The fact that the K5 and the D7000 are sharing (I think), the same Sony sensor it's not surprising that they perform quite similar.
According to your comparaison with the 5D2, on the video session I did, there are differences. Visual differences. The 5D is a FF sensor and features 5MP more than the K5 and 9MP more that the camera involved. Differences are cristal clear in low-light and also the D.O.F you can obtain is more narrow with the 5D. In post-prod you would be much more limited with the K5 or the D7000. The D7000 seems to be an extremely good overall product. The 5D2 for stills has certainly much more to give for the money. We are talking about a 2000euros camera.
But if you already have Pentaxes lenses, the K5 is probably a very good choice. If not, I would probably stick in CaniKon
Fred,
I don't have any Pentax lenses. The Pentax Spotmatic was the first SLR I ever owned about 40 odd years ago, and I remember being thrilled by the through-the-lens experience and built-in light meter. It was a major technological breakthrough for me at the time.
The fact that you get the impression that the 5D2 video has better low-light performance than the D7000 is interesting. It conflicts with the DXO results, except in circumstance where you are using the maximum aperture of the lens.
For example, if I were using the Canon 24-15/F4 on the 5D2, and the Nikkor 24-120/F4 on the D7000, shooting the same dark scene at F4, but at different focal lengths to equalize FoV, then it's true the 5D2, at F4, would have a shallower DoF, and because of its slightly better 'low light' performance due to its larger sensor, should exhibit slightly less noise.
However, I question whether extremely shallow DoF is a desirable thing, especially when the subject is moving and the camera doesn't have auto-focussing in video mode, which I understand the 5D2 doesn't have.
I have enough trouble getting still scenes exactly in focus at wide apertures. Moving scenes could be a nightmare at F2.8 on full frame.
It's often been said that the only impartial, fair and sensible way to compare camera performance is at equal shutter speed and equal DoF. Do you not agree with this principle?
If your speciality is taking shots at F1.2 with FF DSLRs, then fair enough. There are no F0.7 lenses available for the cropped format to match that shallow DoF.
I suppose if you want to be really, really creative and get the tip of the subject's nose in focus, but the eyes out of focus (or vice versa), then the DoF advantages of FF are for you. No argument there.
But I suspect I will prefer to shoot video at less extreme apertures. I might prefer the 'less shallow' and less problematical DoF of F4 on a D7000 as opposed to F4 on the 5D2.
I know this is a lot of blah! blah! technical nonsense for you, Fred, but I really think you would benefit by understanding this issue.
On the DXOMark site, the 5D2 gets a higher ranking than the D7000 for low-light ISO (sports subjects).
This ranking clearly ignores the fair and impartial principle of 'equal shutter speed and equal ISO'. It assumes that for low light shooting you will automatically use the same aperture at the same ISO, irrespective of DoF consequences.
What happens if we equalize DoF and shutter speed by choosing a wider aperture for the D7000?
Well, we have to change ISO of course. The relationships become something like the following.
F4 at ISO 3200 on the D7000 produces the same DoF at the same shutter speed as F5.6 on the 5D2 at ISO 6400, assuming the ISO sensitivities are the same.
However, ISO sensitivities are not the same. The D7000 ISOs are more accurate. At any given 'manufacturer-nominated' ISO, the 5D2's real ISO is about 1/3rd of a stop lower than the 'real' D7000 ISO, which is also a bit on the low side. Example, ISO 3200 on the D7000 is actually ISO 2627, whereas ISO 3200 on the 5D2 is lower at ISO 2133.
Now I understand, Fred, at this point you are going out of your mind with boredom. That delicious chicken and pasta dinner with mushroom sauce you are planning is far more important.
But bear with me. The difference in DoF between the D7000 and the 5D2 is slightly more than a full stop, so, for equal ISO and equal shutter speed we should be comparing the the 5D2 at ISO 6400 with the D7000 at (say) ISO 4800, but this doesn't give us equal DoF. (Take an asprin at this point).
Because of the discrepancy in accuracy of ISO, between these two cameras, the differences in the ISO settings needs to be greater than one full stop, for equal ISO and equal shutter speed.
At such settings, the 5D2 would appear to have no image quality advantage, according to DXOmark results (except perhaps resolution according to lens quality and optimim aperture chosen). In fact, the D7000 is slightly better, but only at an extreme pixel-peeping level, so we can ignore that.
It's a pity that DXO do not test video quality. Trying to find any worthwhile comparisons between the D7000, 5D2 and 7D video output on the internet, is a nightmare. What a waste of time! Most video's I've seem so far are total crap, and none show a direct and meaningful comparison between the main contenders.
Perhaps after a couple of glasses (or more) of Jamieson's Run Chardonnay, it's me who is confused. If so, please set me straight.
http://www.dxomark.com/index.php/eng/Camera-Sensor/Compare-sensors/(appareil1)/680%7C0/(appareil2)/483%7C0/(onglet)/0/(brand)/Nikon/(brand2)/Canon