I briefly tested out the local pro-shop's D800 demo with my 50/1.2 AIS and found that it back-focused significantly using the AF confirmation dot which makes me wonder about the calibration of the AF system in some of these D800 bodies . . . .
I'll be going back on Sunday to give it another try with my 50/1.8G and perhaps a few other lenses.
I've ordered a D800, but will not likely get one until June or so
For general portraits and stationary subjects it's perfect. Studio or Landscape I never noticed a thing, it was only on sport, and with the subject coming towards which is a stern test that I noticed any issues.
I have a 50mm which is a touch out for general portrait which I'll have a look at but if they can all be sorted in cam it's a great function.
I've not got a longer lens as I shoot mainly weddings/portraits but another advantage of the D800 is you don't need such a long lens with the crops being such big files
I've not got a longer lens as I shoot mainly weddings/portraits but another advantage of the D800 is you don't need such a long lens with the crops being such big files
I'm looking forward to comparing the performance of the cam against my D3x and D3s.
I took out my D3 to do a back to back last week but to be honest it wasn't worth comparing such is the standard of the 800, it really is in a different league. Some have been testing the 800 against Hasselblad, something you couldn't do with the D3.
To be fair to the D3 it is four or five years since the sensor was designed so you wouldn't expect it to be close. I will still use my D3's for weddings though, file sizes are perfect!
I still cannot believe it's left up to us to test our own gear and calibrate it. Camera manufactures should profile both lenses and bodies writing a profile into the firmware of each so everything plays together as it should as well as being a whole lot smarter than an offset applied to every zoom setting and focus distance.
Mark_L wrote:
I still cannot believe it's left up to us to test our own gear and calibrate it. Camera manufactures should profile both lenses and bodies writing a profile into the firmware of each so everything plays together as it should as well as being a whole lot smarter than an offset applied to every zoom setting and focus distance.
If everyone would be willing to pay an extra $1k per item, that would be possible. (plus $$$ fees every few years for re calibration)
SloPhoto wrote:
If everyone would be willing to pay an extra $1k per item, that would be possible. (plus $$$ fees every few years for re calibration)
Personally, I am fine with the way things are.
Coming from a manufacturing background there is no way something as simple as doing an automated live view focus distance - phase detect af distance and sticking in the software would cost an extra $1k per unit. 36MP is only going to show up these inadequacies even more.
Mark_L wrote:
Coming from a manufacturing background there is no way something as simple as doing an automated live view focus distance - phase detect af distance and sticking in the software would cost an extra $1k per unit. 36MP is only going to show up these inadequacies even more.
Test fixtures, huge amount of data mapped to each lens, trained labor force, new software in camera to compensate properly. It would be quite expensive and quite prohibitive for lower end gear.
One would think that with the demands of the D800's high MP sensor that the PDAF system would both be more capable of higher accuracy AND be better calibrated right from the factory.
SloPhoto wrote:
Test fixtures, huge amount of data mapped to each lens, trained labor force, new software in camera to compensate properly. It would be quite expensive and quite prohibitive for lower end gear.
FYI, LV = contrast detect, not phase detect.
I know, hence why I said live view focus distance minus phase detect af distance to get any difference. It probably would be prohibitive for lower end gear but professional gear is not cheap by any stretch and even a simple test like this should be economic on pro bodies.
Mark_L wrote:
I know, hence why I said live view focus distance minus phase detect af distance to get any difference. It probably would be prohibitive for lower end gear but professional gear is not cheap by any stretch and even a simple test like this should be economic on pro bodies.
On my screen there was a linebreak at the '-' so I missed it. I was wondering as I have seen you reference it correctly in this type of discussion before.
jhinkey wrote:
One would think that with the demands of the D800's high MP sensor that the PDAF system would both be more capable of higher accuracy AND be better calibrated right from the factory.
PDAF as it stands (Nikon 1 notwithstanding) has no image sensor input. The ways I'm aware of to make it more accurate are to increase resolution or to increase overall size (thereby increasing minimum required aperture). And that would only improve accuracy for properly calibrated bodies.
Given the geometries involved, I can't see factory calibration improving (you'd need tolerancing to within microns, good luck with doing that cost effectively) without moving the PDAF arrays on sensor. Good thing Nikon made the V1 sensor
ausemmao wrote:
PDAF as it stands (Nikon 1 notwithstanding) has no image sensor input. The ways I'm aware of to make it more accurate are to increase resolution or to increase overall size (thereby increasing minimum required aperture). And that would only improve accuracy for properly calibrated bodies.
Given the geometries involved, I can't see factory calibration improving (you'd need tolerancing to within microns, good luck with doing that cost effectively) without moving the PDAF arrays on sensor. Good thing Nikon made the V1 sensor
You mis-understood. The higher MP of the D800 will show AF errors more readily than previous bodies. Therefore it's more important than ever to have a more accurate and better calibrated PDAF system (AF sensors, sub-mirror, etc.) than with previous 12MP or 24MP FX sensors.
Who knows how well the AF system of each body is inspected prior to shipping. It may be that Nikon just tries to manufacture and assemble to tight tolerances and never checks the AF system for accuracy, even on pro bodies. Maybe they do, but perhaps the band of what's acceptable is wide.
After some initial testing my 85 1.4D and my 50 1.4D are both off, this may be the camera that makes me upgrade to the newer G lenses.
I have a Lensalign, I did a little testing with the 85 and it is indeed back focusing. I'm going to seriously sit down with my lenses and test/adjust. I was doing it in my kitchen at high ISO, I need to do it in good light at 100...
My 70-200 VR1 is spot on, it's sharp as a razor!
The 36mp will definitely show when your focusing is off. I'm running this camera through all the paces before I do a job with it.
There's no way Nikon could deliver it perfectly focusing with every lens out there, it's not always the camera but the lens. My 24-70 is my worrisome lens because it's beat to hell and probably close to being trash if not spare parts...
Hi Bruce, after a couple of weeks I'd say the AF is very similar to the D3, I have some more to do and in fact with the fine tune you could argue that it has potential to be better. I've some more action pics to take this weekend and will post an update early next week.
I'm a fan of the fine tune, lenses will have a tolerance and mine was perfect on static subjects, only on moving subjects did it need adjusting. It only took a couple of minutes to sort the 70-200 (I did it at a cycling race I was watching) the screen is fantastic and you can see what is oof or not easily.
On my screen there was a linebreak at the '-' so I missed it. I was wondering as I have seen you reference it correctly in this type of discussion before.