E-5 better in low light than D7000? Yes, maybe.

[ptomsu]

Well - the first firmware update improved things (did you have that?), and it seems there is another one on the way - but apart from testing (where I can see issues as well) I've found the AF to be very reliable in 'real life' even in low light . . . . it's a new AF module (also designed for the MF 645), and Pentax do have a history of getting these things right.



I disagree so radically with this - I have / want / need a full frame 35mm camera, it has obvious advantages. But if I'm going to have a decent small dSLR, then it can't be full frame, because of the lenses - even if Nikon can squeeze a FF sensor into a D7000 sized body (as you say, possible I'm sure), there really isn't much point if you have to use lenses like the 24-70 and 70-200 to take advantage of the sensor.

I think the time it's taken Nikon to update the D700, and Canon the 5DMk II indicates that camera manufacturers on the whole feel the same way.

Added to which the high ISO from the new Sony APS-c sensor is great - really good - Jorgen's argument in favour of the E5 is quite comprehensible, but it hinges around the existence of fast (and excellent) Olympus glass (and the fact that there's nobody selling the K5 on a professional basis where he lives :loco

I think the tragedy of the E5 is that Olympus have obviously done the very best they can with the AF and ergonomics, and their glass is peerless . . . . . . but they're stuck with a mediocre Panasonic sensor (even if they've done their very best with it), and they've made a camera twice as big as it needs to be - which defeats the whole original concept of 4/3. If the E5 was the size of the Pentax, with a sensor the quality of the Sony sensor in the K5, then I'd be there in a second . . . .

First, I had the FW update of the K5, but even with that it did not work for me.

Second - as you can see from all what I am trying, I really also want a smaller DSLR with maybe APSC sensor size and excellent but smaller lenses. Unfortunately I did not find this so far. It is definitely not Canon (as I rule this company out because I simply do not like their approach), it is not Nikon, as they might have the right camera (D7000) but not the right lenses, it is not Pentax as I had the issues we discussed in length and - my last hope - it is also not Olympus with their E5, because the sensor is not the GH2 sensor (this would have totally solved the sensor issue) and they made the mistake to build these huge Pro Grade lenses and a Pro Grade camera body like the E5.

I fully agree, if there would be an E5 in an E1 body with the GH2 sensor and similar excellent image processing as in the E5 plus the same Pro grade lenses like the 2/14-35 and 35-100 only in 2.8, then this would be the dream combination!

I do not agree that the K5 AF was near as good as the AF in the E%. This one is now significantly better! But to put that in relation with what is possible, the D700 AF simply tops that by far, especially in low light and in heavy backlight.

 

[jonoslack]

I do not agree that the K5 AF was near as good as the AF in the E%. This one is now significantly better! But to put that in relation with what is possible, the D700 AF simply tops that by far, especially in low light and in heavy backlight.

Well, Pentax have acknowledged an issue, and said that they'll fix it (relating explicitly to low light). On the other hand, maybe you had a faulty camera? I'm finding the AF to be pretty good (even CAF with fast subjects is pretty good).

Not for a minute suggesting you should go back to Pentax, but if the only problem was the AF, then I'd suggest you wait and see what they come up with. Remember, this is a completely new AF system - it usually takes a little while to come up with the goods. It fits the bill in so many other ways . . . . mind you, so does the Nikon 7000 . . . . and you already have the glass (still not quite sure why you're against it).

Comparing it with the D700 is fine - but not if you want a small camera!

Of course, (relating to other posts), it is possible to have a nice small full frame camera, with the best possible lenses (AF isn't so good though) . . .. . remember the M9? :ROTFL:

 

[Jorgen Udvang]

Sunday today, so I felt inclined to waste some time doing screen captures from dpr. Here are files processed from RAW. I omitted the K-5. It would have been be the winner anyway, and I don't need to see any test shots to prove that. No full frame here, since the D3s only comes up with ISO200. Cameras with IBIS at ISO1600, those without at 3200. I could have given the E-5 an extra stop for the f/2.0 lenses, but chose to be conservative:

 

[Jorgen Udvang]

Then the out-of-camera jpegs. I changed the G2 for the D3s here.

Here, I have reduced the D7000 to ISO1600:

And here I have included the K-5 as well:

 

[Jorgen Udvang]

The lenses used obviously comes into play here as well. In that area, it's difficult to beat Olympus (the 50/2.0), and that is probably also part of the reason for the E-5's superb rendering of detail. But there's another side of that question as well: While Olympus makes three superb, professional quality standard zooms (14-54/2.8-3.5, 12-60/2.8-4.0 and 14-35/2.0), Nikon only makes one, the 17-55/2.8, which is a beast, Pentax also makes one, the 16-50/2.8, which seems to be the weakest part of their pro line-up while Sony has none for the crop sensor. Although I've never fallen in love with standard zooms, they are very important for bread-and-butter work. I prefer them to be small, lightweight and sharp.

One can obviously buy third party lenses, but my experiences with the Tamron are going downhill.

 

[Jorgen Udvang]

I became curious, so here are some that include the K-5 and the 7D as well:

And since I do shoot in daylight now and then as well, here are two at ISO100:

Please be aware that they seem to have changed the watch. The A55, 7D and D3s shows the old one with a black line in the hour markings and different hands that gives more shadow contrast.

Another interesting fact: The RAW files of the E-5 are around half the size of the files from the other cameras. With double backups, that's a saving of $3-500 per year just in hard-disks.

 

[Paratom]

I have made this thought many times...smaller sensor means more DOF means you shoot wider open for same DOF so you can use one stop lower ISO.
This works as long as we talk about zooms because Oly offers some nice and (faster) zooms than available for ff.
As soon as you come to primes its a different story and the selection for 4/3 gets pretty much limited. No real fast portrait lens for example, and for same speed and DOF flexibility like a 50/1.4 you would need a 25/1.0 lens.

The next thing is the question what do you gain? The f2.0 zooms are not really smaller nor lighter than equivalent the f2.8 ff zooms.

On the other side, due to sensor size you have high pixel density which will limit the 4/3 sensor to less MP than what would be possible on a ff frame sensor (if the glass resolution is comparable).

So for me it wouldnt be that much about 1 or 2 steps difference in high ISO quality, but the question would be if 12MP is enough for the prit sizes intended and also if the lenses one wants are offered.

If one is a zoom guy-Oly seems to offer very nice lenses, for prime guys the selection gets limited IMO.

 

[Jorgen Udvang]

I have found that 12MP is enough for all I do. Another side of this is that there are few lenses that can resolve more than that on a crop camera. When looking at the low ISO samples from dpr, the 12MP E-3 shows considerably more detail than the 18MP 7D. That is probably a combination of lens and filtering, but there isn't really much point in an 18MP camera if the lenses can't render that much.

This is of course what Olympus has been pondering all the time, without getting much response, but it seems to be true: 12MP and top notch lenses will render more detail than 18MP and lenses that were designed for full frame.

When it comes to DOF, it's very much a question of taste. I mostly shoot portraits at f/2.8-4.0 on full frame. That means that the Sigma 50mm f/1.4 will do the job very nicely when used on a 4/3 camera.

 

[Paratom]

I have found that 12MP is enough for all I do. Another side of this is that there are few lenses that can resolve more than that on a crop camera. When looking at the low ISO samples from dpr, the 12MP E-3 shows considerably more detail than the 18MP 7D. That is probably a combination of lens and filtering, but there isn't really much point in an 18MP camera if the lenses can't render that much.

This is of course what Olympus has been pondering all the time, without getting much response, but it seems to be true: 12MP and top notch lenses will render more detail than 18MP and lenses that were designed for full frame.

When it comes to DOF, it's very much a question of taste. I mostly shoot portraits at f/2.8-4.0 on full frame. That means that the Sigma 50mm f/1.4 will do the job very nicely when used on a 4/3 camera.

Jürgen,
thats exactly what I mean, the smaller the sensor the earlier max. MP will be limited by the resolution of lenses (assuming we are talking about glass with comparable resolution).
I have to agree so that one can do a lot with a good 12MP image.

 

[marlof]

Please be aware that they seem to have changed the watch. The A55, 7D and D3s shows the old one with a black line in the hour markings and different hands that gives more shadow contrast.

I think it's the same, but the light causes some shadowplay on those images. In the other images, without the black lines, you can also see difference in detail. The hour markings and the hands are probably sharply curved. When they're lit straight on, or harshly, the shimmering silver color doesn't show much detail. In other light, you can see where the shadow is falling. Or maybe some sensors deal better with highlight detail, I don't know.

 

[ptomsu]

@ Jorgen

the last samples you copied including the K5 just show very nicely what were also my findings with the K5. I found it at higher ISO having less noise (ISO 1600 and higher) but also less details. The E5 in my opinion - even showing already substantial noise, has much better detail The 7D for me comes on place 3 and the D7000 is the clear looser.

I think this all relates very well to the quality of lenses. And WRT lenses the Olympus is the clear winner, while the Pentax and Canon are somewhere in the middle and the Nikon with the lens tested here is the obvious looser!

 

[Jorgen Udvang]

I think it's the same, but the light causes some shadowplay on those images. In the other images, without the black lines, you can also see difference in detail. The hour markings and the hands are probably sharply curved. When they're lit straight on, or harshly, the shimmering silver color doesn't show much detail. In other light, you can see where the shadow is falling. Or maybe some sensors deal better with highlight detail, I don't know.

Not the same. You see it clearly if you compare the ISO100 shots by the 7D and the E-5. The E-5 shows more detail everywhere, still the black lines are gone. You can also see that the "counterweight" on the seconds hand has a different shape.

The glass also has a different shape. It's more rounded in the new version.

 

[ptomsu]

I had another day of shooting the E5. Today I limited a lot of my shots to ISO200 and Vivid setting. In combination with the 35-100 and 12-60 the results are stunning! Especially when you start using selective light metering, manual focus fine adjust etc. Once setup that way, the E5 performs flawlessly in good to medium light (even darker forest).

I also tested again the noise from 200 - 6400. I would say till ISO 800 there is no big issue with noise, if exposed correctly! At 1600 and even 3200 noise is clearly there, but I could live with it. Even up till ISO 5000 noise is quite acceptable. But at 6400 the results are kind of useless - at least for me. Maybe a future FW update might bring improvements here.

After these 4 days of shooting and even taking into account the higher noise issues I must say I do not like to return the E5 and especially these 2 lenses. While I can clearly proof that the D700 performs better under low light, I must say I am still considering if I should not switch to the E5. Because my main area of use would be in the range between 200 - 1600 and given the fast zooms (2.0) there can be great results achieved up till ISO 3200 comparable from a D700 - at least for me. BUT with lower ISO settings the E5 clearly outperforms the D700 WRT IQ in each and every area.

So hard decision still. If I only would know what Olympus is going to bring as next Pro DSLR????? The lenses are stellar without any doubt!

 

[bcf]

The only area wher I see clearly more detail from the E-5 is in the Martini logo on the bottle ("suppliers of Martini vermouth"). Elsewhere, it's either a wash or in favour of the D7000. Given that I am comparing the D7000 at ISO 1600 with the E-5 at 800, and that the E-5 is bigger than the D7000, as big as a D700, I fail to see the advantage...

If the E-5 was clearly smaller, I could be interested. But at this size, I fail to see the point. And the f/2 4/3 zooms are as big as their full-frame f/2.8 Nikon counterparts.

 

[jonoslack]

the last samples you copied including the K5 just show very nicely what were also my findings with the K5. I found it at higher ISO having less noise (ISO 1600 and higher) but also less details.

But you only had the kit lens with the K5 Peter, and we all agree that's a dog - that 100 ISO picture looks oof to me (maybe that bears out your focusing problem!).:ROTFL:

all the best

 

[jonoslack]

The only area wher I see clearly more detail from the E-5 is in the Martini logo on the bottle ("suppliers of Martini vermouth"). Elsewhere, it's either a wash or in favour of the D7000. Given that I am comparing the D7000 at ISO 1600 with the E-5 at 800, and that the E-5 is bigger than the D7000, as big as a D700, I fail to see the advantage...

If the E-5 was clearly smaller, I could be interested. But at this size, I fail to see the point. And the f/2 4/3 zooms are as big as their full-frame f/2.8 Nikon counterparts.

HI Bernard - you hit it on the head for me. The E5 is lovely, and clearly really well thought out - but the trouble with those crops is that, of course, the point of focus is slightly different for each camera for each shot - trying to compare critical amounts of detail seems dangerous.

But the real issue is that the E5 with an F2 zoom is as big as a D700 with an f2.8 zoom. . . . . in fact, it's also pretty much the same size as the Sony A900 with the Zeiss 24-75 . . . . . . .

 

[jonoslack]

This is of course what Olympus has been pondering all the time, without getting much response, but it seems to be true: 12MP and top notch lenses will render more detail than 18MP and lenses that were designed for full frame.

HI Jorgen
I think it is true . . . . but it isn't the whole truth. Added to which, if you have a noisy 18mp image you can downsize it to a much less noisy 12mp image (without losing much detail). This is something I discovered when printing from a D700 and a Sony A900 - that the obvious difference in noise at 100% pixel peeping disappeared if you downsized the Sony file.

Basically, you can do your own 'pixel binning' to create a better file at lower resolution . . . if you start off with more resolution.

Mind you, if you want the detail, then just take away the AA filter!

 

[Jorgen Udvang]

The only area wher I see clearly more detail from the E-5 is in the Martini logo on the bottle ("suppliers of Martini vermouth"). Elsewhere, it's either a wash or in favour of the D7000. Given that I am comparing the D7000 at ISO 1600 with the E-5 at 800, and that the E-5 is bigger than the D7000, as big as a D700, I fail to see the advantage...

If the E-5 was clearly smaller, I could be interested. But at this size, I fail to see the point. And the f/2 4/3 zooms are as big as their full-frame f/2.8 Nikon counterparts.

At high ISO, the E-5 doesn't show more detail when there's only one stop difference, but if you mount the 17-55 on the D7000 and the 14-35 on the E-5, the E-5 has a lens that is one stop faster. In addition, the E-5 has IBIS, which will gives another 1-3 stops. Then we are talking ISO 800 or even 400 compared to 3200. At ISO100, the differences are so dramatic that there isn't really any competition. The difference is probably partly in the lens, but since cameras work badly without lenses, it's still a valid comparison.

The most interesting part is still that the difference in MP count seems to evaporate completely, which is food for thought.

I agree that the fast standard zooms don't give 4/3 any size advantage, but mount a telephoto lens, like a 70-300 or a Bigma, and you have 30% more reach. Another good example would be the Zuiko 150/2 vs. the Nikkor 200/2. While the cameras would have exactly the same reach (300mm eqv.), the Zuiko is smaller, lighter and much cheaper. You could choose to compare with the Nikkor 180/2.8 instead, but then you lose one stop and VR and get less reach.

 

[Jorgen Udvang]

HI Bernard - you hit it on the head for me. The E5 is lovely, and clearly really well thought out - but the trouble with those crops is that, of course, the point of focus is slightly different for each camera for each shot - trying to compare critical amounts of detail seems dangerous.

But the real issue is that the E5 with an F2 zoom is as big as a D700 with an f2.8 zoom. . . . . in fact, it's also pretty much the same size as the Sony A900 with the Zeiss 24-75 . . . . . . .

Jono,
The E-5 is stopped down to f/6.3 and the cameras wioth APS-C sensor to f/9.0 on 50mm lenses. I would be very surprised if they missed focus with that much DOF. I don't particularly like dpr, but they are very meticulous when it comes to details like that. They know that they'll get a lot of pepper if they aren't.

 

[bcf]

Jono,
The E-5 is stopped down to f/6.3 and the cameras wioth APS-C sensor to f/9.0 on 50mm lenses. I would be very surprised if they missed focus with that much DOF. I don't particularly like dpr, but they are very meticulous when it comes to details like that. They know that they'll get a lot of pepper if they aren't.

I find it troubling that for the Martini logo, the E-5 seems to be sharper, but is not in the rest of the photo (I think). The difference is particularly striking for the paper clips - is the plane of focus different?