Comparing all (!) my cameras in low light.

[christilou]

I won't bore you with the pictures but having put up the Christmas tree yesterday evening I did the obvious thing and began to see what sort of photos I could get in the gloom with just the lamps on in the living room and the tree lights on. I began with the Pentax K-01 using the 1.8 FA77, auto iso. Having been not much impressed by that I began to try the other cameras from my stable,

Olympus EM5 + PL25mm + Oly 45mm
Fuji X Pro 1 + 35mm 1.4
NEX 5R and Sigma 30mm 2.8

Out of that little lot, the one that performed the best by quite a margin, was the EM5 and those two fast lenses. I had no camera shake, the iso stayed at around 1600 and the focus was the fastest and the most accurate. I was expecting the Fuji and it's 1.4 lens to out perform the EM5 but even though it's focus time was pretty reasonable, it still gave me a bit of camera shake and the pictures didn't come out nearly as well. The NEX was using it's alternative focus method whereby it simply lights up in a green square around the whole frame if it finds something (anything?) to focus on and it missed quite often! Not a fast lens but still. The FA77 was painfully slow to focus but I expected that. Nothing even remotely scientific but I was quite impressed as there have been occasions where I think that the buttons on the EM5 are just too fiddly and maybe I should sell it. It now has a stay of execution

 

[krugorg]

Have you been using the FA 77 on the NEX, Christilou, or did you keep the K-01?

Oh yes, and do you happen to have the kit lens for the NEX? My understanding that it and maybe only one or two other Sony lenses (including the CZ24 after a firmware update) are compatible with the PDAF hybrid focusing on the new NEXs. I wonder if you were using one of these lenses, if the Sony would do better locking focus in your test?

 

[Ray Sachs]

You know Christina, I've been doing some thinking and non-methodological testing on this question lately and I'm coming to something like the same conclusion (although with fewer cameras to compare). When I first got the X-Pro at the beginning of April I couldn't believe how good it was in low light. Better than anything I'd used before by quite a margin (except the X100, which wasn't as good, but wasn't far behind). A few weeks later I got the OMD and was amazed at how close it came to matching the Fuji. Over time I've come to the conclusion that there's very very little real world difference in the quality of the sensors and the OMD's better low light AF and its IBIS more than makes up for any deficiency.

I took the OMD with the 14mm and the X-Pro with the 18mm out one night and shot the same low light scenes back to back with both lenses at f2.5. Same effective focal length, same aperture. I've generally found both cameras very very good at ISO 3200 but with an edge to the X-Pro at 6400. But not all ISOs are created equally and I found that I was generally getting the same (or very nearly the same) shutter speeds with the OMD at 3200 as with the X-Pro at 6400. So, really, no functional difference. Now, the Fuji has its own creamy quality to its files, it's own way of shooting, it's own feel, all of which I flat out LOVE, so that camera isn't going anywhere. But it's not strictly necessary in terms of providing a real benefit in low light, relative to the OMD.

As happy as I am with those cameras, I've been going in circles with compacts. The RX100 has an amazing sensor (pretty close to the OMD/X-Pro at 3200 and f1.8 at a 28mm equivalent) but I just don't like the camera AT ALL, I love the LX7, but its really not much good above 800 - that wouldn't have seemed like much of a downside a year ago but now it feels damn near unacceptable and limiting in anything but the best light. This also screwed up my ability to appreciate the GRD3 any longer too, a camera I really loved but just sold for lack of use. The X-10 splits the difference between these cams pretty well, but getting the most out if the EXR technology in varying conditions requires a fair amount of thought. These decidedly first world "problems" will likely resolve themselves soon enough when someone comes out with a compact as functional as the LX7 with a sensor as good as the RX100, but this could be a year or two away. At this point, I'm seriously thinking about bailing on this segment altogether and just getting an EPM2 or EPL5 body to use as my "compact". When I want to stick it in a coat pocket, with the 14 or 17mm lenses, it essentially the same size as the LX7 and when I want to use a slightly larger lens, it's not too much of a burden.

When it comes down to it, there are trade offs among all cameras, but there's just no substitute for pure sensor-based horsepower. It gives you so much more flexibility in almost any situation. And now it's available in such small packages with relatively good handling. In the past several months, I seem to have become a LOT less willing to put up with limits and trade offs that I just accepted as inevitable as little as a year ago. These are VERY good times to be into camera, including very small ones.

-Ray

 

[christilou]

Kyle, I sold and then re-bought the K-01, there's just something I like about it, especially with the FA77 I have a NEX5R which came with the old style kit lens but I must admit to never having used it. I should give it a go though and see what happens. I bought the NEX mainly so that I could use my Contax G lenses although I think that focus peaking on the Pentax is perfectly useable too.

 

[Armanius]

I found that when I use the XE1, I need to have shutter speeds in excess of the "1/35mm EFL" formula to avoid getting some softness due to shake or vibration. No such issue with the OMD. As much as I want to like the XE1 more (being the most recent purchase), I'm still taking the OMD out with me over the XE1 90% of the time.

 

[christilou]

Ray, I've found in many situations in low evening light that the EM5 and a fast lens is much better at getting a useable shot than any of my other supposedly better sensor cameras. I've just bought the Fuji X10 in a sale for a good price and look forward to being able to have this with me at all times. Although I did consider getting a tiny Olympus body, the small Oly pancakes are not particularly fast and the Lumix 20mm is by all accounts rather slow to focus on the Oly bodies.

 

[Chris2500dk]

For the NEX, have you tried disabling the AF assist light? My RX100 was giving me the annoying "big green box" in low light, and turning off the AF assist slowed focus down a bit, but made it use the normal AF box which is vastly preferable.

 

[Crsnydertx]

I found that when I use the XE1, I need to have shutter speeds in excess of the "1/35mm EFL" formula to avoid getting some softness due to shake or vibration. No such issue with the OMD. As much as I want to like the XE1 more (being the most recent purchase), I'm still taking the OMD out with me over the XE1 90% of the time.

Thanks for saving me $$, Armando - I had the XE-1 in my cart at B&H but now it's pffft.... As I have no interest in the OMD, I'll be standing pat for a while...

 

[Luckypenguin]

I took the OMD with the 14mm and the X-Pro with the 18mm out one night and shot the same low light scenes back to back with both lenses at f2.5. Same effective focal length, same aperture. I've generally found both cameras very very good at ISO 3200 but with an edge to the X-Pro at 6400. But not all ISOs are created equally and I found that I was generally getting the same (or very nearly the same) shutter speeds with the OMD at 3200 as with the X-Pro at 6400. So, really, no functional difference. Now, the Fuji has its own creamy quality to its files, it's own way of shooting, it's own feel, all of which I flat out LOVE, so that camera isn't going anywhere. But it's not strictly necessary in terms of providing a real benefit in low light, relative to the OMD.

Ray, are you still shooting the Fuji in jpeg and the Olympus in raw, and if so do you think that the difference could be in the tone curve of the Fuji jpeg engine?

When DxOMark measured the E-M5 they found that it was overstating the ISO sensitivity by approx 1 stop, and when I did my own comparison between the E-M5 and the NEX-5N at higher ISOs the Sony would produce an image around 2/3 stop brighter with identical settings. Assuming that the Sony was more accurate with it's reported ISOs, it essentially confirmed DxO's numbers for me. I thought that this was a bit cheeky on the part of Olympus but then the camera is genuinely good in low light so it didn't really matter all that much. If your findings are typical then it would indicate that the Fujis might be overstating their higher ISO sensitivities by anything up to 2 stops which is no small amount. From what I have seen they're obviously still fantastic low light cameras, but maybe not by quite as large a margin as I had thought.

 

[stanleyk]

I shoot the Fuji all the time at shutter speeds as low as 1/30 (see the kitten shots). I have just not had any of the problems people have with the camera locking focus but it may be what I'm shooting. I started out using a K1000 and film so all the bells and whistles on modern cameras are confusing for me. The closest thing to auto I use is Aperture Priority. I had thought about selling my Nikon gear and getting a OMD because I did really like the EP camera I had. But...I didn't like the size of it. I just love that D700 so much when I use the 35mm F1.4 though. I rarely use but when I do it's almost a magical as the Fuji. I think I'm happy with the X100, X Pro, and D700. I just cannot see myself wanting another camera. I am selling some of Nikon zooms to get the new Fuji lenses when they come out. Why I bought a 70-200 I will never know. I just never use that lens. It was really expensive so I kept telling myself I would use it but I finally realized a couple weeks ago when I was doing some portraits for a coworker's family and it never came out of the bag, that I wouldn't. Same with the 24-70. Great lens, it never goes on the camera. Live and learn.

 

[Ray Sachs]

Ray, are you still shooting the Fuji in jpeg and the Olympus in raw, and if so do you think that the difference could be in the tone curve of the Fuji jpeg engine?

When DxOMark measured the E-M5 they found that it was overstating the ISO sensitivity by approx 1 stop, and when I did my own comparison between the E-M5 and the NEX-5N at higher ISOs the Sony would produce an image around 2/3 stop brighter with identical settings. Assuming that the Sony was more accurate with it's reported ISOs, it essentially confirmed DxO's numbers for me. I thought that this was a bit cheeky on the part of Olympus but then the camera is genuinely good in low light so it didn't really matter all that much. If your findings are typical then it would indicate that the Fujis might be overstating their higher ISO sensitivities by anything up to 2 stops which is no small amount. From what I have seen they're obviously still fantastic low light cameras, but maybe not by quite as large a margin as I had thought.

Yeah, JPEG for the Fuji, raw for the Oly. Because those are the ways I get the best results from both. Not sure about how the ISO ratings equate to some theoretical "neutral", just how they seem to compare. I'm not one to look at numbers a lot, or do these types of comparisons often, but I was curious to see if it verified what I'd observed over months of shooting. The other time I did similar tests was with the Ricoh GRD 3 and 4. The 4 did better at higher ISO, but they'd basically just overstated the ISO ratings on the new one relative to the old one - they were nearly identical in practice.

The X-Pro and OMD are both really fine low light cameras, but any differences are minor the point of insignificance, although the files do look different, so it's reasonable to prefer one to the other. I tend to prefer the Fuji slightly, but I'm really happy with both.

-Ray

 

[john m flores]

I won't bore you with the pictures but having put up the Christmas tree yesterday evening I did the obvious thing and began to see what sort of photos I could get in the gloom with just the lamps on in the living room and the tree lights on. I began with the Pentax K-01 using the 1.8 FA77, auto iso. Having been not much impressed by that I began to try the other cameras from my stable...

I'm curious Christina, where did you find the K-01 + FA77 lacking? I've found my K-01 very usable at ISO3200. It should also be noted that the 77 has a much longer equivalent focal length than the other cameras you mentioned.

 

[christilou]

John I think that it was mainly the focusing with the FA77 on auto focus. Yes, I could use it manually but it's quite difficult to hold it steady for a low light shot and get a decently sharp picture. I too have found that iso 3200 is perfectly useable I've spent the last couple of days playing with a couple of manual lenses, the SMC K 55 1.8 and the SMC M 50 f2. and have come to see that the FA77 is much easier to manual focus too!

 

[Biro]

I wonder if the E-PM2 and E-PL5 would give the same low-light performance as the OM-D. Same sensor but I wonder how much the E-M5's five-axis IBIS has to do with it.

 

[Luckypenguin]

The key difference between the IBIS in the E-M5 and that in the Pens (and any lens-based IS system for that matter) is not necessarily the magnitude of motion that it can correct for but the range of motion that it can correct for. If the motion of the camera is up-down or left-right then you're fine, but the E-M5 can also correct for pitch, yaw, and roll.

 

[Ray Sachs]

And the practical implication of the difference is that for still subjects, I can shoot the OMD a much slower shutter speeds than I ever could with any of the previous models (and I either owned or shot with the EPL1, EP2, EP3, and EPL3) or that I expect I'd be able to do with the EPL5 or EPM2. For dynamic scenes with movement, it doesn't really make much difference unless you're going for motion blur, but there's a difference. I couldn't quantify it precisely, but I've done things with confidence with the OMD that I never even bothered to do badly with other models. Like 1/2 second exposures handheld at waist level, with just the strap to add a bit of stabilization. Where the movement is blurred past the point of recognition but the background is sharp. In honestly, I haven't found myself using it for this type of application much (at least after the first few weeks). But it also comes in really handy by stabilizing the view in the EVF prior to the shot, which makes composing with a really long lens (say, 300mm, or 600 equivalent) SO MUCH easier than it ever was before. You can actually fix the focus box on your target without the view jumping around in the finder. And for video there's no comparison. The IBIS in the other models is basically non-existant for video and its nice to have a Pany lens with OIS. But with the OMD its remarkable how steady the handheld camera stays while shooting. I don't do much video, but the difference is really plain even for someone like me who doesn't.

So there are real advantages for some types of use, minimal if any benefits for other types of uses. And some of the greatest benefits have nothing to do with low light - once you can shoot at 3200 or 6400 there aren't many shots you'd have trouble with in low light even without IBIS - I do fine with my X-Pro and there's no stabilization on that.

-Ray

 

[Gary]

As a one mirrorless system type of guy, this thread makes me feel warm all over reenforcing I did good by choosing the OM-D. That OM-D just does so many things well in such a small package. Ray mentioned that with clean high ISO the relance on IBIS/IS is minimized ... and it is, but having clean high ISO and five axil IBIS is even better than clean high ISO alone. A great one-two punch which allows the photog many more choices in documenting/interpreting an image.

Gary

PS- I almost picked-up a Fuji a while back (to go with the Leica I almost picked-up), this thread is melting my resolve.
G