lack of sharpness... for lack of a better way of putting it.

Discussion in 'Wedding and Event' started by jason_inskeep, May 16, 2010.

  1. I was recently asked to shoot for a fairly large social event, more specifically a high school graduation. It was a last minute thing, the original photographer bailed and they needed some one who could fill the bill. This is not my normal type of shooting. I really did enjoy the whole process and wouldn't mind doing more of it if I got the opportunity but... There are a couple of issues!
    First the coordinator of the event wanted a wide angle shot of a stage full of students at the end of the ceremony. every thing seems to be sharp except the students!
    The second problem is kind of the same as the first, I have some fairly fast glass; 24 2.8, 50 1.4, 20-35 2.8, and I was shooting at a fairly high ISO ( between 800 and 1600) in an auditorium. How do you shoot in these situations?
    Lastly, a friend of mine suggested that I " put the flash in E-TTL, put the camera in Program mode, and stay no more than 20 ft away... ". Tried that in some test shots on stage before the ceremony started and due to the spot lights I was having a problem keeping the subject from getting way over exposed, I even put the flash down 3 stops, and the back ground turning totally black. The other problem I was having was never being able to get the camera to focus on the subject. P mode sugested 1/60th at f/4 and every time I fired off a test shot I got a black background, almost white subject ( in the face at least) and a lot of motion blur. I took the flash off, put the camera in manual, set it at 1/50th and 2.8 and all the sudden all my problems went away, at least mostly, for that part of the shoot. Is there something that I may not be doing right with my flash?
    I guess there is actually one more thing, how do you compete with other peoples flash and guys who are doing the stage lighting? It seemed like the stage ligting wasn't consistant all the time, I know it wont neccessarily be across the stage but shooting at the same spot and from the same spot with the same camera setting for every shot should get about the same results for every shot. I also had a problem with parents shooting in the back ground with every thing from point and shoots to slr's with flash guns in the back ground and that ruined a few shots. Is there any good way to make up for these types of problems?
  2. Basically what you want to do, is make sure you have a good exposure for the background (by metering like you normally do without a flash). This will allow you to have good background exposure, you would then use the flash to fill, and if the subject was slightly under exposed, now they will be correctly exposed with the flash, and the background will still be exposed.
    ETTL should manage that fairly well, if you're not quite sure about settings, expose for your setting without flash (in manual) and then turn the flash on.
    I believe there is also a way to do this with Tv mode... I think by setting to your fastest flash sync speed, and then the camera picks the aperture.
  3. In this situation you'll need to power up your flash to compete with the stage lighting and the best results would come from a multiple flash setup. A dark background isn't necessarily a bad thing.....background lighting would be the answer if you want to avoid it. Your images are likely soft due to the shallow DOF and some motion blur with the students.
  4. I did try setting a good exposure for the back ground as well... the problem ended up being that the spot light was so bright that the for ground ended up being just a touch over exposed with or with out flash, flash did seem to make it worse though.
  5. David,
    So really just crank up the flash and keep the black back ground, that would be the right way to go in the future? I am pretty new to this type of photography and wasn't sure if that was acceptable! If it is then I am glad to support it.
    Unfortunately at the moment I do not have the equipment to put out more lights. Untill very recently I was very much into available light only photography.
    So in auditoriums and churches how do you usually set up your cameras and on camera flashes?
  6. Stage lighting is lighting is usually very bad and always very difficult. We can't possibly teach you everything you need to know about lighting and flash in a forum thread. And I can't say what I would have done, I wasn't there. Off the top of my head, I might have used spot-metering in aperture priority mode. I probably would have staked out my locations and used the flash in manual. Bright spots, dark background, muddy middles, are going to test any TTL system, Canon's especially. And most TTLs don't do well with spot metering. If you wanted the background lit, you would need to light it! It's dark because the spots are brighter than the background! Hence the use of spots. You can't light both the background and the foreground (up against spots no less) with a shoe mount flash.
    In terms of image sharpness, post the image. Otherwise we are just speculating.
  7. Just a couple of questions before we tackle the problem. I see you are using fairly fast lenses. Were you shooting them wide open much of the time?
    What camera did you use?
    Were you using a tripod? I sincerely hope so.
    For graduation pictures inside you seem to have a very limited lens selection. You are fine for the full stage shots but for the individual students you really need the 70-200 F2.8 (vr) or such so you can isolate them during the presentation of diplomas. You will get much more appealing shots of the individual students that way. Assuming you are not on the stage with them you would be shooting 'up' pretty dramatically at around 80 mm using the 50. So I have to ask if you were using the 50 and then cropping like mad to get your closeups?
    You mention a 50th at F2.8. Even with a flash this is a pretty low number. Most would say that the minimum speed you should use at 50mm with the crop factor (assuming you are using a crop sensor) is a 70th of a second and I would be much happier over 125th. You are really flirting with motion blur as your subjects are moving and adding camera shake into the equasion. I would be willing to bet that this is the answer to your blur problem.
    Other people's flash should not be a problem except for the very rarest of occasions. If you can answer the questions and post a picutre or two we should be able to help more.
  8. Stage lighting sucks. I can't stand working in it because there will always be hot spots, usually in the worst places you could ask for them.
    Don't use program for flash photography like that. That's a good setting for interiors with low white ceilings but it won't work in your setting without looking like cr*p. Shoot a test frame in P, check your histogram and shutter speed/aperture values - if everything is exposed ok, put it on manual or Av and plug in those values. Then turn your flash on.. probably -1.5 FEC ish?
    On the equipment side - fast glass is good, but not for groups. Keep it above f/4 if you can.
  9. Here is a previous thread. Realize that there are professional photographers who specifically do graduation ceremonies. They have the gear and the organizational plan in place to know how to control and affect such things as parents, positioning of the diploma pass, and even stage lighting.
    One thing I can say for sure. I wouldn't know how to advise you because as John says, I wasn't there. This kind of situation, you need to call upon your knowledge of photography and experience in shooting in many different lighting conditions. Shooting weddings, I run across the spotlights being much brighter scenario a lot.
    I can say that manual camera mode would be the one to use, and you'd have to decide upon how to handle the lighting for the action that is occuring. Generally, if there are bright spotlights, you need to put your flash/camera EV at or above the brightness of the spotlights--then use the flash to 'fill in'. This is not necessarily an easy thing to do for an uncontrolled situation and does not result in nice looking photos sometimes--the black or blank background and 'flashy' looking images. However, sometimes it is necessary. This might be avoided with some pre planning and a talk with the lighting manager. Or the set up of off camera lighting, if allowed and feasible. Many times for graduations, it isn't feasible. However, if you are the 'official' photographer, you do have some say.
    On your main problem of soft images--probably motion blur. Secondly, hand holding shake. This basically goes back to the guidelines--there are two sets. The first one is for stopping different speeds of subject motion. The second one is for hand holding. If shooting with ambient as primary, these two sets must be followed or you will get soft images.
    Flash has the ability to freeze motion just by it's duration, but your ambient must be underexposed to allow this to happen. It is called dragging the shutter. I'd do some research into the concepts I just mentioned. In your situation, all of these things would come into play. For instance, for the presentaiton of diploma shot, you might drag the shutter, but for the wide shot of the stage, you probably need to go with ambient only. This is why I said above that you need to call upon your knowledge and experience based on what you are shooting.
    Plus, you might also have DOF issues, as David S. mentioned. I don't think anyone else's recipe is going to solve your problems. You have to think this through yourself.
  10. I doubt that you need a tripod and don't see what difference which camera you are using would make. As John & Nadine indicate above, it's hard to cover all the possible solutions and dynamics in a single forum thread.....feel free to do a search on flash and pay particular attention to balancing flash with ambient light. Marc Williams had an excellent thread several months back, that would be a good starting point.
    For myself, I stay with an ISO of 400 and would shoot at f/5.6 or f/8.0. Fast lenses are nice to have but when you are shooting at f/5.6 or higher, they make very little difference. I like the diffusion and spread that I get from using an omnibounce on my 580EX set at a 45 degree angle (see illustration). Ideally, I would have a second flash equipped with an omni set near my camera position but placed on a pole about 5-8 feet above my lens with the omni at a 45 degree angle but also feathered slightly towards the back end of the group and centered towards the rear area of the background (set this 1 stop hotter than your on-camera or on-bracket flash). The background is much more likely to go dark if there's alot of room between your subjects and the rear area of the church or stage. Occassionally you can compensate to some degree with digital fill while in the PS stages but try to get the lighting as close as possible at time of capture rather than relying on digital fixes. A dark background is much more acceptable if you've got good lighting on the faces of your's often more of an issue for photographers and peers than it is for your clients.
    I would also suggest that you study portraiture lighting and concentrate on the different ratios between the different lights, the basic light dynamics are the same. Good luck.
  11. Here's the illustration for the omnibounce that I mentioned above......the 2nd flash positioned higher above the camera would be feathered at approx. 30-35 degrees.
  12. This diagram isn't scaled of course, but I think it gives you a rough reference for what I'm trying to convey when using the 2nd omni with a high ceiling and attempting to throw more light in to the background and the higher subjects palced further from the lens.
  13. David, you seem to be a fan of the omnibounce diffuser but my experience with gymnasiums is that the ceilings and walls are too far away to render these things effective. They eat light and your flash recycles too slowly for the type of event the OP was shooting. Maybe in a smaller room they work where the walls are pristine white. (But under those conditions basic bouncing would suffice).
    And I've never shot a fast paced graduation but I've watched those "specialty" pros that do this for a living. It appears to me they are just using a lot of direct flash in ETTL. With the changing subject distances and varying lighting I tip my hat to those that can do this on manual flash power settings.
  14. For group shots on the stage I'd prefer to shoot at F5.6 to F8. Although the 2.8 lenses are great they can also have a shallow depth of field.

    Your flash should stop movement from anyone moving slightly. Without seeing the stage, but based on experience, maybe try setting your camera on manual at 1/30 of a second and if your are 20 feet away try direct flash at full power. If F5.6 is still too dark try F4 with an ASA setting of 800. My feeling is the flash on ETTL or TTL is only good to about 10 feet max. After 10 feet the ETTL light falls off causing the people to go dark. So you kind of have to go full power to light up the group from 20 feet.

    Are you going to use a short ladder or something? If you are below the stage shooting upwards the images most likely look a bit weird.

    As far as the stage lights go, think of them as fill, and over power the stage lights with your flash at full power. Stage lights can really cause frustration to a photographer, such as red and blue lights shining on the heads of people. If you are close enough usually your flsah will over power the the stage lights.

    You probably don't have any powerful strobes, but perhaps you can rent them from a local photo store. This is actually how I deal with stage problems. I will set up one or 2 of these monolights and fire them with a wireless radio slave. Your exposures will pretty much be dead on all of the time.

    Lastly, shoot in RAW so you can color correct bad lighting situations if needed
  15. If I were shooting w/ flash indoors in a situation like that I would avoid P mode and would more likely use manual mode and let the TTL flash get the correct exposure.
    However, based on the lens I was using and the amount of ambient light and the camera's light meter, I'd have an estimate on what shutter speed and aperture to use in manual mode.
  16. RT, I am a fan of the omnibounce, it doesn't lose anymore than one stop of light compared to direct flash and it does a nice job of opening up the background. I routinely use it in halls with ceilings over 30 feet high (see image below). Bouncing a speedlight into an umbrella cuts the power much more than using the omnibounce. When used correctly the omni does not eat light.
    The dual-light setup that I've suggested above would do a more than adequate job of lighting the subjects and providing some fill to the background in the situation described by the OP. I would likely shoot manual, f/5.6 at 1/60th............I've been a pro for more than 16 years and have shot more than 500 weddings.
  17. I disagree that the flash is going to stop motion blur in this case. The OP mentioned the hot spots that the stage lighting was causing. Under these hot lights the flash is not going to do the trick. Hand holding at 1/30th is a recipe for a disaster under these circumstances. The OP as much as admitted this. Then we have students walking briskly across a brightly lit stage........right?
    Why tell us what kind of camera? Well beause it will determine the max ISO we can recommend and the difference between crop sensor and full frame effects the speed at which we can expect to hand-hold the camera. (Yes I am aware of that some say it makes no difference.) I would not shoot a 50 mm lens at 1/50th with a 10D but I might with a 5D. I would not shoot at ISO 1600 with the same 10D but would not hesitate to do so with the 5DMark II.
    I have shot a bunch of graduations and other similar events. Newspaper folks do it all of the time. If he keeps his speeds up he will have better luck with stopping motion and reducing camera shake. Sharper pictures. Blasting hot-spots with more flash will not get rid of the hot-spot. It will just make it brighter. It is easer to deal with hot-spots in composition and post production.
  18. With hotspots, flash can still freeze motion, but the hotspot EV has to be about 2 stops below the flash EV, which is why I said above, that the resulting images would look very 'flashy' and the backgrounds would go black, dark. Not a lot one can do about hotspots but overcome/match them or eliminate them by turning off the spots. Sometimes going higher ISO makes hotspots and actual lights look worse because they become hot 'blobs'. This is definitely a tricky situation that automated modes are just not designed or equipped to handle.
    As for making hotspots brighter when flash is used--I'm sure you know the fill flash theory--the flash opens up shadows a lot more than highlights, so while you don't want to blast with flash, filling with flash on a hotspot can be workable.
    I just don't think we can give Jason much useful information, since the specifics are not known.
  19. thank you to all who have answered...
    i was using a 5D for most of the ceremony, occasionally a 30D. the choice of lenses i agree was not the best but my 70-200 is currently in the middle east due to a rather awkward situation. i had 8 hours to prepare for the shoot and i couldn't even get in the building more than an hour before the whole thing took place. i was definitely the last minute photographer, the original guy bailed more than likely because it seems he may not have gotten paid to do the work.
    i did try dragging the shutter a bit, i usually have good luck with this, have done it quite a bit, not so much this time. between over exposure and for some reason a lot of blurr i had better luck with the flash off the camera totally... i dont really know why but did. the only other thing i did notice was with a 24mm lens my E-TTL flash was saying that the focal length was 14mm, not sure what was up with that and i will have to figure that out.
    i asked about what i was allowed to do and was given permission to " go were ever you need to go to get the shot", i dont shoot these types of things so... i was pretty much up on sage with a 50mm, not really the best place in the world but...
    i am going to try and post a picture that i seemed to be having a lot of problems with... its a long distance shot...
  20. there is quite a lack of sharpness in the student area... i had the camera focused on them... still a major lack of datail.
  21. Unless you are shaky, it was subject motion or misfocus. I checked the exif, and this was with a 30D with a 24mm lens. You are right on the border, though of hand holding safety at 1/45th shutter speed. To rule out misfocus, check other parts of the image to see if you can find sharp areas. If it was subject motion, there should be areas of blur and others areas that are sharp. Not all the students are going to move as one. It is impossible to tell from an online image. Also, be sure you are not overreacting to the definition of head size on a cropped sensor camera at 100 percent magnification.
  22. "RT, I am a fan of the omnibounce, it doesn't lose anymore than one stop of light compared to direct flash and it does a nice job of opening up the background. I routinely use it in halls with ceilings over 30 feet high (see image below)."
    David, I am certainly not "critiquing" your use of the omnibounce. Indeed, I am genuinely intrigued by that fact that someone with your experience and talent can actually use this thing so successfully. I never could.
    Of course I never used them as a "pair" like you do. Regarding dimly light venues, I use a bounce card attached to my on-camera flash and while I find frontal subject exposure to be spot on via ETTL, I can't seem to get rid of that "cave" look as there really isn't much ambient to pick up. So if I understand you correctly, you basically drag a 12-15 foot "stick" around with an attached speedlite/omnibounce to throw light behind your subject?
  23. At this resolution I can't tell what is soft and what isn't. That said, another possible issue with "unknown" softness: the dreaded UV filter (especially wide open). If you had a UV filter on the lens, that alone could rob you of sharpness.
  24. a UV filter?
    i am not sure that i have the UV varieties on my lenses i believe they are just the "clear protective" sort.
    is that an issue as well?
  25. the above picture is what i was refering too... just a lack of detail. the school wanted these shots and it just didn't work unless they plan on using them no bigger than what i put up for the original picture i posted! they are an utter waste. but how would you go about keeping this from happening in the future if you were a young and inexperienced rookie like my self?
  26. Did you do the other checking I suggested? Just from what is posted, I would say misfocus or shake--does not look like subject motion.
    Get the trial version of Focus Magic and play with it. It might make the image good enough to print bigger. It will look a lot better printed than on screen (in the software), so print 100% crops as you go to check yourself.
  27. i believe they are just the "clear protective" sort.
    is that an issue as well?​
    YES! Especially on wide open glass. Take a look at this:
    Now you don't have to understand all that. Lord knows I don't! But the point is that lenses are finely tuned, precision engineered optical devices that bend and focus light onto your sensor. Anything you put in front of that can have a profound effect on how that all works. I am not suggesting there isn't a time and place for various types of filters See Thom's excellent article: What I am saying is that any filter will degrade image quality.
    Back to this image, I agree with Nadine that it will look better printed. I am not so sure I would call the shot a waste. It's an f/2.8 shot at quite some distance. If someone from the school actually called this shot a waste (also keeping in mind that these are the same people that apparently didn't want to pay for the shot), I would be a bit put off!
  28. RT, Occasionally I will use a second off-camera flash on a stand or simply placed atop a nearby table or chair. However, a single omni is capable of throwing alot of light into the background (avoiding the dark cave look) when used properly, remember it must be angled at approximately 45'll completely lose this light spread if the flash head is pointed straight on at 90 degrees and if it's pointed straight up (similar to a light-sphere) you will lose alot of light/power. You also need some distance from your subjects to allow the light to open-up. Here's a shot of the dance floor from the same room with the high ceilings that I posted above, all the overhead lights were completely off and besides my flash there were only the wall sconces lighting the room. This was taken with an omnibounce 580EX (set +1/3rd) on a bracket, f/5.6, 1/60th, at 400 ISO. The little bit of blow-out on the guy's white shirt was due to my post-work, the bride's white dress and nearby white table (closer to the camera) has no loss of whites.
  29. Sorry for the slight digression, but I couldn't let the following statement pass without a strong rebuttal:
    [John D]: "...What I am saying is that any filter will degrade image quality...."
    The question, of course, is how much any given filter will degrade IQ.
    The obvious and only real answer to this question is NOT by vague reference to a general article about lenses, or by quoting an "expert". Rather, it is by performing a trivial, controlled experiment that will take you at most a minute or two. Put your camera on a tripod, set it to manual focus and manual exposure, and then take one shot of a typical subject: first with a filter, and then again, without the filter under study. Unless you intentionally have sun or other bright light shining directly on the front element of the lens, my guess is that with any reasonable quality filter and any lens from 20 to 100 mm FL, you will be hard pressed to see any difference between the two shots. If a difference occurs, it will likely be a slight reduction in contrast with the filter, not a loss of sharpness. I know this because I have done this test many times over decades of experience in optics and specifically, photography. Of course, if you are shooting directly into the sun with normal lenses, or are using an $8000 500mm/f4 VR, yes, filters can cause a loss of IQ, but for most people, most of the time, the negative effects of filters are vastly exaggerated.
    If performing that simple experiment is too taxing, you might prefer reading about others who have performed closely-related, actual tests on lenses with severely damaged front elements:
    This is the sort of damage to the front elements it took to begin to see a noticeable loss of contrast in normal shooting situations: and
    Do the filter on-off experiment for yourself, and I'll bet you will change your low opinion of (reasonable quality) filters in normal shooting situations. Unless you like spending money replacing lenses, the pragmatic approach is not an all-encompassing generalization such as you offered, but is to protect your lens most of the time, but remove the filter when you know (from tests or experience) that it is likely to cause problems.
    Tom M
  30. William Michael

    William Michael Moderator Staff Member

    Just commenting on the image presented, my thoughts are, that there are many issues which are compounding making the image look “unsharp” (not in any order of Priority or impact):
    Tech Specs: EOS30D + EF24/2.8
    Shooting Specs: F/3.5 @ 1/45s @ ISO800, Available Light, Pattern, HH (Assumed).
    1. The lens: The EF24/2.8 isn’t the sharpest, you are pushing it at F/3.5. IMO, F/4 would be the absolute safe limit if one wanted an enlargement - F/8 would be best.
    2. The Tv: 1/45s hand held (assumed) – not for an enlargement – at 1/45s you are beyond the safe limit of hand holding (Camera Shake), for an enlargement. I “think” you have a little camera movement.
    3. The Tv: 1/45s is also just beyond the limit for adults sitting (at that distance) IMO (Subject Movement) – 1/60s is safe, IMO. I do NOT “think” you have Subject Movement
    4. The Tv: At 1/45s you are into the faster Tv where mirror slap might impact on the shot – for this shot use a Tripod and Mirror Up and Remote Release.
    5. Filter – no way – especially not for a wide angle lens, this quality WA lens and using it nearly wide open. A filter increases CA and losses sharpness, generally when the lens is wide open or near wide open.
    6. Filter – no way – using the lens wide open increase susceptibility to Veiling Flare which presents as loss of mid-tone contrast and general dullness or a smoky appearance – which is often misinterpreted as “un-sharp”. I do not think the EF24/2.8 is a very well baffled lens
    7. Focus – I am not convinced you nailed focus on the group I would like to see a 100% of: - the steps the six people in the front row centre and compare that to the 100% you have supplied.
    8. Post Production I am not convinced that your sharpening procedure is the best it could be, but I am not expert in this area, however please see FYI, (as much as we can do with web images - but I think there is a significant impact?):
  31. William Michael

    William Michael Moderator Staff Member

  32. William Michael

    William Michael Moderator Staff Member

    “Do the filter on-off experiment for yourself, and I'll bet you will change your low opinion of (reasonable quality) filters in normal shooting situations.”
    I agree, I have so done. I am a manic at testing things and knowing all the limits.
    I can make my 24L look atrocious using it with its $200 Slim line u-bute UV Pro filter on it . . . using it wide open or near to wide open in a lighting scenario with lots of multidirectional lights and or reflections.
    I am not taking a side with either you or John or making any reference to the article he linked to – I didn’t even open it up – I make reference to only your quote I have extracted above . . .
    I have no issue with any “reasonable quality filter” in “normal shooting conditions” . . . (and my addition) with “normal” lenses –
    Wide angle lenses and (some) zoom lenses and Kit Zoom Lenses are “special” in this regard.
    What I am stating is:
    - this is not a “normal” lens,
    - it is not a normal shooting conditions
    - the lens is near wide open
    all far from “normal”
    Stick a $25 UV filter on that 24/2.8, set the lens at F/5.6 and shoot with Front Key & Front Fill Light in a Pro Studio all day and you will never know the filter was there . . .
  33. I think there are a few things going on with this shot. Shutter speed was too low, Lens was wide open, Focus may be off and lastly I feel that 8mp is just not enough resolution to get enough detail into a shot like this in the first place. As an 8x10 once it has been shaprened it may look OK but I doubt these is enough detail to pull off an 11x14 or a 12x16. Back in the film days this would be a good example of when to shoot 6x7 or 6x9 medium format.
    I tried sharpening your crop and sharpended it looks a bit better.
  34. this is a section of the stairs and the people in crowd at 100%. they are in better focus i think but still not quite perfect.
    i do agree that with a bit of sharpening the print may look quite a bit better. i hadn't applied any sharpnening becuase my printer has suggested not doing that till we get together to actually print it at a certain size and on a certain media. but i do agree that it looks better.
    1/45th hand held is pushing it i know... i was in an awkward situation, still am, my tripod and some of my lenses are still in the middle east! so that has been a disaster. i borrowed own from a friend but his was a bit of a pain to use and use quickly so it didn't get used at all. the school wanted the shots but they didn't seem to want the shots really set up and controlled by me, they just said i could go any where i wanted to get them and they would take care of the rest.
    the 24 2.8 isn't the best lens. it does have a b+w filter on it, both those things could cuase some issues as well, but i know i have been able to get some really sharp results from this lens with the filter on, i think even at larger apertures.
    thank you all again.
    so use a tripod and a smaller aperture for the shot next time? or get closer and use lights?
  35. Most pros can handhold that lens at 1/45th but a tripod certainly wouldn't hurt. I still believe that by powering up the flash, even with this distance from the stage, that you can add alot more light and stop down to f/5.6. Looks to me that this shot was taken from a balcony area which should have given you enouh room to bounce the light from the ceiling and given you generous fill as I've tried to simulate in the before & after images below. Post-work with digital fill, sharpening and general tweaking can also improve this image.
    BTW, people around the forum love talking about the specs on lenses and other tech data, I'm wondering if you took the time to clean the lens just before you started shooting the job......
    One other suggestion. Look for a local pro studio or two that have covered some weddings at this location and see if you can identify or ask them about how they approached the same situation.
  36. The question, of course, is how much any given filter will degrade IQ.​
    And whether any such degradation is relevant to the effect of the filter. IE, I use a GND often. I know it will degrade the image, but the effect of the filter has an overall positive effect.
    The obvious and only real answer to this question is NOT by vague reference to a general article about lenses, or by quoting an "expert". Rather, it is by performing a trivial, controlled experiment that will take you at most a minute or two.​
    We will just have to agree to disagree. I could care less how it performs in a well controlled situation. My link to the lens was merely to outline how very complex lenses are. That by putting more glass in front of a lens you will degrade IQ. A point you agree with BTW; you disagree with how much! And "how much" can't be quantified since every interaction with light will be different. Will it always be noticeable? No. True story: friend shoot AAA ball (usually from the 3rd base line, on the field), with a Nikon 70-200. Prints up to 8x10. Some prints are tack sharp, others are just a bit soft. There was no discernible pattern. Literally, one shot would be tack sharp, the other wouldn't, and the focus would change (he knows what he is doing). Sent the lens to Nikon twice. Came back both times with a clean bill of health. I asked if he was using a UV filter, he was. I told him to take it off for one game. He did and every image was tack sharp. You never know how the filter will adversely affect the image until after you take the image. And by then it might be too late.
  37. We will just have to agree to disagree.
    - Sounds like a plan. :)
    My link to the lens was merely to outline how very complex lenses are.
    - I think the difference in POV arises because I'm a physicist and have led multi-person PhD teams in optics for decades. I know how complex they are, but I live in a causal and very predictable world when it comes to light. If my boss or my funding agencies felt I couldn't predict when a given optical system would have problems, I would have been out on my ear a long, long time ago. ;-)
    I could care less how it performs in a well controlled situation.
    - I misspoke when I said, "...take one shot of a typical subject: first with a filter, and then again, without the filter...". What I should have said was take enough on-off pairs of shots to satisfy yourself that you have amply covered your typical shooting situations, whether they be landscapes, studio portraits, various lighting situations, etc. You'll then get a very good idea when a particular combination of filter and lens tends to drop in IQ. I've found the results can be simply summarized for most normal lenses: Worry about loss of IQ when there is a strong light source either in the frame or outside the frame, but hitting the front element of the lens.
    True story: friend shoot AAA ball ... I asked if he was using a UV filter, he was. I told him to take it off for one game. He did and every image was tack sharp...
    - Sorry, but that anecdote just doesn't convince me because a huge number of things could have changed between the two games. For example, all too frequently I'll shot for a minute or two before I notice that I've accidentally bumped the little manual / single-shot / continuous focus switch into the manual position.
    It comes down each person doing their own cost-benefit analysis w.r.t. the use of protective filters. If I don't use one, the cost is the risk of permanent and cumulative damage to the lens (depending, of course on the hazard of the situation). If I do use a protective filter, the cost is possible loss of IQ, but since I know that the situations that provoke this are strong light sources hitting the front element of the lens, I can decide rationally to use it or not. With that in hand, I find blanket policies for their use or against their use to be ill-informed.
    Tom M.
    PS - Thanks to William W for the adding "reasonable lenses" to the list of caveats. I agree.
  38. William Michael

    William Michael Moderator Staff Member

    "so use a tripod and a smaller aperture for the shot next time? or get closer and use lights?"

    I think you are barking up the wrong tree looking for a "this" or "that" answer.

    The whole thrust of my points contained above is that there are several technical issues which are COMPOUNDING to make the result not acceptable.

    As we are now aware that there is NO Post Production Sharpening, I think this is a major player - and as you have confirmed that the area of steps and six people in the front row are sharper than the body of the Stage Group, the for this file I think correct sharpening and “Focus Magic” will enhance the image tenfold.

    As to what to do next time ? ? ? I reiterate - it is my opinion that it is NOT one single element you need to address ALL in the list mentioned.

    And there are now more things on that list. . .

    Considering that you have confirmed that the front area of the stage is sharper than the centre area and you also are of the opinion that you focussed on the main group:
    9. You should review your Focusing Technique.
    10. Test the Lens’s Focus
    Also check exactly what AF selections you were /are using . . .
    11. SPECIFICALLY were you using AI FOCUS?
    12. On WHAT exactly did you focus – Canon AF works on CONTRAST differences – so in this scenario the top of the line of wood and the piano would be a good contrast difference.
    13 Did you use Centre Point AF ONLY – it is the most sensitive.
    14 Is your Centre Point AF exactly inside the RED SQUARE on your 30D – on a 20D it is not – this is exacerbated when using a wide lens.


    Also not wanting to be greedy and have just “my list” I think the point about 8mp, made by Stuart Moxham is excellent and I am kicking myself that I did not include it in my original list . . . :)


    The point is (as I understand the question) you are disappointed in the quality of this image AS AN ENLARGEMENT - that is how I have approached the question.

    So the bottom line is every little element has a multiplicative effect . . . IF we are discussing the end result being an enlargement

    As another example . . .

    I agree that most Pros can HH @ 1/45s – if we want a completion I can get to 1/2s on a good day for a long shot at the Wide:
    And I can get to 1/8s for a Tight Shot at the telephoto:
    Both these images are pretty fine at 5x4 prints held at 15 inches . . . but the point is if we WANT an ENLARGEMENT all those “Hand Holding limits” go out the window . . .

    Back to your image here . . . (assuming that is a 10ft Grand Piano) You were about 100ft from the stage and that stage is about 2/3 occupied with 30 people wide.

    If you enlarge that full frame crop to a 10 x 8 print you have about 6 inches of people across the 10inch print, each person (wide) occupying about ¼ inch . . .

    If my rough mental arithmetic is correct . . .

    IF you had camera movement of only 0.2mm you will get a resultant blur of 1/16th inch shown across the bodies of those 30 people shown in the print – 1/16th inch is a perceivable blur on a 10x8 held arm’s length . . .

    and 0.2mm camera movement is not very much at all.


    PS: "but i know i have been able to get some really sharp results from this lens with the filter on, i think even at larger apertures."
    Yes I am sure you have - I bet the lighting was different, or the subjects were closer or you were not enlarging as much and the Tv was faster . . . and etc - all accumulative - no one element works alone.
  39. I'm seeing grain more that focusing problems. What was your ASA and your camera? If it's not your ASA setting my guess is the focus. I'm not seeing blurring from movement. I 'm thinking our camera focused on the wood in front of the people. Can you enlarge the wood for us to see?

    Some camera lenses need to be adjusted ever few years or so.
  40. While I agree with the warnings about filters, glass sharpness, etc. I'm assuming that you want reasonable facial detail so that on a banner enlargement -which is what student shots usually want- viewers can look and view a banner print closely and clearly recognize their friends:

    Assuming your uncropped post is completely uncropped and a 5D RAW image has 2500 pixels horizontal, then about half of the horizontal area is unused to the left and right of the students. Therefore there is 1200 pixels horizontal left to be shared by all the students horizontally about 27 students horizontal (don't count the second row for this analysis) that leaves 44 horizontal pixels per student. In this even division of student heads it includes space (about 2/3 space) between each student leaving about 14 pixels horizontal per student face. The software smooths (blurs) this some in your cropped view so the pixels aren't seen. If you considered the vertical pixel count you would have a similar number. Faces are taller than wide - so approximately 14 X 20 pixels per face. JPEG looses more and it looks like you are close to the limit of what you can expect. For most of the students I don't see relative motion blur, so that's not your primary problem. Also wide angle lenses reduce camera motion blur as I'm sure your aware.
    The lighting situation is very apparent although no where near as difficult as stage plays..

    If it were me I would blow the dust off my LF camera and shoot this shot on 4x5 where there is sufficient effective resolution. Also the latitude / dynamic range of negative film is such that with good dodging/burning under an enlarger the lighting variances can be resolved. However I would NOT want to print more than a couple enlargements this way and I personally haven't ever made an exposure equalizing negative to sandwich in the enlarger to make a multitude of identical analog prints. (I've seen this done but it gives me a headache jut thinking about it.) A digital camera does make an effective preview device when mounted to a LF camera by the way.

    A better answer I suggest is to take the wide angle shot like you did as a reference. And then quickly shoot a multitude of telephoto pictures of the students from left to right. After the fact use Photoshop, PaintShop Pro, Gimp or whatever to reassemble the clear faces into the wide shot. Correcting for lighting and any blinking eyes/rolling eyes/unbecoming expressions at the same time. Result: about a 100MP image ready for printing.

    I know what you are thinking - this is going to take about 3 hours to edit. You are also thinking - I don't have to do this for my wedding shots. Intuitively a 17"x20" portrait enlargement with a couple of heads in a typical wall mounted enlargement is not going to have the viewer hovering over it looking for pimples etc at a 3" viewing distance.
    This is different, the main subject is a bunch of small heads and you are in fact going to have the viewer looking ever closer at a 30" wide banner print.
    You don't get something for nothing so if you want a masterpiece lots of extra effort is needed.

    If you have never done that sort of photo-editing before add about 8 to 10 hours of learning curve time. (But once you know how to do it -it becomes easy if not tedious), You are not done until the viewer does not know it was edited. Sorry to say you don't get something for nothing.
  41. William Michael

    William Michael Moderator Staff Member

    Terence -

    I have read your post twice.

    The first read I thought your idea was shooting a panorama and then stitching it – but it is not that - is it?

    So . . . when the faces are pasted back into the original wide shot - are you suggesting that there could be "blocks" of several faces pasted as one section - if the lighting was OK and the expressions OK?

    Would you suggest about a 100mm lens (on APS-C) ? ? ?

  42. Shooting a panorama by stitching is not what I meant but the process is similar. Stitching can be as simple as finding a straight line where each pair of overlapping images merge and joining them there eliminating the overlaps. It could be a straight cut or a percentage blending between the images. The viewer will usually detect a straight cut but the blending method works best. In fact some of the intermediate point and shoot digital cameras have a panoramic algothrithm built in that locates overlap in sequential images automatically and blends them producing a large jpeg from a series of shots. No Photoshop involved. I recently saw the result from such a camera when I was in a Walmart and an amateur was picking up a panoramic print and proudly showed it to me. A little Jaw dropping considering I know how much work is involved with Photoshop. It was because the womens 10 year old daughter read the camera manual while on vacation and without prior photographic experience (the first day she ever took any kind of photo from what I was told) shot a series from a hotel balcony and the result never went thru a PC - straight from the flash card at the kiosk.... I couldn't detect any seam.
    A 100mm lens should get you the detail needed. Tripod mounted of course since now camera motion would become the limiting factor. I wasn't thinking about substituting more than one student at a time but actually that should be a time saver and quite doable. I think that where the blending occurs should be easy behind the second row of students but following contour lines on the robes fold, for example, on the first row would yield the best results. I believe you can get away with this in this kind of situation because the lighting does not change and the students don't move much between your shots. I wouldn't want to consider this method if those facts weren't true.
    I wish a simpler answer were available.
    Even if your camera had a 100MP sensor the other concerns expressed by the other posters would become a major problem with a single shot if everything else was the same. Subject motion, lens sharpness, filter degradation, etc. It is always what is the weakest link.
  43. Jason -
    note that the portion of the photo that includes the stairs and the back of persons heads that you had a link to does not actually appear to be sharper - its just closer to the camera -so the features are all larger. Compare the head sizes to the students head sizes..
  44. William Michael

    William Michael Moderator Staff Member


    Thanks for coming back with the detail. You have confirmed the process I had in my head.

    Thanks for the tip - in the meantime I have been playing with my own testing of the technique - I used my 100/2.8 macro shooting an Orchestra on a Stage (before they began playing).


Share This Page