Canon 5D Mark II Soft Images !

[payamghafoori]

Dear all,

 

I'm a professional wedding and modeling photographer. I recently switched to 5D mk II of Canon and used it with

70-200 f/2.8 L IS and 24-105 f/4 L IS.

Here what bothers me.

 

In even many outdoor locations with bright daylight, aperture even close to 6.3 and shutter speeds of at least

1/100 with IS on and low ISO ... (everything good for sharp images), many shots were captured pure soft (when

viewed at 1:1 scale i.e. 100%). No part is in sharp focus.

 

I check both JPEG (noise reduction set at low) and raw. Both are soft.

 

This may be of no problem in normal print sizes but for wedding photos which may be ordered on very large (say

100x70 cm) this could cause serious trouble.

 

I know many serious and professional folks out there have been dealing with such problem, but here I AM SEEKING

SOLUTIONS! maybe I shall go with IS off, or using only center AF point or what!

 

Please share your valuable experience.

 

Many thanks

 

Payam Ghafoori

www.beautygraphs.com

[porter]

<p>Example shots available?</p>

<p>Suggest you take a pic on a tripod using live view for focusing to test it a bit more accurately. Process your images in Canon's DPP and play with the sharpness. FWIW I like my sharpness set to +6. Seems to render quite sharp results using my two lenses (28mm f/1.8 and 50mm f/1.4). </p>

<p>Hope that helps.</p>

[tony_leinster]

<p>Couple of things to check, are you using a tripod in which case you will need to turn off the IS, also I would recommend only using the centre AF point for critical sharpness. Next step would be to check if your camera has any setting for the fine focus adjustment already ( read your manual, this is hidden in the custom functions), start with this at zero then check if your camera is back or front focussing (search the forums for how to do this) and then adjust if necessary.</p>

[porter]

<p>Oh, one more thing: Take off any filters you may have on. Myself and a couple other photographers I know have had issues with cheap UV filters causing excessively soft images.</p>

[Marcus Ian]

<p>We can't tell you with any certainty whether or not the problem is user related, or equipment related. Since you imply you are having this problem <em>consistently</em> with one excellent, and one very good lens, I would have a tendency to think that it's a user error, not an equipment problem.</p>

<p>You should determine (in absolute terms) how 'soft' the images are in 'ideal' circumstances first (by using a tripod, with IS off, to get as perfectly 'sharp' as possible images). If you can consistently generate critically sharp images that way, the you know that your technique is the problem. <br>

If you find yourself unable to get precisely focused imagery that way (and @f6.3 it should be pretty darn sharp) , then you should consider it an equipment problem, and resolve from that direction.</p>

[john_bellenis]

<p>The consensus here is correct, the only way to figure out if this is a problem with your technique or with a piece of your equipment is to be very analytical and do a series of controlled tests. Without going over the methodology excessively, you will need a tripod, a target set at some different distances, use a selection of apertures, AF points, manual focus and duplicate your tests with both lenses, etc. - detailed notes will help you hugely. If it does turn out to be equipment related, these results will help Canon quickly diagnose the problem.</p>

<p>Once you see, in controlled conditions, what your combinations are capable of, you can then see if there anything that you could be doing better to get close to that handheld. I also would ditch any filters and spend some time with micro focus adjustment.</p>

<p>Good luck - it'd be great if you can report back with your conclusions.</p>

<p> </p>

[brad_trostad]

<p>Are you focusing and then re-composing in servo mode? Did you receive the camera with some amount of MFA already set?</p>

<p>IMHO - quickest way to demonstrate the best possible sharpness of the camera + lens is to use live view, manual focus, tripod, shutter delay, zoom 10x on a very small area of interest and manually focus. This image should be quite sharp and will be representative of your future photos when everything is working and being used properly.</p>

<p>FYI - My 5D2 requires approx +3 MFA for all lenses. Before that adjustment everything was soft. I don't recommened using MFA until you can rule out the previously mentioned things first.</p>

[15sunrises]

<p>Nearly impossible to diagnose without examples.</p>

[craig_meddaugh]

<p>Goodness, another "my equipment sucks" thread with no useful information to help the poster.</p>

<p>OK... we seem to be getting the problem with two lenses that presumably worked on another body. So that leaves us with the body or the photographer. To rule out the body, take out your tripod and find a nice target. Use your best tripod technique and take some shots with AF and then use liveview to get the best possible MF shots. If your MF shots are sharp and your AF shots are defocused, see if you can fix it with the focus adjustment tool. If not, send the camera in for service. If your MF shots are soft, something odd is going on. </p>

[a._t._burke]

<p>Mr. Ghafoori...</p>

<p>I too bought a 5DII that gave much softer results than my old Panasonic FZ30 or even older Minolta 5D. What a disappointment! </p>

<p>My solution: Paint Shop Pro 9 "sharpen more." I tried Photo Shop and Canon's sharpening tools as well as a later Corel Paint Shop Pro. Although all improved the image, it never seemed to be just right. I have had one photo that looked over-sharpened, using "sharpen more," but all the others have been spot on. I usually use the 24-105mm L lens but I find the solution seems to work right for all lenses. Until I used "sharpen more" I could tell no difference between the expensive 24-105mm L and my older, less expensive 28-135mm IS. </p>

<p>Hope it works for you. Within a few months ago, I bought a new, in the box, with license, Paint Shop Pro 9 for around twenty bucks on Sleazebay. </p>

<p>A. T. Burke </p>

<p> </p>

[mark_poseley]

<p>When you view a 5D MKII image at 100% you are looking at an equivalent 52 x 78 inch print, on a screen w/ 72 - 96 dpi resolution. I am not surprised you see blurry photographs.</p>

<p>Print the shot at the sizes you might sell, and see what you get. I can print up to 13 x 19 at home and the 5DII is capable of amazing detail!</p>

<p>Also do what folks above have suggested to make sure hardware is OK. You do not mention what camera you used before; going from a crop to FF can be difficult in the DOF arena. It threw me for a few loops when I made the switch.</p>

[dan_south]

<p>My 5D2 works fine. There are plenty of examples on my website.</p>

<p>Some suggestions:</p>

 

<ul>

<li>All digital images require proper sharpening.</li>

<li>Review your images in Digital Photo Professional before opening them in a more complicated raw editor such as Photoshop.</li>

<li>Make sure you evaluate sharpness at the plane of focus, not in the soft-focus "bokeh" areas of the shot.</li>

<li>Test your lenses on something simple. Test both auto and manual focus (live view magnified) to rule out an autofocus adjustment problem (unlikely but possible).</li>

<li>Test focus on a flat vertical plane with a lot of detail, e.g. a wall covered with posters or patterns with a lot of contrast.</li>

<li>1/100 isn't fast enough to freeze subject movement.</li>

<li>If you are handholding shots, you might be moving after acquiring focus. Use good shooting discipline for best results.</li>

</ul>

 

[peter_rowe]

<p>Payam,<br />This can be very frustrating. I think first off we need to determine if you really have a problem, and the only way to do that is to post a 100% crop (or a few of them) on here so we can all see it. Owners of the 5D MK II will be able to tell you in an instant if the shot is acceptable or not - make sure and tell us what lens you use and leave EXIF data intact if you can or just tell us the shooting conditions.</p>

[philip_wilson]

<p>Like Dan my 5DII works fine and is my sharpest DSLR. Of course images from the camera need sharpening like all digital cameras (even my Leica M8 which has no anti alias filter benefits from gentle sharpening).<br>

With some lenses wide open the 5DII shows up lens issues - it is a while since I had the 24-105 but my copy was quite soft at the edges wide open. The 70-200 should be better but I do not have the MkI F2.8 IS. AF accuracy is important.<br>

If you are concerned that at very large print looks slightly soft under close inspection - this is normal. Your 100 x 70 cm print is printed at about 135 DPI assuming no cropping. You can expand the pixel count using interpolation which will help. If you want to print that large and have an image closely inspected (not viewed from a normal distance) then you will need more resolution which means Medium Format.<br>

In the case of Mr Burke (who has an 8MP bridge camera from 2005 and a 6MP Minolta from 2005 that are sharper than the 5DII) I can only suggest that either his 5DII is broken and he should send it back to Canon or that he is doing something very badly wrong.</p>

[a._t._burke]

<p>Mr. Wilson…</p>

<p>“In the case of Mr Burke (who has an 8MP bridge camera from 2005 and a 6MP Minolta from 2005 that are sharper than the 5DII) I can only suggest that either his 5DII is broken and he should send it back to Canon…” </p>

<p>Megapixels did not count in my evaluation. I judged a cut from within a file of each camera’s output that would size the object the same size on my screen when shown at full pixel per pixel resolution. </p>

<p>You are right in principle thinking that I should send it back for adjustment. However, the first sample came DOA. By the time I got it replaced with this one, I had missed Chanukah, Christmas, New Years and the SWAT dinner. I have no idea how long Canon service and shipping will take. Whatever time it takes will be a significant portion of my remaining ability to use a camera. I’d rather adjust the results from their sub-standard product than have nothing at all. Plus, I don’t know that they can improve the output, as what I’ve got may be as good as Canon can do. </p>

<p>“….or that he is doing something very badly wrong.” </p>

<p>Yes, that is probably right. I’m relative new to photography having only started in the 1920s. Given a little more time, I’ll learn. </p>

<p>A. T. Burke</p>

[philip_wilson]

<p>Mr Burke - my intention was not to cause offence but if your 5DII cannot out resolve 2005 bodies you have a serious problem. On the same size image a 5DII (with a decent lens) will out resolve most bodies out there. My own is clearly superior to my 7D, 1DIIN, Panasonic G1 and Leica M8 in resolution. If you are getting superior resolution from a 2005 bridge camera with a built in super-zoom then you have a major issue.<br>

There are very few tests of these cameras but I did find one for the FZ30. Here are the resolution results http://www.cameralabs.com/reviews/PanasonicFZ30/page4b.shtml<br>

1700 and 1750 Lpph (vertical and horizontal)<br>

The same source tests the 5DII and finds resolution of 3000 and 2950<br>

http://www.cameralabs.com/reviews/Canon_EOS_5D_Mark_II/resolution.shtml<br>

This is what I would have expected given the age, sensor size, pixel density and lens in use on each camera. The fact that you are seeing the 5DII with lower resolution than the FZ30 suggests that you are getting about 1/4 of the 5DII resolution. Without a detailed assessment of your issue it is difficult to say why. Perhaps you can post crops.</p>

 

[jamie_robertson2]

<p>Mr Burke, either your gear is badly out of whack or you haven't got a clue what you're doing. The same advice goes to the OP. Any pro wedding photographer should know how to check his gear by sticking it on a tripod and taking a shot of a static subject, first with AF and then focussed using live view. </p>

[a._t._burke]

<p>Mr. Wilson….</p>

<p>Thank you for your reply, kinder to me than I was to you. </p>

<p>I may not have made myself clear. I know the Canon resolves almost twice the line pairs per picture height as opposed to the FZ30. That’s because it’s a 21 mp camera, compared to and 8 mp camera. I could make a much sharper looking print four feet on the short side out of the Canon. So I agree with your lpph argument. </p>

<p>What I’m talking about is the clarity, size for size. Let’s say I photographed from a distance something like a newspaper or a test chart that would have the ratio of 1200 units to 1600 units. I then photograph it with both the FZ30 and the 5DII. However, I don’t photograph it using the same equivalent length of lens from the same spot. Instead, I adjust the lens on the FZ30 so that it gives me an equivalent image to a 50mm on either a full frame 5DII or a 35 mm camera. Then, to compare like with like, I take another picture with the 5DII farther back, experimenting until the photographed object occupies about 1200X1600 pixels within each file. With the FZ30, I have it set to output both a .tif and a .jpg each time I push the shutter button. I have the Canon set to put out a raw and a .jpg each time I push the shutter button. I then use the Canon-supplied software to convert the raw to .tif. Now I can look at the results to see which captures the best acuity of a 1200X1600 pixel cut and compare the results in both .tif and .jpg form in between the two cameras. It of course will be a larger portion of the file created by the FZ30 than the 5DII because the 5DII has more total pixels available. If I look at those results, the FZ30 looks sharper and more detailed than the 5DII. The 5DII just looks soft, even without side-by-side comparison. </p>

<p>I actually did the experiment and, although I didn’t get the objects sized at exactly 1200X1600 on both cameras, it was quite close. It was at that point that I decided the FZ30 produced much better results, pixel per pixel than the 5DII. That was in both .jpg and converted raw to .tif. </p>

<p>If I had filled the frame on both cameras with the same object, of course I would have more definition with the 5DII because I had more llph. In other words, more total pixels to gather data. The solution of “sharpen more” in Paint Shop Pro 9 brings both the .jpg and converted .tif results to about the same crispness as the FZ30 does right out of the camera. I’m using best .jpg on both. </p>

<p>The OP was actually not asking for a comparison or whether the 5DII produced crisp results or not. He just wanted a solution to a perceived problem. Perceiving I had the same problem and had used about six different types of software, trying to improve the results, my purpose was to share my most successful solution with him. </p>

<p>Although I can’t compare my 5DII to others on the board, who state they get very sharp crisp results, I suspect it is not a so much a difference in perception but that my camera is sub-par. On various photo boards, there have been numerous discussions as to whether the 5DII produces all that sharp results, or not. Although I don’t know, my suspicion is that quality control is not what it should be, and in fact, there is a fair difference between 5DIIs. </p>

<p>When I bought my 5DII, I purchased at the same time the 24-105 “L” lens, a grip, a 580EX flash, and two AA battery pack inserts for the grip. The first 5DII was DOA. The battery door on the grip was partly open because it would not close properly. An Exacto knife fixed that for me. One of the two battery packs needed to have flashing trimmed off it before I could insert it into the grip. I needed to trim some excess mold flashing from the top of the grip to get it to fit snugly to the bottom of my camera. The speed light seemed to work okay, but, where the outer housings were joined, they were not joined evenly. All in all, I thought it was a rather poor overall effort from Canon. </p>

<p>Now that I have trimmed pieces to fit, and at least have a working camera that is not DOA, I’m inclined not to send what I feel is a shoddy product for adjustment by the same people who sent me the garbage I outlined above. I don’t know how long I’d be without a camera. I already lost a number of major significant events that I originally purchased that assortment of junk with to photograph. I also don’t have much faith that the company that would send out garbage would then make that garbage into a diamond, as it begs the question, if they can get it right, why didn’t they do it on the first two tries? </p>

<p>Unfortunately, I think I’m just out the money. I’ve gone back to using the FZ30 except where I need more megapixels so I can create a larger image. Using the FZ30 saves me the trouble of running it through Paint Shop Pro 9. </p>

<p>As a side note, the old Minolta 6 mp 5D gives much truer colors right out of the camera than the Canon 5DII. I wish it was 21 mp. If I had to purchase over again, I would take my chances with a Minolta 900, rather than the Canon 5D. The reason I chose Canon is that I’ve had such good results from their upper level film cameras and had been quite pleased with how well the IS worked on the 28-135 mm lens. That lens still focuses faster and truer than the 24-105L and, in my use, I can tell no difference in the newer generation IS. </p>

<p>I’m glad you got a good 5DII. Evidently Canon can get it right, at least part of the time. </p>

<p>A. T. Burke </p>

[a._t._burke]

<p>Mr. Robertson....</p>

<p>1. If, as you speculate, I don't have a clue as to what I'm doing, how do I get good results from the FZ30 and the Minolta 5D for comparison? </p>

<p>2. I would speculate only wedding photographers think it takes a wedding photographer to be able to use a tripod and focus live view. </p>

<p>A. T. Burke </p>

[philip_wilson]

<p>Mr Burke - thank you for the more detailed explanation. I have never used either of the two cameras you discuss so I cannot talk to them specifically. In terms of the 5DII it certainly seems like you have had bad luck.<br>

On your issue I can make the following suggestion and offer the following thoughts:<br>

1 Is it possible that your lens is not focusing correctly - most lenses are fine but some do need micro adjustment. there are lots of links on this subject but I find this site has good advice http://www.northlight-images.co.uk/article_pages/cameras/1ds3_af_micoadjustment.html Try using live view and a tripod and see if you get better images - this is usually a sign of front or back focus<br>

2 The 5DII RAW images do need sharpening when they come out of the camera. I always notice that my Leica M8 10MP images are much sharper than the 5DII when I look at RAW images from each body. After sharpening the Leica does not have the detail of the 5DII<br>

3 The colours of the 5DII and the DPP default settings tend to give very punchy (over saturated) results. It is something that most digital bodies seem to do these days and I think that the world of the internet is partly to blame - the punchy colours look better on the screen. As a side comment I am shocked at how people don't notice image quality these days. My own favourite is hotels where you get a 16:9 (HD!) TV but most of the programming is in SD. Since you do not get a real remote you cannot change the aspect ratio. I have yet to find anyone else (my work colleagues) who has noticed this until I pointed it out! I process my RAW files in ACR and find I can get good results with care but the only DSLR I have that delivers realistic colours (assuming the WB works and there is no IR contamination) is the Leica.</p>

<p>In terms of build quality the fact is that camera have become "consumer electronics" and have a limited design life, built in obsolescence and lower build quality. My L series lenses are not as well made as most of my old FD lenses and similarly I doubt that any of my digital bodies (1 series included) will last like my F1s. The only way to get top build quality is to buy a product like a Leica. Unfortunately you pay for the build quality - the lower quality (summarit) Leica lenses start at $1500 and the true Leica lenses run from about $2300 to $11,000. Unfortunately in the digital age mechanical construction only needs to be good enough to survive until the electronics are obsolete. All three of my F1s function perfectly (one is almost 40 and the other two about 30) and will probably do so for many years to come. I doubt my 5DII, 7D or 1DIIN will last much over a decade.</p>