Muchas gracias por poner estas fotos y por las explicaciones que añades en ellas.
Es muy posible que, conociéndote, Nikon haya cursado una directiva secreta para que a ti siempre te vendan las cámaras mejor calibradas, y al resto del planeta las demás. No lo sé a ciencia cierta, pero no me extrañaría .
Porque lo cierto es que la D90 que tengo cuando quiero hacer fotos a sujetos bajo luz de sol directa usando medición matricial tiene tendencia a quemar algunas zonas, cosa que me pasa mucho en fotos de flores, por ejemplo, si no compenso a la baja de -0.7 EV hasta -1.7 EV.
También me ha pasado haciendo fotos a rostros de personas a la luz del sol, generalmente aparece quemada una parte de la frente si no compenso a la baja (naturalmente en esos casos es mucho mejor usar un difusor sostenido por un ayudante o el propio modelo y lo hago siempre que puedo, pero eso es otra historia).
Pero fíjate que en ningún momento he dicho que eso sea un fallo de la cámara o que haga "cosas raras". Al contrario, te he dicho que tampoco pasa nada con que una cámara sobreexponga, mientras sepas cuanto y cuando lo hace.
El sistema matricial de la D90 tiene 420 "segmentos", me parece lo más normal del mundo que se puedan escapar zonas de luces altas. A la D800 que tú tienes, con 91.000 píxeles en matricial, se le escaparán muchas menos, eso aparte de que a ti te habrán dado una de las buenas .
Pero parece que la conspiración de Nikon es a escala planetaria, sólo a ti te reservan las "buenas", porque mira lo que les pasa a los demás:
A Richard Butler y Simon Joinson revisores de DPReview:
"Our only real worry about the D90 is the matrix metering, which seems to be so strongly connected to the selected AF point that it allows highlights to clip a bit too often for our liking. There is an option to fine-tune the meter (and assign a different amount of correction to each metering mode), if you find it a consistent problem."
A Ken Rockwell:
"Exposure MeteringA un revisor de la D90 en Flickr:
The meter seems identical to the D300 and D3. The defective meter of the D40 and D80, which often overexposed, is gone.
On the D90, D300, D700 and D3, I usually shoot at 0.0 exposure compensation, or often at -0.7 compensation outdoors if I'm shooting in VIVID picture control with +3 saturation. Forget me, just look at your LCD, and change the compensation (the +/- button near the shutter) for your next shot if your pictures are too dark or too light.
A Jim Keenan, de Digital Camera Review
"The D90 seemed a bit more prone to lose or at least push the envelope on losing highlights in difficult, high contrast situations than the others – perhaps the 1005 pixel system has an edge over the 420 in this regard. There are center-weighted and spot metering options in addition to the 3D Matrix."
"For this preliminary test on Saturday I didn't change any of the D90's other default settings. Other settings pertaining to photo output and Picture Control were left at their default settings.
1. I used the default 3D-tracking (11 points). This is a new feature that I'm completely unfamiliar with because it's a new wrinkle compared to my D80. I'm always skeptical of these advanced features that make the camera so smart that it's trying to second-guess me, But in my initial tests I was quite impressed.
2. Wanting to evaluate the D90's matrix metering compared to the D80, I used matrix metering for all my test shots on Saturday. I spent about 4-6 months with my D80 trying to get matrix metering to work for me and couldn't do it. The results not only tended toward overexposure of contrasty scenes by 1/3 to at least 2/3 stop with blown highlights, but the results were also frustratingly inconsistent. I finally settled on using a combination of center-weighted and spot metering instead.
Thom Hogan, in his excellent D80 manual, describes exactly why this happens. I understand the theory too, but the practical results I got using matrix metering with my D80 were so consistently inconsistent that after almost six months of trying to tame the D80's matrix metering, I finally began using a combination of center-weighted and spot metering nearly all the time.
Unfortunately, the D80 and D90 share the same 3-D matrix metering hardware and firmware. So my dissatisfaction with the D90's matrix metering's propensity for overexposure of contrasty scenes carries over to the D90. However, that's not to say that matrix metering isn't very useful, and very accurate in many situations that I'll demonstrate below.
As DSLR metering methods and algorithms have become so much more sophisticated and complex in recent years, I've come to recognize that there's no "one size fits all" ideal metering mode. Matrix metering will produce excellent results under certain conditions, but center-weighted or spot metering will do a better job in others. It's all a matter of personal choice, knowing how each metering mode works, and being able to choose the right one for the photo you're framing."
A un miembro del foro de Nikonians, USA:
"d90 overexposure issue
One thing that I frequently read everewhere else is about D90's problem with overexposed pictures when in Matrix Metering.
I don't see that much in this forum.
My d90 does overexpose most of the time under bright conditions.
Did someone find a solution for that?
If I decide to go center weighted, wouldn't it be even worse, since no information about the background would be considered at all?"
A varios del foro de DPReview:
I'm getting quite a few overexposed pictures with my new D90.
Settings are standard, Dlight is off, matrix metering, AP mode,
central point focusing. Exposure is at times 2/3 of a stop too high.
I don't think I'm doing anything wrong, and besides I checked it
carefully many times. Since I'm new to Nikon, I wonder if this is a
well known problem.
Thanks for your help,
I usually have mine set to -0.3 or -0.7, at least if shooting jpegs.
A gente del foro de Velocity Reviews:
> "Focus" <> wrote in message
>> I was looking over the reactions as some folks seem to think that it's
>> OK for Nikon to over expose in MM. But is it really?
>> Here what they promise in their advertisement about the D90:
>> "Nikon 3D Color Matrix Metering II with Scene Recognition System:
>> Nikon's renowned 420-pixel RGB 3D Color Matrix Metering II, teamed
>> with the exclusive Scene Recognition System, evaluates images,
>> referencing an on-board database of over 30,000 photographic scenes,
>> for unmatched exposure accuracy."
>> A database of 30.000 photos? None of them had a clear, sunny sky in them?
>> That hardly sounds like a camera that would blow out skies like a P&S
>> shooter, does it?
>> So, Nikon: explain yourself.
> I don't know what you are doing wrong Focus but both my D90s are spot on
> with exposure.
When I was selling cameras, it was a very common scenario
that people would bring back Nikon cameras (D40, 50, 70, 80,
90) claiming the exposure was faulty because they were
getting white skies.
There were 2 ways to correct it and get exposures that one
would consider normal - use centre-weighted average, and
take a reading with the horizon exactly in the middle of the
frame, or on matrix use minus 1 to minus 2 EC.
The matrix metering put far too much emphasis on the land
part of a landscape, and would blow the sky every time. In
fact I would call the land part over-exposed too - medium
greens became insipid yellow greens etc. As you say, the
database of 30,000 images obviously didn't include a sunny
To be fair to Nikon, my own Canon 450D & Samsung GX10 also
overexpose landscape scenes when on their equivalents of
matrix metering - although not as severe as the Nikon. The
Canon & Samsung give washed out but still blue skies, and
only need -1/3 to -2/3 EC to get acceptable results."
En fin, podría ponerte infinidad de ejemplos más de usuarios profesionales, de nivel medio y aficionados a lo que les sucede lo mismo con la medición matricial y suelen entonces compensar a la baja.
Es evidente que Nikon te reserva a ti las mejor calibradas .