PDA

View Full Version : Similar 5D mark ii problems anyone?



anulogy
02-06-2009, 04:03 PM
I have had my 5D mark ii for about a month now. After seeing several posts where peoples pictures are tack sharp, I am starting to think it might be my camera. Shots i take indoors are way off colob (W/B) wise and shots i take everywhere seem unusually soft. I kept these thoughts to myself until i came across this on FM review:


<span style="font-family: Trebuchet MS;"]"Ihave just upgraded to the 5D2 from the original 5D (which I'll call the mark 1 here on in).
It's weird. It does everything on the tin but somehow I'm just not "wowed" by the shots. I seem to be missing focus more often than with the mark 1 and AWB seems way off when shooting interior (and I don't mean flourescent).
I always shoot RAW by the way, with neutral settings. I view the results in Apple's Aperture.
The first body I got I sent back to Jessops and got replaced. I was convinced it was under exposing by almost a full stop, more even at times. But it seemed to depend on the lighting conditions: more prone to error under lower light esp. indoors. Also, I had developed with the mark 1, a style of shooting whereby I'll run off a burst at my subject and expect a 50% hit rate viz focus. And usually got more than that. But with the mark 2 my keepers are much thinner on the ground. But add to this that some shots WITHIN THAT BURST are simply under exposed - even in comparison to the other shots in that burst. Odd. It's like the Digic IV just skipped a beat or something.
I use AF and matrix usually, but leaning more and more towards spot metering in camera, with my thumb married to the Exposure lock button.
Maybe I am doing something wrong? Do I need to calibrate my lenses? I have tested the Mark 2 with a 70-200 ISM 2.8 L and a 24-105L. Both a bit off all the time somehow.
Anyway - the second copy seems identical to the first in all respects. Meaning, still prone to underexposing. Still checking on focus side of this body.
I have rumaged in forums and sensed that Canon my have changed it's exposure reference point, to be more in line with industry standard. Can anyone here verify this? As far as I understand, this would imply a "lower" exposure to be "correct".
Or is it that the 14 bit detail increase means that one should be exposing in a slightly diferent way to take advantage? To grab those bits at the bottom in effect.
I am (clearly) only a keen amateur (heck, I am who this camera is aimed at lets face it) and I would welcome constructive criticism (or total destruction of it's argued well). Please tell me where I'm going wrong.
Thinking of replacing this with a 1ds3, but wondering if this is sheer madness just to get snappier (much) focus.
So - anyone else puzzled by the exposure and foxus side of things?
Thanks."


<span style="font-family: Trebuchet MS;"]I shoot straight to large jpeg / AWB / Neutral settings.


16-35 f/2.8L II
70-200 f/2.8L



I Will post test shots when i get home, but any comments for now?

Colin
02-06-2009, 04:45 PM
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]Not being a MkII owner, I may have nothing useful to contribute... Thoughts anyway.<o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]If you're looking at detail 100%, the higher resolution sensors will look softer, because they can better expose the differences of the lens. <o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]I don't know if it does, but if the sharpening is being done in relation to pixel units it would have finer sharpening, so it can actually look less sharp for the same amount of sharpening, because you don't get the pixel structure itself making a more visible edge. you know, an image may look really sharp if you shrink it down to 640x480 and display it at that resolution, but at full size actually rather dull.<o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]Were you using 'neutral' before? It seems that neutral is designed to compress the dynamic range so that you can best fit it into your output file, and then perform dynamic range, curve, and white balance operations from there. It's not accurate, and it looks worse than accurate. So far, the stuff I've found most reliable, to actually start with something that's reliable, is to shoot in raw, with some small .jpg thumbnails on the side, and then get DPP to go 'Faithful', set sharpening with a 100% view on a point of focus, adjust the white balance, take a look at the RGB levels to ensure that nothing's getting clipped, and after seeing the image for what it more or less is, truly, to begin with (or the best that could be captured, figure out how I'm going to adjust the curves and saturation to best convey what the image has to offer.<o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]The only other thing I'd really suggest is to use RAW files whenever possible. I actually like the Canon software, because I only make basic adjustments for the time being. I tried Lightroom for a while, but found the controls a little too abstract, and on a few pictures, couldn&rsquo;t figure out how to decode a raw file in a way that looked remotely accurate. Subjectively appealing, maybe, but if I didn't like the effect, I couldn't undo it.<o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]I really think Bryan's onto something, or at least sit on the same part of the boat, when it comes to using the Canon software to decode the RAW files, and then apply your post-processing methods of choice for whatever effect you desire.<o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]<o:p></o:p>
<p style="margin-left: 4.8pt; margin-right: 4.8pt;"]<span style="font-size: 7pt; color: black; font-family: Verdana;"]Anyway, that may have been way off target, but.... Sorry[:P]<o:p></o:p>
<p style="margin: 0in 0in 0pt;" class="MsoNormal"]<span style="font-size: small;"]<span style="font-family: Times New Roman;"]<o:p></o:p>

Todd Reichman
02-11-2009, 01:42 PM
As far as the underexposure is concerned I find the mark2 to be much more accurate than the mark 1. After shooting with the mark1 for 3 yearsI had gotten to the point where I knew when the meter was "on" and when it was giving me false info and I knew exactly how to compensate. I'm finding with the mark2 that its far more accurate and my old compesations are now a bit too much. Have you tried using a controlled situation and a light meter to see how far off it really is? Are you shooting in Manual when this happens, AV or somthing else?


As far as focus is concerned, we're seeing much better accuracy over the mark1. Having shot a 1dsmk3 and both 5ds I actually was underwhelmed with the focus performance of the 1dsmk3. Its "faster" to acquire focus but not by a significant amount (not $8000 more significant) and I didn't find it any more accurate.


How are you focusing? Selecting the point, focus-recompose, etc? Your technique my be influencing the results you are seeing. Give us an idea of how you are doing it.


I'm sad to hear that you are unhappy as we are over-the-moon thrilled with the mark2. Totally satisfied in a way that we weren't with the 1dsmk3. I'm finding high ISO performance to be amazing and very sharp, and colors to be better. Plus I think the AWB is a little more accurate and (I'm finding anywat :D) that the focus is significantly improved.


- trr

Daniel Browning
02-12-2009, 01:45 AM
21 MP requires much more stability, focus accuracy, and lens performance compared to 12 MP. The 5D2 will give you 12 MP files that are just as good as the 5D (downsizing to 12 MP or printing at the same size will hide focus errors and camera shake). But if you want to print larger or use the higher resolution, it will require improved technique.