50mm f/1.8 - DPChallenge Forums

Tamron AF 28-300mm f/3.5-6.3 XR Di LD Aspherical IF for Canon

04/29/2006 04:48:51 AM · #26

I took some pictures in my 'apartment' this morning. Yes, my apartment is very open... These pictures are 100% crops. All taken tripod mounted.

I took a number of other test shots and was not able to get better results.

The subject was a little farther away, maybe 15-20 feet.

These are crops from the 30D with the 50mm.

Notice how the third of these images is sharp, but it's nowhere near the center focus point.

then my grid shots.

The second of those two was taken at f/1.8, 1/6400. All were taken with self timer.

This one was with a different lens, the 85mm f/1.8

It looks a lot better, but I think that this is probably because of the increased magnification, which might have made the difference and made it easier to focus. The grid shots were VERY difficult for the camera to focus on.... maybe they were a bad choice because the camera could see through them? Maybe something to do with the regularly repeating patterns?

Oh yeah, I put some diagonal crosshairs in the pic so you could see the center where I focused. I used single point AF, centered.

Message edited by author 2006-04-29 11:01:31.

lfordhere

Canon EOS-20D

04/29/2006 05:58:52 AM · #27

It looks like the camera is lens is back focusing, which is discussed in one of those links I put up there--aside from the grid photos too... Look at the two at the top of the previous entry to this one and see how everything behind what would be the subject is in focus, yet the foreground is out of focus.

deapee

Kodak V803

Nikon AF Nikkor 50mm f/1.8D

04/29/2006 09:58:51 AM · #28

I don't think there's anything wrong with anything you've posted.

Tamron AF 28-300mm f/3.5-6.3 XR Di LD Aspherical IF for Canon

04/29/2006 10:40:35 AM · #29

Thanks Lford. I did check the links but only one of them worked for me. One was a related article (Rob Galbraith) that was discussing the 24mm Tilt Shift lens, the other was not available.

Which one did you see that discussed this issue?

I am having trouble with the concept that this one might be user error because I haven't read the article.

I have other shots that I took with the 50mm of kids that were a bit farther away that suffered the OOF look too, but I guessed that this was user error and threw the pics away... I am not so sure anymore...

No offense Dave, I've got a lot of respect for you, but how could you seriously not see anything wrong with those pictures? They are badly 'backfocused' as Lford says and the target that I focused on - and was told had a lock on - is NOT in focus.

lfordhere

Canon EOS-20D

04/29/2006 10:40:50 AM · #30

Originally posted by deapee:

I don't think there's anything wrong with anything you've posted.

Look at the last two grid shots on this page, the two that almost fill the frame completely, and you will see a haze on the grid from the lens in question, yet sharpness from the other lens [the last photo]

Tamron AF 28-300mm f/3.5-6.3 XR Di LD Aspherical IF for Canon

04/29/2006 11:00:35 AM · #31

Lford. I ran a quick double check on the links you posted and I can't find anything about backfocusing.

Is this a problem with the lens or the camera? Or what kind of problem is it? How can I fix it?

The camera is very new, going on it's second week and I would really like to have it sorted. I've got a lot of learning and a friend of mine who is an ex-photog has offered to pull a contact on a model he knows that is willing to throw us an afternoon just for 6 prints for her portfolio. He tentatively set the 'date' for next Saturday.

I'll probably have to buy a new lens by then too...

lfordhere

Canon EOS-20D

04/29/2006 01:38:32 PM · #32

Originally posted by eschelar:

OOPS! I must have read it on one of the other pages I looked at, but forgot to include the link... Here ya go:

Link 1

I just searched through Google to find links related to this type of issue and then read through the links... Backfocus seemed to stand out at possibly the problem you are having, but it certainly could be something else too.

Message edited by author 2006-04-29 13:41:51.

Tamron 70-180mm F/2.8 Di III VXD

04/29/2006 09:59:49 PM · #33

So if it is backfocus, it is probably only with that lens?

If so, is it simply a 'send it in and get it fixed' issue?

MadMan2k

Sony Alpha a6400

04/30/2006 12:07:43 AM · #34

A little OT, but it seems like my 70-210 will frontfocus a little when focused between maybe 8 and 15ft away. It works pretty well near the close limit, and further then 15ft or so, but that range is almost always off, so I have to shoot at a higher aperture or focus a little forward manually. It's much easier to set the focus on a telephoto than on the 50, as I'm sure you've found out, though.

Nikon AF Nikkor 50mm f/1.8D

04/30/2006 10:46:20 AM · #35

Nobody has commented that I should do anything to fix this issue myself, so I will be taking my camera in tomorrow morning... if they are open...

Grrr. It's Taiwan's Labor day, but I don't get a day off?

Maybe Tuesday.

Frustrating.

We have arranged an outdoors photo shoot if the weather is OK with a couple of models.

I will be picking up a reflector. Sadly, I don't have the funds right now to get a Sigma flash unit quite yet.

I hope they can fix the camera quickly.

deapee

Kodak V803

04/30/2006 11:11:20 AM · #36

My only problem is that the 50 1.8 just wasn't designed to shoot at 1.8 ... well, it was, but it wasn't designed to be super sharp and accurate at 1.8.

The fact is that if you're backfocusing, it will backfocus no matter where you're focusing, not just when you're focusing further than a set distance. If it's having the problem at 40 feet, it should have the same problem (but be even more magnified) at 3 feet.

If you can take a photo at a 45 degree angle of a tape measure, and focus on the 6, and the 6 is what's in focus, then I'd say you're golden...of course, this should probably be done outside on a sunny or overcast day with good light, not in your house under a 60 watt tungsten bulb as, again, neither this lens, or your camera are going to be all that accurate under these conditions.

--

I don't look at any of my pics at 100%, they're always rather dissapointing to be honest. Try taking a shot with auto focus, then try manually focusing at a specific point, and compare the shots -- are they similar?

--

But no, I don't think it's anything you can fix yourself, unfortunately. Sorry for being so brief before, I just really don't think it's an issue if you are getting accurate focus up close. Although without seeing it myself, it's kind of hard to tell. Good luck though man...hope they can fix whatever it is that's wrong.

04/30/2006 11:25:26 AM · #37

Thanks dave. Please note that I am experiencing this issue with the 50mm lens all through the aperture range. I repeated my tests at several points in the aperture range, both in the original hand-held shot with the flowers (when the shutter speed got low, I put the camera on a solid wooden post to prevent camera shake).

I should have been able to get something sharp out of autofocus on a highly detailed area.

My closer shots do not seem to be affected and I wonder if perhaps this is also an issue with my other lenses which I have not shot at larger distances.

I am just hoping to understand the issue better before I take it in to the shop. People in this country have a TERRIBLE habit of saying 'there's no problem' without thinking.

Now that I look closer, the filter that I bought for it, the store owner said it was multi-coated, but looking closer (I bought it in a bit of a hurry as the stores were closed and I had pictures to take), nothing on it indicates that this is the case.

However, I did repeat the tests with and without the filter on the lens, so I've pretty much ruled this out of the equation.

I believe I will attempt to get my money back on that filter tomorrow or trade up to the B&W filter. I don't appreciate being lied to.

Very irritating stuff.

The pictures were taken outside on an overcast day. Light was EASILY sufficient. I am fortunate enough to live on a rooftop where I have pretty much the whole place to myself. Yeah, the entire open rooftop with two covered areas.

It's nice and a half in the summer with my hammock. And it's great for natural light shooting.

I honestly don't see that the test that I carried out with the high detail wire mesh (which contains both vertical and horizontal lines, suitable for making it easy on the center AF point which is sensitive to both axes) is any different than a measuring tape test. Look at the wire mesh pictures. At f/1.8 or at f/4 (both of which were used on several shots, not all of which are shown here), there should at least be SOME indication of sharpness. But if you look closely at the pictures, you will notice that the GROUND, well behind the mesh is quite sharp. This does not appear to be an issue of anything more than focus.

I didn't raise this issue before because I assumed that this was just softness typical of 100% crops. I don't really think that this specific issue is.

Again, this issue only seems to show up once I pass a certain distance.

I've played with a number of other DSLR's and I don't think this is typical performance.

fotomann_forever

04/30/2006 12:10:31 PM · #38

Have you tried MF vs. AF? Let's see if we can rule out the AF system as being at fault.

Sigma 20mm f/1.8 EX Aspherical DG DF RF for Canon

04/30/2006 09:29:45 PM · #39

I'm not sure that that would help anything at all. Being as this is my first DSLR, I have approximately zero experience and confidence in my manual focusing abilities.

The way that I understand that MF works is that you manually slide your focus until the camera beeps, saying that it 'agrees' with your decision.

This is supposed to be far more accurate than just using your eyes for an arbitrary level of 'sharpness' through a rather small viewfinder.

Further, it also adds reliance on the AF accuracy to confirm your decision.

I wouldn't trust my manual focusing abilties as far as I could throw them.

I believe that a significant part of my issue relates to using the lens at a point where the DOF is quite deep. This adds a fair bit of 'uncertainty' to the equation as there is a larger room for error. Both with AF and MF.

I am just not sure whether the camera should be able to handle the situations I am giving it normally.

I don't believe that this is an optics problem because there are portions of the frame that are definitely sharp.

Message edited by author 2006-04-30 21:31:09.

yido

Canon EOS-5D

05/01/2006 03:06:05 AM · #40

Remember shooting a plant can throw off the focusing. You don't know if it's focusing on the background behind the leaves or not. I also noticed that the net isn't all equal distance from the lens. If you use a wide aperature, DOF is shallow and an irregular subject may have some parts in focus and some parts OOF focus.

Now you know why shooting brick wall is so popular. Things on the corner and the center are about equal distance from the lens hence if the focus is correct, everything in the shot should be in focuse. Shooting a scene with things at different distance from the sensor will not aid in detecting if the lens is sharp or not especially if you use wide aperatures.

MadMan2k

Sony Alpha a6400

Tamron 70-180mm F/2.8 Di III VXD

05/01/2006 03:41:28 AM · #41

Originally posted by eschelar:

The way that I understand that MF works is that you manually slide your focus until the camera beeps, saying that it 'agrees' with your decision.

I've never heard of that way.. To manual focus, I've always just turned the focus ring until the part I wanted to be in focus, was. Sometimes you mess up, that's what AF is for (assuming it works :p)

deapee

Kodak V803

Nikon AF Nikkor 50mm f/1.8D

05/01/2006 08:08:58 AM · #42

Originally posted by eschelar:

At f/1.8 or at f/4 (both of which were used on several shots, not all of which are shown here), there should at least be SOME indication of sharpness.

See, the thing is though if you're shooting at 50mm, focusing about 20 feet away at f/4...you're still getting about 13 feet of decent sharpness. I don't see an area in those pics that's really that sharp...which is why I think it's just the way it is.

--

Try a shot at f/16...focus about 20 feet out and see how your 100% crop looks...make sure you're using some kind of tripod though. Everything from about 11 feet to infinity should be acceptably sharp.

--

And as far as the test of the mesh being accurate, I mean it is, but it isn't, because there's a lot of factors involved...wind, atmosphere, reflection, etc, etc...

Take a tape measure, focus on the 6 at a 45 degree angle in autofocus...and see what you get -- shoot at 1.8. If it's backfocusing, you'll know it right off the bat.

As far as manual focusing, just switch the camera to manual, twist the ring until what you want is in focus, and shoot...I've never had my camera beep or anything when I manually focus, and about 95% of my shots are manual focus.

05/01/2006 12:06:33 PM · #43

Interesting stuff about the manual focus. I don't use it because I don't trust my eyes as much as I trust the AF. Perhaps that is part of the issue...

All of the net shots were taken on a tripod, as was the green plant shot without the flowers.

I always understood that it was best to shoot a subject with high detail. Now I am trying to shoot more subjects that have moderate detail.

Thanks your guys for continuing to help me out on this one. This is definitely an area where experience counts.

I took the lens and camera to the shop today. The official Canon store was closed, so I started at the store I bought the camera at. I ended up buying the 50mm lens at a different store because I had to buy it on a sunday in the evening and I just missed the other store... we have a lot of camera stores here...

Anyhow, they ran some tests for back/forward focusing and ended up finding that the camera was FORWARD focusing at a certain length. They found this ONLY with the lens that I have already. They used two other lenses and found that they did not have this problem.

However, they could not replace my lens because for some odd reason I didn't have the warranty card in the box... stupid me, trying to be organized and keep all my warranty cards together... They suggested that I go to the store that I bought the lens at and boy am I glad I did.

They repeated a similar test, although with more 'real world' type conditions. We discovered that the camera was experiencing troubles with other lenses as well.

They ruled out the lens issue, but then I ended up with the opposite story.

To be honest, I put a bit more stock in their results as I was able to work them out myself.

When I went back to the first store though, they were kind enough to load the test shots they took with my camera and other lenses (with lens serial numbers listed) and I will take them down to the Canon dealership tomorrow. I am really hoping that they are open this time... bleh.

So after a big round circle, I'm kinda at square one again.

I'm leaning towards a fault with the camera's AF system though. Doesn't all the literature state that AF is generally able to get within 1/3 of the DOF? Heck, that's why I forked over an extra 600 dollars to get this over the 350XT.

These results are a ways off in my books.