he has now.</p>

<p>As Doug is quietly walking through the woods, stalking the bird, getting as close as he can, the very last thing he ought to be worried about is whether his lens is going to ever so slightly affect the shot he is about to take. Put that one last. First he should consider doing all of the great things Dan suggested he do. </p>

<p>I like sharp lenses. I own some legendary ones. Not a single one of them will make up for a failure in my technique. </p>

<p> </p>

<p>I have enjoyed looking over all of the responses to my question. The breadth of information could be overwhelming but being of a technical sort I have enjoyed all of the posts.<br>

Let me do a little clarification and then some observations. Most all of my photo shooting is done at the lakes and ponds around the high plains of the Laramie, WY basin. The thin and dry air at 7200' + makes for intense and very contrasty light. The only vegetation is sage brush and scattered low growing reeds. This means stalking birds and animals is problematic at best; long shots are by far in the majority (partially due to shooting opportunities and partially because these ponds and lakes are breeding grounds and I try not to disturb the birds). The good news is that as soon as the sun rises the light is sufficient to shoot at 1/1000s at f8. With the VR engaged hand holding 300mm is very doable. Even just before sunrise I can usually shoot at 1/500s at f8. I have uploaded a picture here - <a href="/gallery">Gallery</a> > <a href="/photodb/user?user_id=8268988">Doug Jordan</a> > <a href="/photodb/member-photos?user_id=8268988">Photos</a> > <a href="/photodb/folder?folder_id=1070552">Examples</a> > "Heron 13- 1" as an example of what I am trying to overcome. It was shot hand held at 1/1000, f8 and 300mm. The distance was about 80 yards and cropped from 4280p x 2848p to 1464p x 976p. I used Raw Therapee to process and GIMP to finish.<br>

Now for observations.<br>

Considering all of your input I believe that my best bet is to go for the D7100 for the following reasons.<br>

1. With the D7100 I get some added sharpness due to elimination of the low pass filter.<br>

2. The 24mp gives me about 40% more cropping latitude.<br>

3. The D7100 gives me added dynamic range over the D90 that should help some with the contrasty light here in Wyoming.<br>

4. Being a DX body I am still using the sharpest part of my longest lens so I shouldn't lose much if any sharpness due to the glass.<br>

5. The ability of the autofocus on the D7100 to operate on lower light situations gives me more latitude in using a tele-extender.<br>

6. Without constantly improving my technique all the equipment in the world will not get me where I want to be.</p>

<p>Thanks, Doug</p>

<blockquote>

<p>"I used Raw Therapee to process and GIMP to finish."</p>

</blockquote>

<p>Try other software while you're also considering new equipment. It will make at least as much difference, at a fraction of the price, as buying a new camera.<br>

<br>

In 2012 I waffled between Lightroom and Raw Therapee, giving each a thorough try on various Nikon and Ricoh raw files. I tried many suggestions discussed by experienced RT users. In the end I found Lightroom 4 (and 5) did a better job more consistently and easily.<br>

<br>

Raw Therapee offers a lot of user adjustable parameters but after exhaustive experiments using the same set of Nikon NEFs and Ricoh DNGs as I did with Lightroom, I found that higher apparent "sharpness" in Raw Therapee always came at the cost of more unpleasant demosaicing artifacts, harsh evidence of sharpening, with mediocre noise reduction. Your heron photo looks very much like the typical results I got from RT. I consistently saw grittiness from trying to wring out more detail from Raw Therapee, often the result of the excessively strong default setting for Contrast By Detail Level (try reducing it for less gritty results).<br>

<br>

And RT is incredibly complex to adjust each parameter, while offering no perceptible advantages over Lightroom, Silkypix, Photo Ninja (an excellent alternative for folks who wish to support independent software developers like Picturecode) or Nikon's own raw converter. I got better results combining Nikon's freebie View NX2 raw converter to TIFF, followed by my old copy of Noise Ninja for noise reduction, finished up with whatever other editor I used. It may be possible to get better results with Raw Therapee, but not in a time-effective manner.</p>