FA43 sharpness test

**rparmar** · 08-31-2010, 01:14 PM - **1 Like**

introduction

In My New Pentax 31mm LTD...Is this normal?! "oxidized" wonders if his FA31 is too soft wide open. I decided to test the FA43, not because it would have anything directly to do with that question, but because it might tell us something about post-processing and how many variables are involved in making such a determination.

I had my suspicions, having recently learned a few things about testing protocol in K-x versus K20D ISO test.

So, I set up the camera and lens on a crappy tripod and decided to take some quick shots of a score laid out on our music stand. This inclines the page slightly, so only one vertical section will be in exact focus. I took shots at f/4, f/2.8, f/2 and f/1.9. The camera has all noise reduction and sharpness settings off. I used AF at a distance of about 1m. I shot RAW to DNG files, processed neutrally in ACR, and then took a 600 pixel square crop at 100%, choosing the area that should be most in focus.

test results

FA43 at f/4.0

FA43 at f/2.8

FA43 at f/2.0

FA43 at f/1.9

criticisms

There are several possible criticisms of this test I will make and answer, to stave off the inevitable replies.

1. I knew from the previous tests that the K-x might have issues with mirror slap. The solution is to use a flash to freeze the shot. For the sake of time I did not do that here. But my experience is that this isn't a problem I can perceive.

2. There is no fine detail in the image. But then again, this is not a test of resolution. The letter and note shapes should be enough to test sharpness.

3. There is no colour in the image. Oh well, that makes things simpler.

4. The shots are underexposed. For some reason I did not spot meter. Now I've lost the light, or I'd go back and try again. But I doubt this will have a significant effect on the results.

I'll present some conclusions in the next post.

**rparmar** · 08-31-2010, 01:22 PM

conclusions

As near as I can see the f/4 shot is the sharpest, but it is less easy choosing between the rest. Nonetheless, I would be comfortable saying that wide open the lens is not insanely worse than stopped down one. There seems to be more colour shift at f/1.9, but with the random white balance that seems to be happening, that could be due to some other factor.

Indeed, all variance might well be the result of the auto-focus system. So that is the first conclusion: It is better to manually focus. If nothing else it removes one possible variable. Unfortunately I am not much good at focusing on the K-x.

The second conclusion is that every image looks pretty bad using default settings. One must post-process in order to get anything good out of a digital image. And this seems to be more true with some cameras than others.

Of course this makes a joke of "objective" testing in this way. If we all post-process how will we ever establish a common baseline everyone will agree to? We must fall back on mechanical work-bench testing instead of photo-based evaluation.

In the following image I used the "Auto" adjustment in ACR and added some sharpening. Then I followed my usual post-process routine, which involves contrast and grey scale range corrections, as well as further (subtle) high-pass sharpening.

FA43 at f/1.9 processed

It is apparent that this is a much more attractive image. And for my purposes it is sharp enough. If you disagree, remember that we are looking at 100% crops. No-one does that "in real life".

This result conforms to my previous realisation that, even wide open, the FA43 is plenty sharp, at least in the centre of the frame. I prefer to use it at f/4, but if I need to go faster I do so without any qualms.

oxidized · 08-31-2010, 01:30 PM

Thanks rparmar! Yea, id say the F2.8 result is probably due to the auto focus. When i start to stop down my sharpness tends to improve by similar increments, however at least for the 31mm the differance between 1.8 and 2.8 seems more substantial.

Now if you could only take some unprocessed shots at F1.8 and focus at inifinity

Nesster · 08-31-2010, 01:37 PM

... good effort, these things tend to get picked apart... I would say for my purposes the FA43 is sharp enough as well

On the processed one, I seem to see some artifacts from sharpening or something, e.g. around the G in the box, lower right corner. I know 100%crop and all.. and in the end it's what we see that counts.

**rparmar** · 08-31-2010, 01:40 PM

Originally posted by Nesster

On the processed one, I seem to see some artifacts from sharpening or something, e.g. around the G in the box, lower right corner.

Yes, it's true, but I would go further: there are always artefacts of sharpening. After all, it is a digital process which corrupts an image to fool the eye. Some artefacts we like and some we don't!

**rparmar** · 08-31-2010, 01:42 PM

Originally posted by oxidized

Now if you could only take some unprocessed shots at F1.8 and focus at inifinity

It is pretty hard to find any object at infinity that actually emits/reflects light. Not to mention atmospheric distortion!

But maybe tomorrow I will find something close to infinity.

oxidized · 08-31-2010, 01:45 PM

well, not necessarily infinity, but something close to it. like something far away. just to compare. thanks again the pics and the effort!

Nesster · 08-31-2010, 02:39 PM

do we listen to Ken Rockwell?

If we do, his suggestion is to shoot at infinity, across a big lawn, say, at some trees. You can check the jpeg size (remember, he don't do raw) and pixel peep the leaves.

Marc Sabatella · 08-31-2010, 03:33 PM

That's a really effective demonstration of the effect of testing methods and in particular PP can have on test results, and why comparing results obtained by different photographers using different methods is close to worthless.

Doanh · 08-31-2010, 03:37 PM

When I bought the 43 Ltd, I expected it to be soft wide open based on various reviews. So I was surprised to find it is quite sharp wide open (as stated in the PF review section). Thanks rparmar for the confirmation.

Peter

Marc Sabatella · 08-31-2010, 05:56 PM

Originally posted by rparmar

It is pretty hard to find any object at infinity that actually emits/reflects light.

I can think of one, and it's available for shooting almost every day from sunrise to sunset. But I don't recommend it.

NeverSatisfied · 08-31-2010, 07:08 PM

Originally posted by Doanh

When I bought the 43 Ltd, I expected it to be soft wide open based on various reviews. So I was surprised to find it is quite sharp wide open (as stated in the PF review section). Thanks rparmar for the confirmation.

Peter

I just received mine last week, and I'm very happy with it indeed! I too was surprised at how sharp it is even wide open (some of the reviews said otherwise). I sold off my FA 35 and F 50 to get this, and I already like it a whole lot better than either of those. It's compact and light, the FOV suits me quite well and the construction and handling are very nice. Glad I did it, and the FA 43 will probably end up as the default lens on my K20D now.

**rparmar** · 09-01-2010, 03:59 AM

Originally posted by Marc Sabatella

I can think of one, and it's available for shooting almost every day from sunrise to sunset. But I don't recommend it.

There are also smaller versions at night.

GeneV · 09-01-2010, 05:42 AM

Originally posted by Nesster

do we listen to Ken Rockwell?

If we do, his suggestion is to shoot at infinity, across a big lawn, say, at some trees. You can check the jpeg size (remember, he don't do raw) and pixel peep the leaves.

I prefer a detail that doesn't move, like a brick. Here it is hard to find a day with no breeze.

GeneV · 09-01-2010, 05:50 AM

Originally posted by rparmar

Yes, it's true, but I would go further: there are always artefacts of sharpening. After all, it is a digital process which corrupts an image to fool the eye. Some artefacts we like and some we don't!

That's one way to look at it, but there is not much of an image to "corrupt" until all the digital processing puts the pixels together into an image. Sharpening is as much a part of that process as demosaicing, and in some digital formats they can be part of the same process. This was something that I had trouble getting through my head after years of being a slide film purist. I keep thinking there is something sinful about sharpening, but in reality, some sharpening is absolutely necessary given the way pixels react.

08-31-2010, 01:14 PM - 1 Like	#1
rparmar Site Supporter Join Date: Jan 2008 Photos: Gallery Posts: 8,817	FA43 sharpness test introduction In My New Pentax 31mm LTD...Is this normal?! "oxidized" wonders if his FA31 is too soft wide open. I decided to test the FA43, not because it would have anything directly to do with that question, but because it might tell us something about post-processing and how many variables are involved in making such a determination. I had my suspicions, having recently learned a few things about testing protocol in K-x versus K20D ISO test. So, I set up the camera and lens on a crappy tripod and decided to take some quick shots of a score laid out on our music stand. This inclines the page slightly, so only one vertical section will be in exact focus. I took shots at f/4, f/2.8, f/2 and f/1.9. The camera has all noise reduction and sharpness settings off. I used AF at a distance of about 1m. I shot RAW to DNG files, processed neutrally in ACR, and then took a 600 pixel square crop at 100%, choosing the area that should be most in focus. test results FA43 at f/4.0 FA43 at f/2.8 FA43 at f/2.0 FA43 at f/1.9 criticisms There are several possible criticisms of this test I will make and answer, to stave off the inevitable replies. 1. I knew from the previous tests that the K-x might have issues with mirror slap. The solution is to use a flash to freeze the shot. For the sake of time I did not do that here. But my experience is that this isn't a problem I can perceive. 2. There is no fine detail in the image. But then again, this is not a test of resolution. The letter and note shapes should be enough to test sharpness. 3. There is no colour in the image. Oh well, that makes things simpler. 4. The shots are underexposed. For some reason I did not spot meter. Now I've lost the light, or I'd go back and try again. But I doubt this will have a significant effect on the results. I'll present some conclusions in the next post.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Takumar 1:3.5/28 Sharpness test	eccs19	Pentax SLR Lens Discussion	2	04-10-2009 12:22 PM
Takumar 1:4/100 sharpness test	eccs19	Pentax SLR Lens Discussion	0	04-10-2009 09:18 AM
Sharpness vs Fine Sharpness on K20D	morfic	Pentax DSLR Discussion	2	11-02-2008 10:13 AM
Fine sharpness and sharpness move together on K20D 1.01	morfic	Pentax DSLR Discussion	2	07-11-2008 09:18 PM
Sharpness test, 70-300mm vs 55-300mm	audiobomber	Pentax SLR Lens Discussion	13	07-11-2008 09:28 AM

08-31-2010, 01:22 PM	#2
rparmar Site Supporter Join Date: Jan 2008 Photos: Gallery Posts: 8,817 Original Poster	conclusions As near as I can see the f/4 shot is the sharpest, but it is less easy choosing between the rest. Nonetheless, I would be comfortable saying that wide open the lens is not insanely worse than stopped down one. There seems to be more colour shift at f/1.9, but with the random white balance that seems to be happening, that could be due to some other factor. Indeed, all variance might well be the result of the auto-focus system. So that is the first conclusion: It is better to manually focus. If nothing else it removes one possible variable. Unfortunately I am not much good at focusing on the K-x. The second conclusion is that every image looks pretty bad using default settings. One must post-process in order to get anything good out of a digital image. And this seems to be more true with some cameras than others. Of course this makes a joke of "objective" testing in this way. If we all post-process how will we ever establish a common baseline everyone will agree to? We must fall back on mechanical work-bench testing instead of photo-based evaluation. In the following image I used the "Auto" adjustment in ACR and added some sharpening. Then I followed my usual post-process routine, which involves contrast and grey scale range corrections, as well as further (subtle) high-pass sharpening. FA43 at f/1.9 processed It is apparent that this is a much more attractive image. And for my purposes it is sharp enough. If you disagree, remember that we are looking at 100% crops. No-one does that "in real life". This result conforms to my previous realisation that, even wide open, the FA43 is plenty sharp, at least in the centre of the frame. I prefer to use it at f/4, but if I need to go faster I do so without any qualms.

08-31-2010, 01:30 PM	#3
oxidized Veteran Member Join Date: Dec 2009 Location: USA - Delaware Photos: Albums Posts: 435	Thanks rparmar! Yea, id say the F2.8 result is probably due to the auto focus. When i start to stop down my sharpness tends to improve by similar increments, however at least for the 31mm the differance between 1.8 and 2.8 seems more substantial. Now if you could only take some unprocessed shots at F1.8 and focus at inifinity

08-31-2010, 01:37 PM	#4
Nesster Veteran Member Join Date: Oct 2006 Location: NJ USA Photos: Gallery Posts: 13,072	... good effort, these things tend to get picked apart... I would say for my purposes the FA43 is sharp enough as well On the processed one, I seem to see some artifacts from sharpening or something, e.g. around the G in the box, lower right corner. I know 100%crop and all.. and in the end it's what we see that counts.

08-31-2010, 01:40 PM	#5
rparmar Site Supporter Join Date: Jan 2008 Photos: Gallery Posts: 8,817 Original Poster	Originally posted by Nesster On the processed one, I seem to see some artifacts from sharpening or something, e.g. around the G in the box, lower right corner. Yes, it's true, but I would go further: there are always artefacts of sharpening. After all, it is a digital process which corrupts an image to fool the eye. Some artefacts we like and some we don't!

08-31-2010, 01:42 PM	#6
rparmar Site Supporter Join Date: Jan 2008 Photos: Gallery Posts: 8,817 Original Poster	Originally posted by oxidized Now if you could only take some unprocessed shots at F1.8 and focus at inifinity It is pretty hard to find any object at infinity that actually emits/reflects light. Not to mention atmospheric distortion! But maybe tomorrow I will find something close to infinity.

08-31-2010, 01:45 PM	#7
oxidized Veteran Member Join Date: Dec 2009 Location: USA - Delaware Photos: Albums Posts: 435	well, not necessarily infinity, but something close to it. like something far away. just to compare. thanks again the pics and the effort!

08-31-2010, 02:39 PM	#8
Nesster Veteran Member Join Date: Oct 2006 Location: NJ USA Photos: Gallery Posts: 13,072	do we listen to Ken Rockwell? If we do, his suggestion is to shoot at infinity, across a big lawn, say, at some trees. You can check the jpeg size (remember, he don't do raw) and pixel peep the leaves.

08-31-2010, 03:33 PM	#9
Marc Sabatella Veteran Member Join Date: Sep 2006 Location: Denver, CO Photos: Gallery Posts: 10,685	That's a really effective demonstration of the effect of testing methods and in particular PP can have on test results, and why comparing results obtained by different photographers using different methods is close to worthless.

08-31-2010, 03:37 PM	#10
Doanh Senior Member Join Date: May 2009 Location: Edmonton, AB Posts: 247	When I bought the 43 Ltd, I expected it to be soft wide open based on various reviews. So I was surprised to find it is quite sharp wide open (as stated in the PF review section). Thanks rparmar for the confirmation. Peter

08-31-2010, 05:56 PM	#11
Marc Sabatella Veteran Member Join Date: Sep 2006 Location: Denver, CO Photos: Gallery Posts: 10,685	Originally posted by rparmar It is pretty hard to find any object at infinity that actually emits/reflects light. I can think of one, and it's available for shooting almost every day from sunrise to sunset. But I don't recommend it.

Thread Tools	Search this Thread
	Search this Thread: Advanced Search

08-31-2010, 07:08 PM	#12
NeverSatisfied Veteran Member Join Date: Dec 2007 Location: S.E. Michigan Photos: Albums Posts: 688	Originally posted by Doanh When I bought the 43 Ltd, I expected it to be soft wide open based on various reviews. So I was surprised to find it is quite sharp wide open (as stated in the PF review section). Thanks rparmar for the confirmation. Peter I just received mine last week, and I'm very happy with it indeed! I too was surprised at how sharp it is even wide open (some of the reviews said otherwise). I sold off my FA 35 and F 50 to get this, and I already like it a whole lot better than either of those. It's compact and light, the FOV suits me quite well and the construction and handling are very nice. Glad I did it, and the FA 43 will probably end up as the default lens on my K20D now.

09-01-2010, 03:59 AM	#13
rparmar Site Supporter Join Date: Jan 2008 Photos: Gallery Posts: 8,817 Original Poster	Originally posted by Marc Sabatella I can think of one, and it's available for shooting almost every day from sunrise to sunset. But I don't recommend it. There are also smaller versions at night.

09-01-2010, 05:42 AM	#14
GeneV Veteran Member Join Date: Oct 2008 Location: Albuquerque NM Photos: Albums Posts: 9,830	Originally posted by Nesster do we listen to Ken Rockwell? If we do, his suggestion is to shoot at infinity, across a big lawn, say, at some trees. You can check the jpeg size (remember, he don't do raw) and pixel peep the leaves. I prefer a detail that doesn't move, like a brick. Here it is hard to find a day with no breeze.

09-01-2010, 05:50 AM	#15
GeneV Veteran Member Join Date: Oct 2008 Location: Albuquerque NM Photos: Albums Posts: 9,830	Originally posted by rparmar Yes, it's true, but I would go further: there are always artefacts of sharpening. After all, it is a digital process which corrupts an image to fool the eye. Some artefacts we like and some we don't! That's one way to look at it, but there is not much of an image to "corrupt" until all the digital processing puts the pixels together into an image. Sharpening is as much a part of that process as demosaicing, and in some digital formats they can be part of the same process. This was something that I had trouble getting through my head after years of being a slide film purist. I keep thinking there is something sinful about sharpening, but in reality, some sharpening is absolutely necessary given the way pixels react.