sanblas wrote:
regarding the 5dmkii focus
I would imagine for most shooters the center focus point is fine, however shooting weddings in crappy light with wide open primes it sometimes hunts too long...it also has problems on things IMO it shouldnt like text & dresses. I realize I'm stressing an AF system with low light, wide open shooting where I need accurate focus or the shot is blown.
overall I've been a little underwhelmed with the 5dmkii center point(the only point I use as the rest are worthless) and have actually been relying more on my 5d classic recently for low light work.
it stinks they gave us such a great sensor that can shoot in the dark, but such painfully mediocre at best autofocus
I need to send the mkii to CPS for a clean and check though, it could be sample variation and I have put about 40k shots on the mkii this summer...Show more →
very odd that is would do worse than your 5D classic!
the other day i shot in my shady room under natural lighting at f/1.4 with both 7D and 5D2 and the center point af seemed very similar to me. Once microfocus adjustment was done the results were actually not too bad at all, if not perfect, but really not bad at all. Before microfocus adjustment the results were simply godawful beyond belief though so I really think MFA is VERY important!
under fluorescent lighting the results were note quite as good from either and each had certain subjects they would totally bomb on.
outdoors at greater subject distances they were ok, but would make mistakes with some regularity, the 5D2 almost seemed a little better!
but i do need to test more, haven't used it nearly enough to say, so don't put too much weight on this
a side story:
both are pretty terrible if you try to focus on leafy trees though, very random, slight changes in lighting can mean mostly in focus to almost never in focus (doing a tough f/1.4 test with wide to medium-tele focal lengths). I think leafy patterns and sky shining through the leaves and all can drive the AF systems crazy (i wonder if this is the case with sony/nikon/etc. too?), same story with 20D,40D,50D,5D2,7D, the 5D2 perhaps the best by a slight margin. And yeah f/1.4 is shallow DOF but I am talking not a single leaf on the tree remotely in focus sometimes for as many as 10-15 frames in a row (at other times it can do 8 perfect frames in a row though, especially if it is overcast or gloomy or late evening sun) and the focus seems to be maybe as far as 100', yes feet, off, nothing anywhere in frame close to being in focus.
EDIT:
just did a quick out point test in a very dimly naturally lit hallway
didnt test enough to say whether either camera was more precise or not with the outer points but I can say the 7D outer points do lock on MUCH faster, sometimes the 5D2 are close but I rarely ever quite so fast and sometimes they even need to hunt a little before locking on while the 7D always zeroes right in.
At the very least, the non-center points on the 7D are far faster.
(i'm not sure center point for one shot is all that different between the two, certainly nothing the difference in speed between the alternate points)
keithreeder wrote:
They're on the case now Simon - and results will be even better from a tailored 7D profile than we're seeing from 7D files converted to DNG.
And as I mention on another thread, Cap One does a very good job even now with high ISO DNGed 7D files.
Fantastic shots, by the way - I must get down to Croft some day.
That's good to hear Keith as I am very much looking forward to seeing what C1 can do with the 7D files.
Only two events left at Croft this year, but both are crackers. The European Rallycross Superprix is on Sunday 1st November, and the Christmas Stages Rally is on the 28 September. Both are very good events. Drop me a line if you're going to head down.
skibum5 wrote:
very odd that is would do worse than your 5D classic!
the other day i shot in my shady room under natural lighting at f/1.4 with both 7D and 5D2 and the center point af seemed very similar to me. oddly when i aimed at a painting out in the very dark hallway the 7D actually hunted and hunted and hunted and not the 5D2.
but i do need to test more, haven't used it nearly enough to say
It could be his 5DII AF optics have gotten dirty/fogged. To focus the light needs to go through the reflex mirror and bounce once before hitting the AF sensor. I can say my 5DII AF is a notch more surefooted in low light than my 5D was. Nothing earth shattering but noticeable.
Gochugogi wrote:
It could be his 5DII AF optics have gotten dirty/fogged. To focus the light needs to go through the reflex mirror and bounce once before hitting the AF sensor. I can say my 5DII AF is a notch more surefooted in low light than my 5D was. Nothing earth shattering but noticeable.
note that i've edited my message since you quoted it and the part about hte 7D hunting in the dark halway was only in ai servo mode not one shot, need to look into that more, maybe it was trying to stay on a near target or something (although i had it set to fast switch mode)
adding:
"a side story:
both are pretty terrible if you try to focus on leafy trees though, very random, slight changes in lighting can mean mostly in focus to almost never in focus (doing a tough f/1.4 test with wide to medium-tele focal lengths). I think leafy patterns and sky shining through the leaves and all can drive the AF systems crazy (i wonder if this is the case with sony/nikon/etc. too?), same story with 20D,40D,50D,5D2,7D, the 5D2 perhaps the best by a slight margin. And yeah f/1.4 is shallow DOF but I am talking not a single leaf on the tree remotely in focus sometimes for as many as 10-15 frames in a row (at other times it can do 8 perfect frames in a row though, especially if it is overcast or gloomy or late evening sun) and the focus seems to be maybe as far as 100', yes feet, off, nothing anywhere in frame close to being in focus.
EDIT:
just did a quick out point test in a very dimly naturally lit hallway
didnt test enough to say whether either camera was more precise or not with the outer points but I can say the 7D outer points do lock on MUCH faster, sometimes the 5D2 are close but I rarely ever quite so fast and sometimes they even need to hunt a little before locking on while the 7D always zeroes right in.
At the very least, the non-center points on the 7D are far faster.
(i'm not sure center point for one shot is all that different between the two, certainly nothing the difference in speed between the alternate points)
"
You probably know this already, but the center point AF behaves diffrently in servo / one shot. In low light, my 7D locks quickly in one shot mode, while at the exact same location, will not lock at all in servo mode. I use AF priority / tracking priority.
Edit: skibum posted while I was writing there, so we share the experience of servo / one shot.
Place in
C:\Documents and Settings\[your username]\Application Data\Adobe\CameraRaw\CameraProfiles\
and restart LR. A new profile will pop up under "calibration". So far it works good for me, but I haven't tried fluorescents yet. I wouldn't recommend using the profile for higer ISO shots as it corrects for the defective red-yellow-orange hue/saturation shift, and thereby it will increase chroma noise. Any feedback is appreciated, but sofar it works for me.
alundeb wrote:
You probably know this already, but the center point AF behaves diffrently in servo / one shot. In low light, my 7D locks quickly in one shot mode, while at the exact same location, will not lock at all in servo mode. I use AF priority / tracking priority.
I've noticed this problem with AI Servo locking in shooting moving deer in low light in the woods. I think that by using expansion of the AF points helps. In good light with contrasty targets it's great. I've been busting my head over this for two days now trying different settings. It also has been cloudy and the color of the deer's coats are matching up with the woods, so they are tough targets but I've not had this problem with other cameras. ...Show more →
Yakim Peled wrote:
Have you found 40D's metering so bad that you needed more than 2 stops? I have the 40D for more than two years and never needed more than 1.
Meetering is a starting point best guess from the camera, artisticaly I would prefer more than 2 stops of compensation and it also helps to be able to bracket -3,-1,+1 or something
As before all images shot as adobe jpg's, neutral setting, hightlight tone priority, in camera noise reduction 'standard', and then auto batch processed with my normal routine for post-processing proof galleries from events to boost, saturation, contrast, sharpness and converted to sRGB for display on the web. Night soccer was shot in manual, all others aperture priority.
troutstreaming wrote:
As before all images shot as adobe jpg's, neutral setting, hightlight tone priority, in camera noise reduction 'standard', and then auto batch processed with my normal routine for post-processing proof galleries from events to boost, saturation, contrast, sharpness and converted to sRGB for display on the web. Night soccer was shot in manual, all others aperture priority.
These shots are simply stunning. Thanks for sharing. Well done!
The 7D appears to be fully capable in most situations of pro-level work that was once the bastion of 1Ds bodies only.
As before all images shot as adobe jpg's, neutral setting, hightlight tone priority, in camera noise reduction 'standard', and then auto batch processed with my normal routine for post-processing proof galleries from events to boost, saturation, contrast, sharpness and converted to sRGB for display on the web. Night soccer was shot in manual, all others aperture priority.
Troutstreaming - Thanks for these.
Seattle's my hometown, and I am sorry to have missed (possibly) Griffey's last game. But I do get to go to the Twins' tiebreaker vs. Detroit on Tuesday...
Ok, back to the photos - just to make sure, your "normal routine for post-processing proof galleries" did not add any more NR to that provided by 'standard' in-camera NR? And, I suppose, if anything, the sharpening would increase the visibility of noise? Well then... these look great! Thanks again for posting them. I can't wait to get my 7. (<-- I'm trying to be on the leading edge of the acknowledgement of the digital era, and will [try] henceforth to only refer to it by its number) [Oops, forgot about the EOS 7 (Elan). Ok, it's the 7D.]
