And Canon shipped almost half the total interchangable lens cameras shipped in the latest figures.

By sales figures, I don't think Canon need to be concerned. Maybe this is simply brand inertia, but holding 50% of a market that supposedly has more than 3 competitors is pretty amazing. It's also worth noting that DSLRs used in live-view mode are, essentially, mirrorless cameras.

Nikon are in a very different position to Canon as Nikon are dependent on Sony for their sensors. Sony could put a pretty large dent in Nikon by simply cutting off the flow of sensors, as they are, supposedly, already doing for their "top of the line" silicon.

While I agree, I was making the point more in reference to the posters claiming that "You should use glass designed for your sensor, glass designed for a larger sensor will always perform worse".

In the context of a 2yr old breaking the printer, that typo is awesome!

Glass that is too good for the body? That is a difficult question.

Looking at the whole FF/APS-C question, is the pixel size of a Canon 5Ds that different to the pixel size of a 7DmkII? I would expect comparable pixel level performance at the centre of the frame for any given lens. How will that lens perform on a 5DmkIV? A 1DxII?

A lens designed for a 20MPix FF sensor attached to a 20MPix APS-C sensor will obviously perform worse than a lens designed for a 20MPix APS-C sensor. A lens designed for a 50MPix FF sensor, on the other hand, I would expect to perform about as well on a 20Mpix APS-C sensor. The obvious big penalty to using a lens designed for the 50MPix FF sensor is that it will be larger and heavier than the same focus length designed for a 20MPix APS-C sensor.

Now, within a sensor size, can you get lenses that are too good for a given sensor? I have a Canon 5D (the 12 MPix original), and I would expect it to fail to fully exploit the optical performance of any lens than can exploit the resolution of a 5Ds. Would that stop me acquiring such glass? No. The glass should still perform well. It would also perform well if I upgrade the camera body to current Canon body. If I upgrade the body first, I may discover that lenses which work fine on a 12MPix sensor can't cope with a 30MPix+ sensor.

I don't know if this has been mentioned, but... I have seen some cameras with firmware that will work perfectly with CF (I have some older Canons) cards up to 32GB (maybe larger, I don't have larger), but if you use the camera to format them the result is only 2GB of usable space. Format on the computer and the camera sees and uses the 32GB just perfect.

What makes this issue even more annoying is that if you are non-technical, getting windows to turn that card back into a 32GB card is slightly non-trivial.

What? 20fps isn't fast enough?!

Is that rating out of 100?

My reaction to the image posted is "another seascape sunrise/set, pretty colors, all nicely by theory, but it's still yet another seascape".

I'm not trying to be rude, it is a nice image, but it's really not standing out from the vast horde of nice seascape images we see.

Pelican 1500 or Pelican Air 1525.

Pricey, but indestructible.

Sounds nice.

What is the max flash sync speed?

What type of AF points? If they are cross points, that is really spectacular, but as they aren't saying so, we can assume they aren't. How many cross points?

And still inadequate. Taking the figures from the Sony site of 480 - 650 photos on a charge, that's between 4 minutes and 5.5 minutes of full burst rate capture. Except they claim the camera can't actually do continuous burst mode that long. Real usage figures for the battery life are needed before we can meaningfully comment on this.

What I notice is completely missing from your excitement is consideration of storage options. Where is the XQD card slot? CFast? Without very fast storage, their burst rate claims are interesting. If we assume they are to full-buffer, what is the buffer clear time from full? This is one area where the D5 and the 1DxII appear to have a clear and massive edge. Nikon "claim" 200 shots in burst mode for the D5, real testing showed they are limited by card size on XQD cards. What is real testing going to make of the burst performance of this Sony? I suspect it's going to be lacking simply due to the storage options.

A lot will depend on how the camera performs in real use. If it works as well as hyped, it will be an interesting development. However, there are parts of the spec that make me suspect that it's not going to be the effortless walk-over people expect. It will be interesting to see how Canon and Nikon respond, if they feel a need to respond.

As a first prime, the 50mm 1.8 options are very cheap, if it turns out you don't like using prime lenses. That would be my recommendation for a first prime.

The other thing that is worth bearing in mind when looking at prime glass is are you commited to staying on a DX body? DX specific lenses don't work well on FX bodies.

50mm is commonly seen as a boring uninteresting focal length, as far as I can tell. While it can try it's hand at everything, it's rarely the best focal length for anything: People tend to go longer for portraits, shorter for landscapes. Some people swear by the 35mm as their preference, others the 85mm (or even longer). The 50mm f/1.8 is commonly ignored by the "fast prime" fans, because the f/1.4 is faster - at three times the price, and a lot bulkier and heavier.

What the 50mm f/1.8 does have to it's advantage is it's the cheapest prime, and it's still a Canon prime, and the lightest lens, you are going to find for a Canon FF camera.

Personally I quite like it. I have a 50mm f/1.8 and it is the lens that lives on my 5D.

A cheap, light, Canon prime. What's not to like?

One lens I don't see mentioned is the 50mm F/1.8. Incredibly cheap, small, and light general purpose lens.

If your budget will cover it, the 11-24 would offer you wider angles. I'm not convinced the 16-35 is such a great investment due to the range overlap.

Looking at the long end the 100-400 is the obvious candidate.

I'm also going to be slightly contrary, and suggest getting a speedlite might be a better investment than glass. The 24-105 covers the most used lengths.

Which they will sell to Fuji, Hasselblad, and Phase One. Pentax as well, if they are looking for a new sensor.

Currently all the CMOS MF sensors are Sony, AIUI. It is unlikely Sony are going to rush to enter the MF market directly. The current arrangement means they make money of every MF camera sale, while carrying very little of the risk of serving a small niche market.

I'm not advocating a switch back to Nikon, I'm suggesting that if reach is the prime requirement, carrying both might be worth considering. 100-400 on the 5DmkIV, swapping to the 200-500 on D500. As, IIRC, the D500 has a 1.5crop factor, that would give the equivalent to 300-750 reach for when the 400 isn't quite long enough. At $3,400 this is a fair bit cheaper than a $5-10k lens, and offers a back-up body.

The Tamron and Sigma lenses would offer the reach for below $2k. No, they aren't the match of a Canon super-tele-prime, but at a small fraction of the cost, that might be an acceptable compromise.

The big question is what compromises are acceptable, and I think the Nikon option is a compromise worth including in the list of possible compromises.