Brand new to this site/forum. Have a real brain-teaser. Got a 5DSr a few months ago, just getting around to testing it out. I find that with a solid tripod mount, using cable release, I am unable to capture a sharp image file with wide-open lenses. This is even using mirror lockup (which should be unnecessary). Using an Otus 85 at 1.4 @ 1/4000 & ISO 100, I get completely unfocused images, when they looked razor sharp in the vf & in live-view. I have the exact same condition using a Canon EF 85mm 1.8, IN MANUAL FOCUS MODE.
The situation in both cases begins getting corrected as the lenses are stopped-down, and the images get completely sharp at f4.0. But anything below that, is soft, and downright blurry, and unusable. As this same phenomenon is happening with both lenses, it appears to be something going on with the 5DSr. I am positive there is no vibration induced in the setup. RRS pod + BH-55.
Any ideas?
Sounds like a smaller aperture is used during LiveView and that it opens up for the capture. Don't know how this is supposed to happen though. Can you check that the lenses really are wide open during LiveView?
The fact that it happens with both lenses suggests that it has to do with the camera rather than one lens. Besides, I can certify that Otus 85 is sharp even wide open. That said, this lens is so precise that it is quite possible that what you see in the VF is "almost sharp", and therefore looks OK in the VF and on the LCD, and, when you open it in your computer and view it at 100%, "almost sharp" no longer is good enough. But that can happen on some shots, and should not on all of them. Another possibiity is the diopter. If it is off and you focus wide open, it will look fine in the VF, but be off on every shot, and, if the difference is small, you might not notice it on the LCD but see it big time on the computer screen.Last suggestion, micro-adjust. Did you do it? I have no experience with 5 DSR but plenty with 5D2 and 5D3, where it really made a difference.
Can you use Liveview and zoom in to manually focus and then take the shot? Are you sure NOTHING is sharp? Narrow depth of field can play havoc with such testing.
Yes, what you describe was exactly my procedure. On view-finder view + live-view, the Otus is extremely sharp, however after the exposure is made, the image is blurry. Same deal using the Canon 85mm 1.8 in manual focus. Using the exact same procedure, only on tripod, each successive f-stop sharpens the image, up to 4.0, where the image is in acceptable sharpness.
What you say about the diopter is very interesting. I have it set to about -3.0 so I don't have to wear glasses, and I had to adjust it so it was right for me. But to think it could be "off" is very interesting.
But hang-on a second...doesn't live-view negate that supposition? It's razor-sharp in live-view, and then I snap the shot, and it's blurry, out-of-focus
Regarding micro-adjust, I (mistakenly) thought this was for auto-focusing lenses..you saying I should adjust for the Otus, which is only manual?
Let me clarify - I havn't even loaded the shots onto the computer, I was so distraught after just looking at them in "playback" on the 5DSr's LCD screen, I didn't bother..
You think the computer will show me a sharp file when the 5DSr screen shows it blurry?
Are you focusing with Live View magnified 16x? You need to - I had some surprises when I didn't use full magnification. On the Canon, I also use the Zacuto along with LV 16x - it does make a difference. With the Zacuto, you can get pinpoint focus every time.
It's more of a magnifying loupe - you can hold it by hand or attach it to a frame that sticks to the camera back. I have the 2.5x version; there's also a 3x.
When I manual focus my Otus 55 using LV, i set magnification for 10x and use a Hoodman loupe. The Zacutto stuff is nice but not cheap. If I remember, the Hoodman is around $75.
You can also use focus confirmation using the VF with Zeiss mf lenses. Using focus confirmation is where you can AFMA the lens.
I use a combination of both LV and VF with focus confirmation when manual focusing.
TY - I will check out all these things, and post the results. What's throwing me is 85's not really that long, and at 1/4000 there should be no inducement for blur..isn't that the whole lure of a 1.4 - to shoot it wide open as much as you can?
I know it's a real precise, sensitive focusing helicoid; it's beautiful. What if I shot a portrait hand-held? This has got me rattled. I shot for many a year at the 4 x 5 ground-glass back, so am familiar with this technique. But I appreciate everyone's help, and will ultimately discover what the fault is.
CJ Han wrote:
TY - I will check out all these things, and post the results. What's throwing me is 85's not really that long, and at 1/4000 there should be no inducement for blur..isn't that the whole lure of a 1.4 - to shoot it wide open as much as you can?
I know it's a real precise, sensitive focusing helicoid; it's beautiful. What if I shot a portrait hand-held? This has got me rattled. I shot for many a year at the 4 x 5 ground-glass back, so am familiar with this technique. But I appreciate everyone's help, and will ultimately discover what the fault is....Show more →
First off, you mentioned that you have only checked the images on the back of the camera? If so, you need to put the images on a computer and analyze them there. Look for a plane of focus that is sharp(not necessarily where you thought it should be), it could just be a missed focus versus a blurry image per se.
Good luck manually focussing an f1.4 handheld using manual focus, the focus plane is very narrow. DSLR viewfinders are not exactly suited for manual focus and what you see is not what you get either.
You need to make sure everything is static(not moving) before assuming the lens/camera isn't sharp.
TY for the thoughts, wise words. I am thrown because the "live-view" images are so razor sharp, they could cut you - especially with the Otus. These are reading right off the sensor, and then throwing it to the LCD. At 1/4000 exposure, shake shouldn't exist, and I'm sure it doesn't. I'm new to this board - is it possible to upload images so everyone can see some samples, do you know?
You can put your images on the web yourself if you have a private server or reference to photobucket or the like. Or you could subscribe (and pay) to use the service on FM.
So if I am reading all this correctly you have your R on a sturdy tripod, use LV to determine precise focus at 16x magnification, use 1/4000 SS and still get a blurry image. Correct? Have you found out if it's misfocused blur or motion blur? Did you try to shoot a subject standing on a carpet or other textured foreground to determine if there is a focal plane that actually is in focus?
