7D v. 5D II in Low Light

[travismcgee]

<p>Many thanks, everyone. I really appreciate the information and opinions. I'll read everything very carefully and think about it. </p>

[philip_wilson]

<p>I have had both bodies since they came out. In terms of image quality the 5DII clearly beats the 7D. It is about 1.5 stops better on noise (the 5DII at ISO 3200 is much better than the 7D at ISO 1600). The other bog 5DII advantage at high ISO is that it is much less sensitive to slight exposure errors. The 7D gets very noisy (even in RAW) if the exposure is slightly off - especially when underexposed, the 5DII images are much more robust (I am talking about 1/3 to 2/3 of a stop errors). the 7D advantage is the AF system which is clearly better in poor light - especially low contrast. That said the centre AF point of the 5DII is quite acceptable under most circumstances. I do not know what lenses you own and what your budget is but if I assume no constraints in either then the 5DII is clearly better unless you are shooting indoor sports. Both are very good cameras but neither is perfect.</p>

[ty_mickan]

<p>Dave, you have slow lenses. If you were to purchase some fast primes, you may not need to upgrade your body at all, plus you will open up some creative opportunities.</p>

[tarek_wazzan]

<p>"<em>DxO is drivel. Anybody with eyes in his head who has actually taken the time to figure out 7D and 5D Mk II files for themselves, knows that whatever DxO is telling us, it's nothng whatsoever to do with Real World image quality or noise performance" </em><br>

<em>"dxo is a drivel....."</em><br>

<em> </em></p>

<p>Keith, and Jack..</p>

<p>Really?! how can you be so sure?!<br />it seems some other posters here have opinions that are in principle in agreement with DXO mark on this particular question..</p>

[holzphoto]

5d ii beats the 7d easily for low light.

 

you can occasionally get away w 6400 ISO on the 5d ii. Can't do that so much w/ the 7d

[dcstep]

<p>f/2.8 and f/4 are not slow lenses. You'd need a bag of primes ot cover your current range of focal lengths with little likelyhood of improving IQ.</p>

[ty_mickan]

<p>it doesn't matter which way you look at it, f2.8 and up is slow. Primes don't just offer an advantage in sharpness. A lens is a long term purchase, a body (especially nowadays) is almost a disposable one. Trying to solve the shortcomings with correct lens choice will always be preferable to 'cranking up the ISO'.</p>

[dcstep]

<p>When we were shooting ISO 25, 100, 200 and 400, what Ty says might have been true, but it's no longer the case. Yes, good lenses can last for many years and we should buy the best lenses that we can afford, but not taking advantage of the high-ISO capacities of the latest bodies and incorporating those capacities into our shooting seems foolish to me.</p>

[nathangardner]

<p>The 1D III costs used about the same as a 7D new. It's less MP, but handles noise very well. Haven't seen it side by side with a 7D, but I shoot indoor sports fairly often and, for me, I wouldn't trade the 1D for anything else.</p>

[Dave Luttmann]

<p>something that gets lost on some is the effective real iso. many ff users go on about shallow dof. but when capturing a scene in low light, it's often about being able to capture as much of a scene as possible. with the 7d, i can shoot at f2.8 with the 17-55 and get the same dof as the 5d2 shooter at f4.5. in other words, the 7d shooter can be at iso 3200 while the 5d2 shooter would need to be at iso 8000 to iso 10,000. so much for the high iso advantage.</p>

[Dave Luttmann]

Tarek, I can be sure because DxO places the Pentax Kx ahead of the Canon 7D. I own both, and I can tell you

DxO are fools if they think the Pentax is better in any way.

[scott_ferris]

No Dave that is not true. As I have said countless times, to get exactly the same image from ff you need to stop down one

stop and raise iso one stop, this effectively gives you he same dof, the same noise, same everything. Look up

Equivalence ( http://www.josephjamesphotography.com/equivalence/ ) and you will stop spouting such drivel.

 

So if you shot your 7 D at 50 mm and f2.8 and 800 iso the exact same shot, framing, dof, perspective and theoretically

noise etc could be shot with he 5 D at 80 mm, f4 and 1600 iso and both with the same shutter speed. But I know which

image I would rather process. Your suggesting the ff shooter would need to use 8-10,000 iso to match a 7d shot at 3,200

shows your lack of ff experience and general photographic theory knowledge.

[ty_mickan]

<p>David, the same laws of physics apply no matter the lens you are using. A lens that is two stops faster than another lens is always going to be two stops faster. You can use high ISO on fast lenses if required; there's no free lunches.</p>

[dcstep]

<p>Remember, noise vs. detail is a trade off. I use my 5D2 for low noise and I use my 7D when I expect to crop and want detail, as in nature photogaphy. Noise is only one element in the IQ formula and many people forget that. Our OP was concerned about noise, so that's why we're discussing that aspect primarily.</p>

<p>In the real world, having shot tens of thousands of shots with my 5D2 and my 7D, the 5D2 has better high-ISO performance and better tolerance for under exposure. They're both good, but the 5D2 wins in this particular respect.</p>

[Dave Luttmann]

Yes it is true Scott. You need to bump up the iso a good 1.5 stops. That removes the high iso advantage of

the 5d2, which was my point.

 

Ty, you muissed my point entirely. Go back and read my post again.

[dcstep]

<p ><a name="00Zkrs"></a><a href="../photodb/user?user_id=2257221">Ty Mickan</a> <a href="../member-status-icons"><img title="Subscriber" src="../v3graphics/member-status-icons/sub6.gif" alt="" /></a>, Dec 18, 2011; 10:38 p.m. said:</p>

 

<blockquote>

<p>David, the same laws of physics apply no matter the lens you are using. A lens that is two stops faster than another lens is always going to be two stops faster. You can use high ISO on fast lenses if required; there's no free lunches.</p>

</blockquote>

<p>Actually, the laws of physics have been working in our favor. Compared to when we shot at ISO 25, 100, 200 and 400, we know do enjoy a "free lunch". Decades of working in the film era led to the need for fast lenses, yet some people hold onto those old beliefs without considering the new world that we now work in.</p>

<p>Select your lenses based on the focal length, quality of the bokeh (if that's important in your application) and the match with the speed of your camera. In the digitial world, look at the corrected IQ, not the RAW, uncorrected, IQ, since what your end product is based on the processing that you do. High quality programs, such as Lightroom and DxO, adjust for geomertric errors, CA, vignetting, etc. for every focal length and every aperture for the top lenses. With these corrections, the high quality lenses, like Canon L-series, now approach the IQ of the very best prime lenses.</p>

 

[scott_ferris]

Dave, read the link, or get a 5D Mk II. There is more than one stop difference in IQ, especially at high iso, and you only

need to increase Iso one stop to achieve equivalence. So raising the iso one stop doesn't negate the ff cameras high iso

advantage even if you want to achieve equivalence. Where are you getting your "a good 1.5 stops" from? Certainly not

experience.

[Dave Luttmann]

I've used both extensively Scott. After post processing, the difference is about a stop or so...which vanishes in

print. And spare me the "read the link" nonsense. I've shot tens of thousands of frames on the 5d2, 7D, 1ds,

1ds2, D700, k20D, etc, etc. I know full well what the differences are, and how they appear on print.

[tarek_wazzan]

<p>".. dxo are fools if they think the pentax is better in anyway"</p>

<p>Dave, not to stray way off topic but, dpreview.com finds that the pentax kx is clearly better than the 60d ( which has a sensor essentially identical to 7d i believe) in low light/high iso performance, are they fools too?</p>

[scott_ferris]

<blockquote>

<p>"I've used both extensively Scott"</p>

</blockquote>

<p>Dave, then post some examples supporting your position, I have on many occasions, often shooting comparable subjects to those posted to graphically illustrate and support my observations.</p>

<blockquote>

<p>"And spare me the "read the link" nonsense."</p>

</blockquote>

<p>It is saying stuff like that that got you down your blinkered dead end in the first place. The physics is not intuitive at first, but it is physics, you can have whatever opinion you want but if it doesn't agree with the physics then you just come across as a luddite.</p>

<p>In all your only one stop difference preaching it is easy to forget that ff sensors are 2.6 times bigger than Canon crop sensors, that means crop sensors need to be 2.6 times more efficient to be equal. That they all but manage that at low iso is remarkable. But compare a 1/4 mile time with a 200 hp car and a 580 hp car, the smaller car might keep up initially but sooner or later the more powerful one will pull away, those are the same differences in sensor sizes. The bigger car won't go 2.6 times as fast, but if you need to get there 2 seconds faster everybody would take the 580 hp. Sure most of the time in normal road use 200 hp will get you there plenty quick enough, but if you need the 580 sometimes you need it.</p>

[Dave Luttmann]

<p>Scott, you do understand what 2.6x equals in terms of f stops, don't you? I've not gone into any blinkered dead end....it's you that doesn't seem to understand basic math. How many stops does 2.6X equal? And stop with the childish car HP analogies.</p>

<p>There is nothing in my statement that disagrees with physics nor basic math. The difference of 2.6X equates to, give or take, a stop and a half. And as I said, in order to maintain the same DOF in a dark reception for example, I could be at f2.8 with the 7D....and to maintain the same DOF, the 5D2 user would need to be at f4.5 or so. That means to maintain the same shutter speed, you'd need to be 1.5 stops faster.....so if I'm at 3200....you'd need to be at 8000 to 10,000 iso....1.5 stops more or so.</p>

<p>There is nothing in what I wrote above that disagrees with math or physics. And in my example, the 5D2 at iso 8000 isn't really going to be any better than the 7D at 3200....is it?</p>

<p>By the way, we've posted examples numerous times.....you just chose to ignore them.</p>

<p> </p>

[scott_ferris]

So you havent got any then? Have a good week.

[ilkka_nissila]

<p>So, your lens on 7D at f/2.8 is as sharp as the lens on 5D Mk II at f/4.5? Wow, that must be some expensive optic! All my lenses are sharper at f/4.5 than at f/2.8. Not to mention at f/2.5 vs. f/1.4.</p>

[KenPapai]

<p>Here is an image I shot last week: 7D at ISO 6400, with minimal (raw, via Adobe's ACR 6.x) noise reduction. 16mm f/2.8 at 1/15 handheld. It was DARK in there. Still a lot of detail at 6400 with the 7D and very usable. (no one else has posted an image to this thread)</p><div></div>

[danielleetaylor]

<p><em>It depends on the low light situation, if you are talking about point light sources, night skies and star trails etc. then the 50 f 1.4 does have a light gathering advantage as absolute aperture opening size is more important than f stop.</em></p>

<p>Astrophotography is a unique application and is not relevant to any other low light/high ISO situation that I'm aware of.</p>

<p><em>Your defense of crop camera additional dof is predictable but misguided. The crop camera is never "better", for want of a better word. If you want the dof of the crop camera stop the ff one down one stop and bump the iso up one stop, </em></p>

<p>The difference is closer to 1.5 stops. It is not "misguided" to believe a 7D at ISO 800 will produce a better image than a 5D2 @ ISO 2500, it's simple fact. FF fans love to talk about shallow DoF except when it works against the photographer.</p>