How do you test a new lens to make sure it's not defective in some way?

I'm brand new to photography. Bought my first DSLR a month ago (K10D - love it) and just received my new Sigma 17-70 a couple days ago (love it, too).

I have two questions:

1) How do you know if you have a good production lens or a "lemon"? I have read that not all lenses are created equal.

What do you look for? If there is a problem, is it pretty obvious or does it take a highly trained eye?

2) Are jagged lines (e.g. ship's rigging) due to the monitor? In trying out the lens, I took some photos of ships' rigging, and many of the the lines, especially the ones against the blue sky, look almost like dotted lines in some cases. At best they are a bit jagged and not smoothly distinct.

I haven't printed any of the photos --- all I've done is look at them on my two computers. I've used two different monitors (LCD and and CRT) and it is noticeable on both. Also, it seems to be worse depending what photo browser I use. For example, the Pentax Photo Browser seems to have the problem more than the Windows Viewer.

The photos are JPEG at highest resolution.

Thanks in advance for any help. (This is a superb board!)

I am going to go out on a limb and say that those dotted lines are a result of viewing a photo at less that 100%. When you view photos at a reduced size without having actually resized them, your system has to make some quick and dirty assumptions and it doesn't always look great.

As for testing lenses, its a matter of preference, I guess, but I have never tested a single lens or camera and I don't believe that I am an worse for it. The purpose of my lenses is focus an image of the world onto my sensor and that is what they do. If in the course normal activities I noticed that there seemed to be something wrong with a lens, in either construction or output quality, I would look into it, but I don't see the sense in looking for problems.

--



Gallery: http://www.pbase.com/jess_ulm
Photolog: http://photos.apt131.com

TampaJack said:

I'm brand new to photography. Bought my first DSLR a month ago (K10D

• love it) and just received my new Sigma 17-70 a couple days ago

(love it, too).

I have two questions:

1) How do you know if you have a good production lens or a "lemon"? I
have read that not all lenses are created equal.

What do you look for? If there is a problem, is it pretty obvious or
does it take a highly trained eye?

Click to expand...

TampaJack said:

2) Are jagged lines (e.g. ship's rigging) due to the monitor? In
trying out the lens, I took some photos of ships' rigging, and many
of the the lines, especially the ones against the blue sky, look
almost like dotted lines in some cases. At best they are a bit jagged
and not smoothly distinct.

I haven't printed any of the photos --- all I've done is look at them
on my two computers. I've used two different monitors (LCD and and
CRT) and it is noticeable on both. Also, it seems to be worse
depending what photo browser I use. For example, the Pentax Photo
Browser seems to have the problem more than the Windows Viewer.

The photos are JPEG at highest resolution.

Click to expand...

Beware: you are entering the realm of pixel-peeping.

For both exercises, make sure you are looking at 100% (actual size). At this magnification, 1 camera pixel = 1 screen pixel. Any other resolution and you are evaluating the interpretive algorithm of you browser. Also, the LCD monitor should be at its native resolution or else a similar problem will occur.

Look at several different kinds of scenes. If everything looks sharp and you are happy with what it looks like, then stop there. If you see something questionable, post it on the forum and you'll get plenty of opinions. There aren't many bad copies of this lens floating around, so you may just want to shoot and not worry about it. Especially if nothing has come to your attention yet.

For the sail rigging, try looking at this at 100% as well. The effect is probably just a viewer artifact from the resizing. I bet it would look fine at 100%. If it does, it will print fine as well. Occasionally, the effect will occur at 100% as well. This is the result of of phenomenon know as aliasing, where the line is so faint that the square grid of the sensor doesn't know how to interpret the correct color for the line.This isn't a lens problem and can be caused by a lens being TOO sharp for the sensor. It's no big deal, and a nice problem to have most of the time.

--

Judging a photographer on the basis of equipment is like speculating one's physique from a gym pass.

Hi,

I also have a Sigma 17-70. My first shots already seemed to be less sharp on the right side than on the left side. I tried not to pixel-peep and tended to search for a mistake at my side rather than the lens' side. During the time I realized that with landscape shots at infinity still very often the right third of th pics were significantly softer than the center and the left side.

Later I did a comparison of the Sigma 17-70 with the DA* 16-50 and posted it here:
http://forums.dpreview.com/forums/read.asp?forum=1036&message=24339683

The reactions revealed that my copy of the Sigma was obviously severely decentered.

If I would have tested the centering of the lens right after purchase - and returned it immediately, I would have had many more satisfying images.

So, be careful with those experienced people who tell you "just take photos and beware of pixel peeping". If you are experienced and do know what to expect from a lens, then this method works fine. If you are uncertain because you are a newbie to DSLR photography, two successful tests will help you to be happy with your equipment, and to know quite for sure, it is you who is responsible for shortcomings and not the lens or camera

• the centering test (cf. above) and
• the auto focus test (c.f. http://www.focustestchart.com/chart.html )

Both tests together require about one hour work.

It is worth taking this time - and afters concentrating on taking pictures. At least, this was my experience.
Yours
Belenus

--
Regards
Belenus

When I get a new lens I generally perform 2 tests with it.

1. Focus chart test. You can find it on the internet. I repeat the focus test a couple of times to verify results. Make sure you do this in good light. View results at 100% of your computer screen and determine if the focus is accurate.

2. Newspaper test - I generally grab the Business section with the stock reports (smallest print of the newspaper) and either hang it vertically or lay it down on the floor. Shoot it so the newspaper fills the viewfinder with a tripod. View results at 100% on a computer monitor. Compare the 4 corners of the images for sharpness, contrast & overall clarity. Ideal results are when the 4 corners look the same. If one or two of the corners differ from the others you have a bad lens.

--



Learning to light at http://www.Strobist.com

http://www.trappedlight.com
http://groups.yahoo.com/group/LAShooters/

Another quick, cheap and dirty lens test that I like is the "brick wall test".

Just photograph a large symmetrical brick wall. (Most brick walls are symmetrical.) Photograph nothing but bricks and mortar. (That is, no windows or doors or anything else.)
Then examine the image carefully. Are the edges in focus as much as the center?

Are the edges as sharp as the center? Can you see the individual grains of sand in the mortar everywhere?
Any chromatic abberation at the edges of the picture?
Do left and right, top and bottom edges appear the same?
Can you see any differences between the center and left, right, top, or bottom?

--
'Now I know what it's like to be high on life.
It isn't as good, but my driving has improved.'
== Nina, on 'Just Shoot Me', 13 Jan 2006.

What great advice! Thanks all.

I looked at the sail rigging photos at 100% and the lines look just fine. And looking closely at all sides and corners, I couldn't find any visual evidence of a lens problem or inconsistency.

But I will try the other tests asap. The brick wall idea and newspaper ideas are both easy to do, and I'm pretty sure that if the lens passes those two tests, it's good enough for me.

I have to admit, I'm a little leery of getting TOO caught up in tiny details because I don't want my affinity for my new lens to drop for a flaw that few people would ever notice. I am comfortable with the idea that a $350 zoom lens will always have a minute flaw or two under the best of circumstances.

I just want to ensure there's no major flaws. I can supply those on my own! LOL.

Here's my method of testing a lens:

1. Shoot RAW.
2. Use a tripod.
3. Find some subject like a brick wall as your test subject.

4. Position the camera and tripod at a distance of about 50x the focal length. For a zoom lens, this means moving it several times.

5. Test apertures from wide open through F11. After F11, diffraction reduces image quality in all lenses.

6. Import the images into your image processing software without making any adjustments to them.

7. Evaluate images at 100%. It is best if you can compare a lens against some other lens of the same focal length.

Joe

The jaggies you see inwill viewing your photos might be due to viewing less than life size. Try viewing at 100% and see how it looks.

For a new lens I try to avoid shoting charts and brick walls. I usually go out and start shoting a lot of photos undera lot of different situations including:
1. At close focucing distance
2. At infinity distance
3. At mid distance
4. Low contract lighting
5. High contrast lighting
6 Use all the f-stops
7. If zooms the max, mid and min focal lengths.

The different focus distances to see how the auto focus works (front or back focusing problems) It also will show the the brokeh of the lens (how it renders the out of focus areas (creamy is good))

Low contrast to see how the micro contrast of the lens is.

High contrast to check on the purple fringing.

I will then pixel peep (whole phot but really look hard at the center and corners) those shots to see how the lens performs. Since I don't normally shot charts and brick walls I don't really care about how they shot those subjects. What I care about is how it performs in real life.

Dave

Jessica, you just poured ice water down the backs of all the pixel-peepers who inhabit this forum...

BTW, for the most part I agree with you. The best way to test a lens is to take pictures with it and see if the results satisfy you. That said, I do have a set of "standard' subjects which I shoot when I check a new lens to be sure it's not a dog. I think that only two or three of the more than 200 lenses I currently own were in that category.
--
Jim King - Retired Colormonger - Suburban Detroit, Michigan, USA; GMT -4h (EDT)
Pentax user for over 45 years. Photo gear and collection listed in my profile.



* * * * *
A fanatic is one who can't change his mind and won't change the subject.

• Sir Winston Churchill

* * * * *
The difference between genius and stupidity is that genius has its limits.

• Albert Einstein

Jim King said:

Jessica, you just poured ice water down the backs of all the
pixel-peepers who inhabit this forum...

Click to expand...

What can I say. I m a pixel peeper when it comes to checking out a new lens or a new camera. Maybe it is because I am old (54 years old) and and my eye sight is not like it use to be. Also I am cheap and don't want to print out a bunch of 11"x14" prints (size of my typical enlargements).

Jim King said:

BTW, for the most part I agree with you. The best way to test a lens
is to take pictures with it and see if the results satisfy you. That
said, I do have a set of "standard' subjects which I shoot when I
check a new lens to be sure it's not a dog. I think that only two or
three of the more than 200 lenses I currently own were in that
category.

Click to expand...

You should know about Pentax lenses. You have the best collection I know of.

Dave

Doesn't a flaw have to be pretty blatant in order for the retailer (or manufacturer) to take it back?

A couple of the posters recommend some pretty thorough testing. Is it realistic to tell a manufacturer or reseller that "at F11, in low contrast lighting, in the 50-60mm range, the lower left quadrant seems a bit soft"? Will they actually replace the lens for you, or will they just politely laugh?

It seems to me that it would have to be a blatantly obvious flaw in order to get some attention. Wouldn't that type of problem tend to show up under almost all settings, meaning that a couple basic tests will reveal it?

The brick wall is all blurry!



Sorry, had to do it.

Not a pixel peeper but would obviously agree to test lens if not satisfied with my image results.

Be well,

AZ
--

'There is an art, or, rather, a knack to flying. The knack lies in learning how to throw
yourself at the ground and miss.'Douglas Adams'.



Images at: http://www.pbase.com/frumsimchasphoto
Events at: http://frumsimchasphoto.com

Hello alinla:

I would like to add one more test to yours...Basically your newspaper type test but only at infinity...Some call it a brick wall test...Any scene with a flat plane of focus will do...The reason being I've found some lenses will focus OK close up but not at infinity or vice versa...

Pixel peepers Rock On...

LW

alinla said:

When I get a new lens I generally perform 2 tests with it.

1. Focus chart test. You can find it on the internet. I repeat the
focus test a couple of times to verify results. Make sure you do
this in good light. View results at 100% of your computer screen and
determine if the focus is accurate.

2. Newspaper test - I generally grab the Business section with the
stock reports (smallest print of the newspaper) and either hang it
vertically or lay it down on the floor. Shoot it so the newspaper
fills the viewfinder with a tripod. View results at 100% on a
computer monitor. Compare the 4 corners of the images for sharpness,
contrast & overall clarity. Ideal results are when the 4 corners look
the same. If one or two of the corners differ from the others you
have a bad lens.

--



Learning to light at http://www.Strobist.com

http://www.trappedlight.com
http://groups.yahoo.com/group/LAShooters/

Click to expand...

Larry,

That sounds good. Except the newspaper test is to determine decentering of the lens elements not really a focus test. I usually cover focus testing using other tests.

--



Learning to light at http://www.Strobist.com

http://www.trappedlight.com
http://groups.yahoo.com/group/LAShooters/

TampaJack said:

Doesn't a flaw have to be pretty blatant in order for the retailer
(or manufacturer) to take it back?

Click to expand...

Yes.

TampaJack said:

A couple of the posters recommend some pretty thorough testing. Is it
realistic to tell a manufacturer or reseller that "at F11, in low
contrast lighting, in the 50-60mm range, the lower left quadrant
seems a bit soft"? Will they actually replace the lens for you, or
will they just politely laugh?

Click to expand...

Most likely

TampaJack said:

It seems to me that it would have to be a blatantly obvious flaw in
order to get some attention. Wouldn't that type of problem tend to
show up under almost all settings, meaning that a couple basic tests
will reveal it?

Click to expand...

Maybe.

After giving a new lens a real world workout and then pixel-peeping to see any faults and hopefully more important how the lens behaves in situations you will use it in. If the real-world workout reveal any potential problem then I might result to shoting a brickwall, newpaper or a chart. Just to have something to give the dealer whn I return the lens. So far I haven't had any initial problems with my lenses (DA18-50, DA50-200, DFA50, DFA100, FA77, FA31, DA21, DA14 and DA*50-135).

Dave

TampaJack said:

I have to admit, I'm a little leery of getting TOO caught up in tiny
details because I don't want my affinity for my new lens to drop for
a flaw that few people would ever notice. I am comfortable with the
idea that a $350 zoom lens will always have a minute flaw or two
under the best of circumstances.

Click to expand...

This is even true for expensive lenses - as I had to learn. Klaus at photozone ( http://www.photozone.de/8Reviews/index.html#pentax ) once wrote that there is virtually no lens that is 100% perfectly centered. But some are obviously worse than others and not acceptable. Slight differences in corner sharpness are within the tolerances and are nothing to care about in practice.

Furthermore lens design is always a compromise between many factors like:

• distortion,
• vignetting
• zoom range
• speed
• center sharpness edge and corner sharpness (wide open versus stopped down and wide end versus long end)
• CAs and purple fringing (some Pentax lenses suffer from CAs - which can be removed by post processing but are great in other respects)
• auto focus accuracy
• manual focusing comfort
• closest focusing distance (close up or macro)
• bokeh
• contrast
• color rendition
• a n d weight and bulk
• a n d price

Some of these criteria contradict others to a certain degree and no lens can be perfect in all respects (just compare the DA70 limited and the FA 77 limited at Photozone and in the forum discussions)

I write this because I had been initially disappointed by the DA* 16-50, having had the too high newbies mis-expectation that this rather expensive lens should be excellent in many respects (e.g. corner sharpness) already wide open.

TampaJack said:

I just want to ensure there's no major flaws. I can supply those on
my own! LOL.

Click to expand...

This is the adequate attitude
--
Regards
Belenus

taged

Just want to thank everyone for taking the time to respond with such detail and care.

I bet this thread will be of use to many people for a long time to come.

Hi,

just to add a small hint to the very right things said above:

When doing the centering test with a newspaper or brick wall, make sure that your camera is pointing exactly perpendicular to the subject. At small distances, wide apertures and/or large focal lengths, DOF can become very narrow (less than one cm in some cases), so any deviation from 90 degrees angle (horizontal AND vertical!) can cause out-of-focus areas possibly mis-interpreted as decentering.

Perfect 90° angles are difficult to obtain- I once used a mirror for this. Draw a small cross on the mirror and point the camera on a tripod towards it. Make sure the cross is exactly in the middle of the viewfinder and the image of the lens is symmetric around it. Then your camera should be perpendicular to the mirror. You can now put a newspaper etc. flat on the mirror and take the shot.

Marc