DA* 16-50 how to test quality

Burgerbrater · 09-30-2009, 08:11 AM

Ok I am a total newbie with this lens business. But I want to purchase the 16-50mm and I was wondering what is the best way to test sharpness, distortion etc. I read somewhere that someone taped some sort of a grid on his wall and photographed it. But how do you guys do it?

thanks

Taikonaut · 09-30-2009, 08:40 AM

Just take a lot of pictures to see if you like it or not.

**Ash** · 09-30-2009, 08:45 AM

Hi again.
First do a focus test using a proper focus test chart - these are available online - and follow its instructions to test it properly.

Once you're convinced focusing is spot on, then proceed to shoot flat subjects (brick-walls come to mind, and are commonly picked) at varying focal lengths wide-open - this will give you an imperfect idea of whether the lens is centrally sharp (which it should be) and reasonably sharp at the edges (which it should be also for a medium to far flat subject - beware interpreting corner softness using a close subject!)

Distortion is evident at 16mm, so don't be put off by that, but it shouldn't be too objectionable.
Then last of all, shoot real-world images with it (most important step) - you must be happy with these results if you are going to be at all happy with your lens.

All the best for that.
Hope this helped.

Burgerbrater · 09-30-2009, 09:01 AM

Originally posted by Ash

beware interpreting corner softness using a close subject!)

What do you mean by this?

Wouldnt a newspaper, taped to my apartment wall do the same thing? And how far must I stand away from the brick wall?

**Ash** · 09-30-2009, 09:39 AM

Look at the distances from the lens to the flat object you're shooting.

The closer the object is to the lens, the greater the discrepancy between the distance to the middle of the object and the distance to the corner of the object in the frame. The greater this difference, the more obvious gaussian blur (due to a thin depth of field) will be in the image, which is NOT the same as corner softness. This is further exacerbated by having the lens set at 16mm at f/2.8 with a close object. If you don't believe me, try doing the sharpness test with an object at minimum focus distance and see how much of the image will be sharp.

So your newspaper on the wall test is technically fine, but be prepared to see corner softness wide open because of how close the newspaper will have to be to the lens. This is why I am careful about calling a lens 'soft' in the corners.

Burgerbrater · 09-30-2009, 10:06 AM

Originally posted by Ash

Look at the distances from the lens to the flat object you're shooting.

The closer the object is to the lens, the greater the discrepancy between the distance to the middle of the object and the distance to the corner of the object in the frame. The greater this difference, the more obvious gaussian blur (due to a thin depth of field) will be in the image, which is NOT the same as corner softness. This is further exacerbated by having the lens set at 16mm at f/2.8 with a close object. If you don't believe me, try doing the sharpness test with an object at minimum focus distance and see how much of the image will be sharp.

So your newspaper on the wall test is technically fine, but be prepared to see corner softness wide open because of how close the newspaper will have to be to the lens. This is why I am careful about calling a lens 'soft' in the corners.

Ok sounds very complex. Please forgive me if I keep asking and asking. I dont completely understand it yet. How far would you position the camera from the wall to begin with?

**Ash** · 09-30-2009, 10:26 AM

Sorry if I wasn't clear and using too much jargon. It's hard to explain well without physically showing you.
The best way to be sure you're getting the best test results is if you have a wall further than 5m or so from the camera - at that distance, the differences mentioned before would be greatly reduced (relatively, not absolutely - but that's what matters more here).

This ball-park figure does vary depending on the lens and focal length, so 5m may not be enough for a telephoto lens. The DoF scale is helpful in this regard to get an idea of distance to subject sharpness at a particular aperture (but is not available on DA lenses AFAIK...)

filorp · 09-30-2009, 03:11 PM

So Ash, if you want to check decentering the best would be to do it from the close distance because the closer you place the subject the more visible becomes differences betw left and right corners? am i right?

**Ash** · 09-30-2009, 03:24 PM

Yes, I believe so for testing decentering (not for corner sharpness).
The small area of sharpness in the image should be seen in the centre as long as the central focus point is used.

Burgerbrater · 10-01-2009, 05:15 PM

Hey Ash,

today I went to Adorama personally. I asked specifically how many of the 16-50mm they have in stock. I was told just one. strange because so many people said theat Adorama has never had problems returning or exchanging lenses. At any rate I got the "last copy". And told the sales person that I wanted a good copy. Apparently he played dumb and didnt know what I was talking about. I told him that a lot of people have severe problems with this lens etc. etc. Anyway I bought the lens. Serial number: "9027713" Went back home and plugged it onto my K20d. Then the shock. The motor wasnt working. I couldnt auto focus at all. I saw some movement in the display but not much. I manually focused andd played around a little and eventually gave up. Then after an hour I picked it up again and strangeley there was mor activitiy in the focusing display. I then pressed the AF button on the back of the camera, focused it manually once, let the button go and kaboom now the autofocus works perfectly. I have to get used to not hearing the jigsaw like focus annymore since it is dead silent but it works.

Because of all this I couldnt test the sharpness and the softness and all that but I ll do it tomorrow with a brick wall.

However I shot the same Image again that I did with the Sigma 18-50mm ex and I can tell you so far there is no comparisen. The lesne is 30 percent sharper in the middle, higher contrast. Its like putting on glasses when you know that your eyes are getting bad. Also the manual focus is very very good. Smooth and precise. However I will have to do more tests before I conclude. After all I dont want decentration again.

so long,

Alex

**Ash** · 10-01-2009, 05:19 PM

Hi Alex.
It's important to rule out user error before labelling the equipment as faulty - glad that it worked out in the end, but AF should work straight from the word go, not needing any fiddling around with anything on the camera.

All the best with this fine lens.

Burgerbrater · 10-01-2009, 07:05 PM

By the way. Can front or back focus be corrected with the k20d?

**Ash** · 10-01-2009, 07:30 PM

Yes, to a degree.
Each lens can be calibrated separately on the K20D's focus adjustment feature, but don't correct major FF/BF.
None of my lenses have so far needed any adjustment...

**pentaxman** · 10-01-2009, 11:07 PM

Originally posted by Burgerbrater

Hey Ash,

However I shot the same Image again that I did with the Sigma 18-50mm ex and I can tell you so far there is no comparisen. The lesne is 30 percent sharper in the middle, higher contrast. Its like putting on glasses when you know that your eyes are getting bad. Also the manual focus is very very good. Smooth and precise. However I will have to do more tests before I conclude. After all I dont want decentration again.

so long,

Alex

I'll have to disagree with you on that one. The Pentax has some advantages over the Sigma 18-50 but sharpness and contrast are not two of them. I would have liked the 16mm focal length and weather sealing, but I bought three Pentaxes and all were decentered. The Pentaxes also had much higher CA and purple fringing.

I hope you got a good Pentax. I tested the 16-50s by shooting a lens chart against a wall. It is very important to get the lens axis exactly perpendicular to the wall. I did this by putting a mirror in the middle of the chart and adjusting my camera position until the reflection of the lens was in the exact center of the mirror. I tested my setup with known good lenses to be sure that I did not have anything misaligned.

Burgerbrater · 10-02-2009, 12:26 PM

Hi Ash,

I have tested the lens. I am dissapointed. It seems to me that not matter what f stop the right side is now decentered.

I put examples in a picasa album. Would kindly take a look and see what you think.

thanks,

Alex

Picasa Web Albums - alexconrady - Pentax 16-50m...

09-30-2009, 08:11 AM	#1
Burgerbrater New Member Join Date: Mar 2009 Posts: 23	*DA 16-50 how to test quality** Ok I am a total newbie with this lens business. But I want to purchase the 16-50mm and I was wondering what is the best way to test sharpness, distortion etc. I read somewhere that someone taped some sort of a grid on his wall and photographed it. But how do you guys do it? thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
K-5 video quality	dileepkrp	Video Recording and Processing	16	10-10-2010 12:45 PM
Fuji Quality II	woof	Film SLRs and Compact Film Cameras	3	02-21-2009 12:54 AM
dc.watch 6 DSLRs image quality test	yakiniku	Pentax News and Rumors	9	11-12-2008 05:02 AM
Quality between 2 lenses...	mchud	Pentax SLR Lens Discussion	6	10-01-2008 09:30 AM
Interesting DR Test and BG Noise Test of the K20D	RiceHigh	Pentax DSLR Discussion	14	06-25-2008 09:49 AM

09-30-2009, 08:40 AM	#2
Taikonaut Forum Member Join Date: Apr 2009 Location: New York Posts: 62	Just take a lot of pictures to see if you like it or not.

09-30-2009, 08:45 AM	#3
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	Hi again. First do a focus test using a proper focus test chart - these are available online - and follow its instructions to test it properly. Once you're convinced focusing is spot on, then proceed to shoot flat subjects (brick-walls come to mind, and are commonly picked) at varying focal lengths wide-open - this will give you an imperfect idea of whether the lens is centrally sharp (which it should be) and reasonably sharp at the edges (which it should be also for a medium to far flat subject - beware interpreting corner softness using a close subject!) Distortion is evident at 16mm, so don't be put off by that, but it shouldn't be too objectionable. Then last of all, shoot real-world images with it (most important step) - you must be happy with these results if you are going to be at all happy with your lens. All the best for that. Hope this helped.

09-30-2009, 09:01 AM	#4
Burgerbrater New Member Join Date: Mar 2009 Posts: 23 Original Poster	Originally posted by Ash beware interpreting corner softness using a close subject!) What do you mean by this? Wouldnt a newspaper, taped to my apartment wall do the same thing? And how far must I stand away from the brick wall?

09-30-2009, 09:39 AM	#5
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	Look at the distances from the lens to the flat object you're shooting. The closer the object is to the lens, the greater the discrepancy between the distance to the middle of the object and the distance to the corner of the object in the frame. The greater this difference, the more obvious gaussian blur (due to a thin depth of field) will be in the image, which is NOT the same as corner softness. This is further exacerbated by having the lens set at 16mm at f/2.8 with a close object. If you don't believe me, try doing the sharpness test with an object at minimum focus distance and see how much of the image will be sharp. So your newspaper on the wall test is technically fine, but be prepared to see corner softness wide open because of how close the newspaper will have to be to the lens. This is why I am careful about calling a lens 'soft' in the corners.

09-30-2009, 10:06 AM	#6
Burgerbrater New Member Join Date: Mar 2009 Posts: 23 Original Poster	Originally posted by Ash Look at the distances from the lens to the flat object you're shooting. The closer the object is to the lens, the greater the discrepancy between the distance to the middle of the object and the distance to the corner of the object in the frame. The greater this difference, the more obvious gaussian blur (due to a thin depth of field) will be in the image, which is NOT the same as corner softness. This is further exacerbated by having the lens set at 16mm at f/2.8 with a close object. If you don't believe me, try doing the sharpness test with an object at minimum focus distance and see how much of the image will be sharp. So your newspaper on the wall test is technically fine, but be prepared to see corner softness wide open because of how close the newspaper will have to be to the lens. This is why I am careful about calling a lens 'soft' in the corners. Ok sounds very complex. Please forgive me if I keep asking and asking. I dont completely understand it yet. How far would you position the camera from the wall to begin with?

09-30-2009, 10:26 AM	#7
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	Sorry if I wasn't clear and using too much jargon. It's hard to explain well without physically showing you. The best way to be sure you're getting the best test results is if you have a wall further than 5m or so from the camera - at that distance, the differences mentioned before would be greatly reduced (relatively, not absolutely - but that's what matters more here). This ball-park figure does vary depending on the lens and focal length, so 5m may not be enough for a telephoto lens. The DoF scale is helpful in this regard to get an idea of distance to subject sharpness at a particular aperture (but is not available on DA lenses AFAIK...)

09-30-2009, 03:11 PM	#8
filorp Veteran Member Join Date: Jun 2008 Location: Aberdeen Scotland Photos: Gallery \| Albums Posts: 398	So Ash, if you want to check decentering the best would be to do it from the close distance because the closer you place the subject the more visible becomes differences betw left and right corners? am i right?

09-30-2009, 03:24 PM	#9
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	Yes, I believe so for testing decentering (not for corner sharpness). The small area of sharpness in the image should be seen in the centre as long as the central focus point is used.

10-01-2009, 05:15 PM	#10
Burgerbrater New Member Join Date: Mar 2009 Posts: 23 Original Poster	Hey Ash, today I went to Adorama personally. I asked specifically how many of the 16-50mm they have in stock. I was told just one. strange because so many people said theat Adorama has never had problems returning or exchanging lenses. At any rate I got the "last copy". And told the sales person that I wanted a good copy. Apparently he played dumb and didnt know what I was talking about. I told him that a lot of people have severe problems with this lens etc. etc. Anyway I bought the lens. Serial number: "9027713" Went back home and plugged it onto my K20d. Then the shock. The motor wasnt working. I couldnt auto focus at all. I saw some movement in the display but not much. I manually focused andd played around a little and eventually gave up. Then after an hour I picked it up again and strangeley there was mor activitiy in the focusing display. I then pressed the AF button on the back of the camera, focused it manually once, let the button go and kaboom now the autofocus works perfectly. I have to get used to not hearing the jigsaw like focus annymore since it is dead silent but it works. Because of all this I couldnt test the sharpness and the softness and all that but I ll do it tomorrow with a brick wall. However I shot the same Image again that I did with the Sigma 18-50mm ex and I can tell you so far there is no comparisen. The lesne is 30 percent sharper in the middle, higher contrast. Its like putting on glasses when you know that your eyes are getting bad. Also the manual focus is very very good. Smooth and precise. However I will have to do more tests before I conclude. After all I dont want decentration again. so long, Alex

10-01-2009, 05:19 PM	#11
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	Hi Alex. It's important to rule out user error before labelling the equipment as faulty - glad that it worked out in the end, but AF should work straight from the word go, not needing any fiddling around with anything on the camera. All the best with this fine lens.

10-01-2009, 07:05 PM	#12
Burgerbrater New Member Join Date: Mar 2009 Posts: 23 Original Poster	By the way. Can front or back focus be corrected with the k20d?

10-01-2009, 07:30 PM	#13
Ash Loyal Site Supporter Join Date: Mar 2007 Location: Toowoomba, Queensland Photos: Gallery \| Albums Posts: 23,920	Yes, to a degree. Each lens can be calibrated separately on the K20D's focus adjustment feature, but don't correct major FF/BF. None of my lenses have so far needed any adjustment...

10-01-2009, 11:07 PM	#14
pentaxman Site Supporter Join Date: Oct 2007 Location: Washington Photos: Gallery Posts: 250	Originally posted by Burgerbrater Hey Ash, However I shot the same Image again that I did with the Sigma 18-50mm ex and I can tell you so far there is no comparisen. The lesne is 30 percent sharper in the middle, higher contrast. Its like putting on glasses when you know that your eyes are getting bad. Also the manual focus is very very good. Smooth and precise. However I will have to do more tests before I conclude. After all I dont want decentration again. so long, Alex I'll have to disagree with you on that one. The Pentax has some advantages over the Sigma 18-50 but sharpness and contrast are not two of them. I would have liked the 16mm focal length and weather sealing, but I bought three Pentaxes and all were decentered. The Pentaxes also had much higher CA and purple fringing. I hope you got a good Pentax. I tested the 16-50s by shooting a lens chart against a wall. It is very important to get the lens axis exactly perpendicular to the wall. I did this by putting a mirror in the middle of the chart and adjusting my camera position until the reflection of the lens was in the exact center of the mirror. I tested my setup with known good lenses to be sure that I did not have anything misaligned.

10-02-2009, 12:26 PM	#15
Burgerbrater New Member Join Date: Mar 2009 Posts: 23 Original Poster	Hi Ash, I have tested the lens. I am dissapointed. It seems to me that not matter what f stop the right side is now decentered. I put examples in a picasa album. Would kindly take a look and see what you think. thanks, Alex Picasa Web Albums - alexconrady - Pentax 16-50m...