NikoDoby said:
Surprised that no one's really talking about this new 2,016 pixel RGB sensor . . . If this is what Nikon puts in their "prosumer" camera imagine the next generation metering on the DX00 and D4!

Yes, I was rather impressed with the sheer pixel-count of the new RGB sensor as well. On my D90, I usually shoot in spot-meter mode about 90% of the time, and only 10% of the time in matrix mode, but a smarter, 2,016-count sensor could possibly change that ratio. I think NIkon raised the ante pretty high for the D7000's feature set, and that also makes me really excited to see what's next for Nikon's FX bodies in 2011.

It's not even close to comparable to my D60 - I don't have a frame of reference for any other camera like a 90/300/700, etc...I still typically shoot in center weighted, but last weekend I was unable to get a shot of my daughter in the shade on a see saw and direct light in the scene behind her with center weighted and had to switch to matrix which got the shot. It also worked better than spot.

So just wondering, has anyone tested how accurately the D7000 meters?

Super Shooter said:
What about the iTTL flash control? That should be a big improvement too with this new meter. I wonder if Nikon is just going to put this same meter in all of it's pro cameras too. Maybe this is a pro meter?

I guess nobody cares about iTTL flash control then.

A very noob question here, does the point and shoots have so much better metering since they use the original sensor (10 ~ 18 mp) for the metering, or do they not? I have never heard a pns having a second sensor for metering, and if they do use the main sensor, they are bound to have metering thats way way beyond 1000 ~ 2000 px dedicated sensors?

Oh btw, a bit of introduction, I have been trolling these threads for the past 2 months and not signing up out of lazyness :D Finally though, I'm here so a barrage of questions for you guys lining up :D
I have a IXUS 95 IS pns, long for a D7000 and am 3-4 months old in the photography game.

Jack the Ripper said:
its allright paperman, use the center to focus on you want, lock the focus, then recompose.

:-) Not that easy with a tripod ... I'd rather focus manually.

jonnyapple said:
So you never answered my question about whether or not you're a scientist.

I'm a retired aircraft mechanic.

its allright paperman, use the center to focus on you want, lock the focus, then recompose.

Thanks Jonnyapple , for the effort at least . I'm not sure if Drab's note means it technically is possible or the exact opposite .

In any case , it looks like I will never have a focus point at the bottom of my screen . The designers are probably assuming that anyone shooting landscape on a tripod must be using manual focus.

Hahaha! My old optics professor would probably murder me. The missing info is in my next paragraph where I said there are prisms between the sensor and the (desired) focal plane. I understand image formation with a lens even if my understanding of rangefinders is a little sketchy.

So you never answered my question about whether or not you're a scientist. How do you understand/explain phase detected autofocus? I guess no matter how it works we've come a long way from having just one focus point (even the split-screen viewfinders for manual focus are based on the same principle).

jonnyapple said:In order to get the parallax measurement, light from both sides of the lens has to get to the point you want to have a focus sensor,

jonnyapple - you're not assuming what I think you're assuming, are you?
http://amasci.com/miscon/lens1.html

I'm not absolutely sure, Paperman, but I think it comes down to physics why they're close to the center. In order to get the parallax measurement, light from both sides of the lens has to get to the point you want to have a focus sensor, and on the edges of the image you don't get light from both sides of the lens, or at least not with some lenses.

This is at best over-simplified and possibly wrong, but it's the way I think about it. I only dared answer because you asked if anyone had an 'opinion' about it. ;-) I think technically you actually have to have light from both sides of the lens hitting prisms just before the sensor, each of which selectively directs light coming from one specific part of the lens to the sensor, but I think it works out to be similar to what I described before.

So to answer your question, I'm pretty sure they could have AF points further from the center for some lenses with a slightly larger image circle (PC lenses, for example). They should also definitely be able to put them as close to the bottom edge of the frame as they can to the side edge of the frame. I'll bet someone has done a cost-benefit analysis and it's not worth the extra design and production dollars because not everyone will use them or will even have lenses that would let them use them.

I know we're talking about metering and not AF , but I'd like to know if anyone has an opinion why the focus points are only in the center elliptical area in DSLRs and still why not spread all over.

I shoot landscapes all the time and usually have to focus almost to a point down the bottom of the frame - to get best overall focus ( that is where hyperfocal distance occurs usually ) and of course it is not possible .

I know it will also effect the metering points and add costs but still wonder why it can't be done on a $1000-5000 camera .

What about the iTTL flash control? That should be a big improvement too with this new meter. I wonder if Nikon is just going to put this same meter in all of it's pro cameras too. Maybe this is a pro meter?

@Johnnyapple Thanks for the double Rainbow...

Now if I could what he had...

I do suspect that the metering will be more responsive (at least I should think Nikon is banking on it), so I hope they don't let us down.

Mike

TaoTeJared said:
Fairly cheap handy-cams seem to do the "idiot mode" much better than most DSLRs.

This is true.

I like the way that sounds. Thank you for the good explanation. You guys need to write manuals. :)

Never paid too much attention to the AF to Meter connection details - all faster and more focused than my inept eyes so I let it do its thing.

I really could care less about video. I have no need for it and I have almost no interest in it at this point. Fairly cheap handy-cams seem to do the "idiot mode" much better than most DSLRs.

TTJ, ditto what Drab said, and I don't understand what kind of phase focus you're expecting in video. You won't see that without some kind of beam splitter in front of the sensor.

Drab said:
As johnny and I said, the meter has a TON to do with subject tracking AF through Nikon's Scene Recognition System.*

AF points are "dumb" in that they just know if the split-phase lines match or not. No usable "resolution" for subject recognition, no color recognition. This is where the meter sensor comes in. Nikon uses the meter segments, as a low-res "video feed".

Nikon maps the AF active sensor to a "frame" from the meter and then use the next "frame" from the meter to tell which AF sensor to make active next by seeing where the object which /was/ under the AF sensor is now.

The higher res the metering sensor, the harder it will be to confuse the subject recognition routines.

THIS is why I'm excited about the new metering. As I said, I never had an issue with any of the matrix meters, but I'd love even better subject tracking. And if the D700 replacement gets better tracking through new meter tech AND AF points where I actually want them - woo hoo!

Exactly- which is also why im stoked about the next gen fx bodies...hoping for more frame coverage and obviously larger MM coverage...

In terms of metering I am pretty old school. I do use matrix some times when I dont wantto think but most times its on centre weighted. Wil have to try out the new metering when I get the D7000 but I dont think it will have a big impact on the way I shoot.

TaoTeJared said:
AF has almost nothing to do with with the meter (except phase focus for video maybe). We shouldn't get the two mixed up.

As johnny and I said, the meter has a TON to do with subject tracking AF through Nikon's Scene Recognition System.*

AF points are "dumb" in that they just know if the split-phase lines match or not. No usable "resolution" for subject recognition, no color recognition. This is where the meter sensor comes in. Nikon uses the meter segments, as a low-res "video feed".

Nikon maps the AF active sensor to a "frame" from the meter and then use the next "frame" from the meter to tell which AF sensor to make active next by seeing where the object which /was/ under the AF sensor is now.

The higher res the metering sensor, the harder it will be to confuse the subject recognition routines.

THIS is why I'm excited about the new metering. As I said, I never had an issue with any of the matrix meters, but I'd love even better subject tracking. And if the D700 replacement gets better tracking through new meter tech AND AF points where I actually want them - woo hoo!

It sounds good to me. I shoot center or spot most of the time, but am amazed by the current matrix when I do use it.

I just wonder if it has more to do with video or in essence just spiting the 1005's into two. (Not really but you get the jest.)

AF has almost nothing to do with with the meter (except phase focus for video maybe). We shouldn't get the two mixed up.

Abhinav said:
johnny : you've poor mans Leica .thats a great cam for sure .I hope it is released in India .I will buy one to replace my pns :)

So ,coming to the point of uber metering sony nex :) how accurate it is ? when it comes to landscapes ,night ?

I don't know if I'd call it a leica...

I use manual exposure (or shutter priority when I've got a nikon lens on it because you have to change the aperture manually with the adapter), so I can only say that I really like the live metering histogram that you can have it show on the LCD and it's as accurate as your skill in reading a histogram allows. You have to keep in mind that the histogram shows post-jpeg curved data and you're storing raw, so I like to live on the edge of overexposure. Live view on the Nikons would be so much more useful with that. Does anyone know if there's a Nikon body with that feature?

Drab said:
From the sounds of it, it will be used as much for the 3D AF tracking (color-awareness @ AF point) as for metering. Ignoring the D80 debacle, I've not seen much of a difference in metering accuracy between any of the bodies since Matrix came out.

EDIT: And if doubling the number of metering points is really the bee's knees then the D300 should have been a no-brainer over the D90. ;)

Maybe, the D300's meter can be a bit defective at times. For all the ways the D300 improved over the D200, IMO the D200 still meters better. It's perhaps one of the best at metering. Of course, I'm speaking from anecdotal personal experience.

Considering that I noticed a big improvement in matrix metering between the D90 and D300, I'm going to guess that the new system will be a very good. I sure hope it helps to improves the 3D tracking, because right now I don't find it useful, even with the 51 points of the D300.