what degree of sharpness should I expect?

**sholtzma** · 10-02-2007, 02:57 PM

This may be a misguided question; and if so, please forgive my naivete. I have a K10D and have been using various lenses with it. Often I magnify the shots (either in-camera or after downloading into the Pentax Browser), in order to see the degree of sharpness I have achieved. Even with my sharpest lens (a Tamron 90mm MF macro), my shots are at best sharp up to 100% magnification. Once I reach 200%, I cannot get sharp shots, even with a tripod, flash, wireless release, stationary subject, etc. Am I expecting way too much for 200%?

m8o · 10-02-2007, 03:08 PM

If you go over 100% I've found quality becomes a function of the software you're viewing/resizing with, and whether some processing is being done to anti-alias and perhaps sharpen and in what manner of the pixels which are now 2x2 (4 pixels) for every one pixel @ 100%.

Alternatively, zoom in 200%, then step back from your screen. Now further back, now further ... is it somewhat sharp again.

(I kid, but there's some truth to it too)

Peter Zack · 10-02-2007, 03:15 PM

Absolutely you're 'pushing' it too far. With the best lenses you should see some loss at 100%. it is a guarantee at 200%. As far as I'm concerned magnifying a shot that much is only good for careful editing a shot like cloning out a dust spec or something similar. You would never crop a shot that much to be printed so this is not the way to evaluate the sharpness.

Cropping shouldn't be more than 20% from the original IMO and if you want to get closer then get a longer lens or extension tubes for your macro. BTW, you own one, if not the best macro lens around for the focal range and it will tell you the limits of your setup if used correctly.

**sholtzma** · 10-02-2007, 06:42 PM

Thanks!

Part of what I was trying to determine was how accurately the camera was autofocussing and how accurately I was manually focussing. And yes, I would never be cropping anywhere near that much. And yes again, the Tamron 90mm MF macro is a tremendous lens; in fact, I am worried about replacing it (say, with a Sigma 150mm) with something that turns out to be inferior. Any suggestions for a longer (and autofocus) macro, while we're on that subject?

Peter Zack · 10-02-2007, 07:23 PM

I've seen pictures from the 2 Sigmas 150mm and 180mm that both look good. owners report they like them quite a bit. Other than older discountinued A* and FA* lenses I can't think of any others. The main advantage would be being able to get 1:1 while being further away from the subject.

For me, I'll magnify the shot myself to 50 or 100% to double check focus but that's about it unless I'm trying to edit out something small.

Check this site for the reviews of those lenses here:
PhotoZone
They may be tested on other bodies but they will give you some idea of the quality.

philmorley · 10-02-2007, 07:53 PM

Hi

the sigma 150 is not available for pentax. You only option (aside from fa*200 for af is the sigma 180mm, which many people seem very happy with. The voigtlander 125 will give you a touch more length but it is MF.

As for sharpness, get some newsprint or something with detail, put the camera with your tamron lens on it on a tripod (or table / something solid) set to about f5.6 with mirror lock up. That will be about as sharp as you will get. I check everything at 100% for doing stock, sometimes I have go to 200% for fine fixing / cloning etc but I really have found the need to think about sharpness at greater than 100%.

Phil

Ignacio · 10-05-2007, 06:01 PM

Hi Sholtzma,

Phil's suggestion to use newsprint is a very good one, but, given your self confessed "naiveté", you may not know what to do with the newsprint. Basically, what you're doing is trying to isolate the different elements of your setup in the test. Procedure:

1) Place the camera such that the film plane ends up just off the parallel to the newsprint on one axis. For ease of later description, I am going to assume that the newsprint is taped on the wall and your camera is mounted on a tripod in front of it perfectly level except for pointing ever so slightly upwards. This way, the film plane horizontal axis is parallel to the newsprint and the vertical axis just a few degrees off vertical.

2) Set the aperture to the optimum for that lens (depends on the lens, but a good starting point is one or two stops down from fully open).

3) Focus somewhere near the middle of the visual field.

4) Take the shot respecting all sharpness requirements (mirror lock up, remote shutter release, heavy tripod, image stabilization off, etc.)

Now you should have an image of text with three horizontal bands. A blurry section on top merging gradually into a critically sharp section near the middle (doesn't matter if you didn't get it exactly in the middle) merging finally into another blurry section at the bottom. The sharpness variation will be gradual, not abrupt. If it's too abrupt to see it clearly, you're pointing the camera too far up. If you can't see any bands, you didn't go far enough off vertical.

The exact middle of the middle section will be the most sharpness your equipment can produce (should be considerable with that lens). If you didn't get the film plane exactly parallel in the horizontal plane, you may find some variation in the horizontal direction as well, but it should be negligible compared to the vertical direction.

5) Now, to test your autofocus or manual focus with this knowledge of the best your equipment can do in hand: pick a sentence, word, or letter, depending on your magnification, and try to focus on it. This part of the test becomes more sensitive (challenging) if you angle the film plane more than for the previous test.

Have fun,

Ignacio

mattdm · 10-05-2007, 08:09 PM

Originally posted by sholtzma

This may be a misguided question; and if so, please forgive my naivete. I have a K10D and have been using various lenses with it. Often I magnify the shots (either in-camera or after downloading into the Pentax Browser), in order to see the degree of sharpness I have achieved. Even with my sharpest lens (a Tamron 90mm MF macro), my shots are at best sharp up to 100% magnification. Once I reach 200%, I cannot get sharp shots, even with a tripod, flash, wireless release, stationary subject, etc. Am I expecting way too much for 200%?

Err, yeah. At 200%, you're seeing 4 imaginary pixels made up out of assumptions based on the surrounding pixels. (In fact, due to the whole bayer filter thing, this is the case for colors even at 100%). The fact that you can zoom at all beyond 100% and get something that looks sharp is simply a testament to the impressive work people have done with graphics algorithms.

The *real* picture can't be sharper than the amount of data you have. Any detail that looks sharp beyond 100% is simply false data.

10-02-2007, 02:57 PM	#1
sholtzma Site Supporter Join Date: Sep 2007 Location: Salisbury, NC Photos: Gallery Posts: 1,043	what degree of sharpness should I expect? This may be a misguided question; and if so, please forgive my naivete. I have a K10D and have been using various lenses with it. Often I magnify the shots (either in-camera or after downloading into the Pentax Browser), in order to see the degree of sharpness I have achieved. Even with my sharpest lens (a Tamron 90mm MF macro), my shots are at best sharp up to 100% magnification. Once I reach 200%, I cannot get sharp shots, even with a tripod, flash, wireless release, stationary subject, etc. Am I expecting way too much for 200%?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Abstract 90 Degree Bend	clarkebd	Post Your Photos!	1	08-18-2010 09:31 PM
How much sharpness should I expect from the 100mm Macro WR?	iocchelli	Pentax SLR Lens Discussion	4	07-31-2010 09:01 PM
Sharpness VS Fine-Sharpness in k20d	wasim_altaf	Pentax DSLR Discussion	8	10-12-2009 11:41 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

10-02-2007, 03:08 PM	#2
m8o Veteran Member Join Date: Jul 2007 Location: 40°-55'-44" N / 73°-24'-07" W [on LI] Photos: Gallery \| Albums Posts: 3,092	If you go over 100% I've found quality becomes a function of the software you're viewing/resizing with, and whether some processing is being done to anti-alias and perhaps sharpen and in what manner of the pixels which are now 2x2 (4 pixels) for every one pixel @ 100%. Alternatively, zoom in 200%, then step back from your screen. Now further back, now further ... is it somewhat sharp again. (I kid, but there's some truth to it too)

10-02-2007, 03:15 PM	#3
Peter Zack Pentaxian Moderator Emeritus Join Date: May 2007 Location: Edmonton Alberta, Canada Photos: Gallery \| Albums Posts: 10,643	Absolutely you're 'pushing' it too far. With the best lenses you should see some loss at 100%. it is a guarantee at 200%. As far as I'm concerned magnifying a shot that much is only good for careful editing a shot like cloning out a dust spec or something similar. You would never crop a shot that much to be printed so this is not the way to evaluate the sharpness. Cropping shouldn't be more than 20% from the original IMO and if you want to get closer then get a longer lens or extension tubes for your macro. BTW, you own one, if not the best macro lens around for the focal range and it will tell you the limits of your setup if used correctly.

10-02-2007, 06:42 PM	#4
sholtzma Site Supporter Join Date: Sep 2007 Location: Salisbury, NC Photos: Gallery Posts: 1,043 Original Poster	Thanks! Part of what I was trying to determine was how accurately the camera was autofocussing and how accurately I was manually focussing. And yes, I would never be cropping anywhere near that much. And yes again, the Tamron 90mm MF macro is a tremendous lens; in fact, I am worried about replacing it (say, with a Sigma 150mm) with something that turns out to be inferior. Any suggestions for a longer (and autofocus) macro, while we're on that subject?

10-02-2007, 07:23 PM	#5
Peter Zack Pentaxian Moderator Emeritus Join Date: May 2007 Location: Edmonton Alberta, Canada Photos: Gallery \| Albums Posts: 10,643	I've seen pictures from the 2 Sigmas 150mm and 180mm that both look good. owners report they like them quite a bit. Other than older discountinued A* and FA* lenses I can't think of any others. The main advantage would be being able to get 1:1 while being further away from the subject. For me, I'll magnify the shot myself to 50 or 100% to double check focus but that's about it unless I'm trying to edit out something small. Check this site for the reviews of those lenses here: PhotoZone They may be tested on other bodies but they will give you some idea of the quality.

10-02-2007, 07:53 PM	#6
philmorley Veteran Member Join Date: Mar 2007 Location: in a house in Armidale, Australia Posts: 472	Hi the sigma 150 is not available for pentax. You only option (aside from fa*200 for af is the sigma 180mm, which many people seem very happy with. The voigtlander 125 will give you a touch more length but it is MF. As for sharpness, get some newsprint or something with detail, put the camera with your tamron lens on it on a tripod (or table / something solid) set to about f5.6 with mirror lock up. That will be about as sharp as you will get. I check everything at 100% for doing stock, sometimes I have go to 200% for fine fixing / cloning etc but I really have found the need to think about sharpness at greater than 100%. Phil

10-05-2007, 06:01 PM	#7
Ignacio New Member Join Date: Sep 2007 Location: Vancouver, BC, Canada Posts: 8	Hi Sholtzma, Phil's suggestion to use newsprint is a very good one, but, given your self confessed "naiveté", you may not know what to do with the newsprint. Basically, what you're doing is trying to isolate the different elements of your setup in the test. Procedure: 1) Place the camera such that the film plane ends up just off the parallel to the newsprint on one axis. For ease of later description, I am going to assume that the newsprint is taped on the wall and your camera is mounted on a tripod in front of it perfectly level except for pointing ever so slightly upwards. This way, the film plane horizontal axis is parallel to the newsprint and the vertical axis just a few degrees off vertical. 2) Set the aperture to the optimum for that lens (depends on the lens, but a good starting point is one or two stops down from fully open). 3) Focus somewhere near the middle of the visual field. 4) Take the shot respecting all sharpness requirements (mirror lock up, remote shutter release, heavy tripod, image stabilization off, etc.) Now you should have an image of text with three horizontal bands. A blurry section on top merging gradually into a critically sharp section near the middle (doesn't matter if you didn't get it exactly in the middle) merging finally into another blurry section at the bottom. The sharpness variation will be gradual, not abrupt. If it's too abrupt to see it clearly, you're pointing the camera too far up. If you can't see any bands, you didn't go far enough off vertical. The exact middle of the middle section will be the most sharpness your equipment can produce (should be considerable with that lens). If you didn't get the film plane exactly parallel in the horizontal plane, you may find some variation in the horizontal direction as well, but it should be negligible compared to the vertical direction. 5) Now, to test your autofocus or manual focus with this knowledge of the best your equipment can do in hand: pick a sentence, word, or letter, depending on your magnification, and try to focus on it. This part of the test becomes more sensitive (challenging) if you angle the film plane more than for the previous test. Have fun, Ignacio

10-05-2007, 08:09 PM	#8
mattdm Veteran Member Join Date: Feb 2007 Location: Boston, MA Photos: Gallery Posts: 2,948	Originally posted by sholtzma This may be a misguided question; and if so, please forgive my naivete. I have a K10D and have been using various lenses with it. Often I magnify the shots (either in-camera or after downloading into the Pentax Browser), in order to see the degree of sharpness I have achieved. Even with my sharpest lens (a Tamron 90mm MF macro), my shots are at best sharp up to 100% magnification. Once I reach 200%, I cannot get sharp shots, even with a tripod, flash, wireless release, stationary subject, etc. Am I expecting way too much for 200%? Err, yeah. At 200%, you're seeing 4 imaginary pixels made up out of assumptions based on the surrounding pixels. (In fact, due to the whole bayer filter thing, this is the case for colors even at 100%). The fact that you can zoom at all beyond 100% and get something that looks sharp is simply a testament to the impressive work people have done with graphics algorithms. The real picture can't be sharper than the amount of data you have. Any detail that looks sharp beyond 100% is simply false data.