Best "bang for buck" lenses (K-5)

**bobbotron** · 01-21-2016, 12:11 PM

Originally posted by normhead

Well then, what does it do?

This thread... is derailing. :P

Norm, I'm considering selling my 16-45 now that I have the 15, if you're interested in testing it I would sell it to you for a "I hate selling gear" Ontario deal. :P

kh1234567890 · 01-21-2016, 04:00 PM - **1 Like**

Originally posted by normhead

Well then, what does it do?

It depends on a particular software, but essentially you warp each of the three colour channel images, using a polynomial with coefficients read from the lens profile, to line them up. An iterative algorithm can also be used if there is no lens profile (you select an area and the software uses that to minimise the channel offsets).

How good it ends up depends on a number of factors (polynomial used, accuracy of the coefficients, effect of each of the colour channels spanning a range of wavelengths, amount of decentering of your particular sample of a lens, etc. etc.), but it can be made to work well if done before demosaicing. There is no reason why this approach should affect microcontrast.

normhead · 01-21-2016, 04:25 PM

Originally posted by kh1234567890

It depends on a particular software, but essentially you warp each of the three colour channel images, using a polynomial with coefficients read from the lens profile, to line them up. An iterative algorithm can also be used if there is no lens profile (you select an area and the software uses that to minimise the channel offsets).

How good it ends up depends on a number of factors (polynomial used, accuracy of the coefficients, effect of each of the colour channels spanning a range of wavelengths, amount of decentering of your particular sample of a lens, etc. etc.), but it can be made to work well if done before demosaicing. There is no reason why this approach should affect microcontrast.

I never know how to interpret that word.

**robgski** · 01-21-2016, 07:17 PM

I am currently renting the 10-17mm for the third time, it's a great lens, and if you are reasonably careful when composing the shot, you can avoid the CA

northcoastgreg · 01-22-2016, 08:32 AM

Originally posted by normhead

Having light scattered on your image where it's not supposed to be doesn't affect micro-contrast?

I'm not sure what is meant by "having light scattered on your image." Are you sure you're not confusing chromatic aberration with spherical aberration? Spherical aberration will most definitely affect micro-contrast, as is easily seen in any image from a soft-focus lens.

Originally posted by bobbotron

This thread... is derailing

Well, not necessarily. Because the issue of whether CA correction in post has any affect on image quality does actually have quite a bit to do with the topic of best bang for buck lenses, and here's why: CA correction in post is much cheaper than optical correction of CAs. So conceivably we can better glass (in terms of resolution, contrast, flare control, etc.) for less money if we are willing to accept software correction of CAs.

There are two broad attitudes toward this issue: that of the purists, and that of the pragmatists. The purists are offended at the very idea of any kind of software correction for lens' faults, regardless of whether it has any impact on "real world" image quality. The pragmatists only care about what actual images look like, and they'll gladly accept software corrections provided they don't have any real impact on image quality. Now while I wouldn't necessarily say that one attitude is "wrong" and the other "right" (it's all a matter of opinion), I would suggest that if you really want to get the best bang for you buck with lenses, you have no choice but to be a pragmatist. For those looking for the best bang for the buck lenses, the only question they should ask regarding CAs are: "Do they clean up easily in post?" If the answer is yes, there's no need to bother one's head about them.

normhead · 01-22-2016, 09:15 AM

Yet we seem to have no real world examples to support either case.

Whether or not images clean up in post is irrelevant if you aren't proficient with the software that cleans it up. It's hardly cost efficient to buy a lens based on the need to buy software to clean it up, unless, where relevant you include that in the cost of the lens.

So are we talking lens and software combinations bang for buck... or just the lens. I don't remember software being part of the package.

kh1234567890 · 01-22-2016, 10:34 AM

Originally posted by normhead

Whether or not images clean up in post is irrelevant if you aren't proficient with the software that cleans it up. It's hardly cost efficient to buy a lens based on the need to buy software to clean it up, unless, where relevant you include that in the cost of the lens.

So are we talking lens and software combinations bang for buck... or just the lens. I don't remember software being part of the package.

The extra cost is effectively zero since the CA and distortion corrections are built into the camera firmware (as long as you shoot JPEG, use Pentax lenses and are ok with the slightly slower saving time) or can be done in the free software (PDCU) that comes with the camera (if you shoot RAW). Even if you use paid for software (eg Lightroom etc) it is not exactly hard to tick the box that turns the corrections on.

normhead · 01-22-2016, 10:46 AM

That's a lot of caveats. I don't believe those were stated in your original post. You could have fooled somebody.

01-21-2016, 04:00 PM - 1 Like	#62
kh1234567890 Veteran Member Join Date: Sep 2010 Location: Manchester, UK Photos: Gallery Posts: 2,653	Originally posted by normhead Well then, what does it do? It depends on a particular software, but essentially you warp each of the three colour channel images, using a polynomial with coefficients read from the lens profile, to line them up. An iterative algorithm can also be used if there is no lens profile (you select an area and the software uses that to minimise the channel offsets). How good it ends up depends on a number of factors (polynomial used, accuracy of the coefficients, effect of each of the colour channels spanning a range of wavelengths, amount of decentering of your particular sample of a lens, etc. etc.), but it can be made to work well if done before demosaicing. There is no reason why this approach should affect microcontrast. Last edited by kh1234567890; 01-21-2016 at 04:10 PM.

01-22-2016, 10:46 AM	#68
normhead Pentaxian Join Date: Jun 2007 Location: Near Algonquin Park Photos: Gallery \| Albums Posts: 40,451	That's a lot of caveats. I don't believe those were stated in your original post. You could have fooled somebody. Last edited by normhead; 01-22-2016 at 11:11 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Best bang for the buck long 300mm tele for Q?	raider	Pentax Q	82	07-11-2018 06:50 AM
Wide angle best bang for the buck?	ismaelg	Pentax SLR Lens Discussion	19	04-04-2015 01:09 AM
Best bang for the buck in telephoto lenses?	pentaxmz	Pentax SLR Lens Discussion	20	08-13-2014 07:04 AM
"Best" lenses for K-3	hcc	Pentax SLR Lens Discussion	11	01-15-2014 05:51 AM
Best "bang-for-the-buck" flash?	lastdodobird	Flashes, Lighting, and Studio	24	02-13-2009 11:32 AM

01-21-2016, 12:11 PM	#61
bobbotron Site Supporter Join Date: Jan 2015 Location: Ottawa, ON Photos: Gallery Posts: 3,349	Originally posted by normhead Well then, what does it do? This thread... is derailing. :P Norm, I'm considering selling my 16-45 now that I have the 15, if you're interested in testing it I would sell it to you for a "I hate selling gear" Ontario deal. :P

01-21-2016, 04:25 PM	#63
normhead Pentaxian Join Date: Jun 2007 Location: Near Algonquin Park Photos: Gallery \| Albums Posts: 40,451	Originally posted by kh1234567890 It depends on a particular software, but essentially you warp each of the three colour channel images, using a polynomial with coefficients read from the lens profile, to line them up. An iterative algorithm can also be used if there is no lens profile (you select an area and the software uses that to minimise the channel offsets). How good it ends up depends on a number of factors (polynomial used, accuracy of the coefficients, effect of each of the colour channels spanning a range of wavelengths, amount of decentering of your particular sample of a lens, etc. etc.), but it can be made to work well if done before demosaicing. There is no reason why this approach should affect microcontrast. I never know how to interpret that word.

01-21-2016, 07:17 PM	#64
robgski Loyal Site Supporter Join Date: Nov 2012 Location: Shenandoah Valley, Virginia Photos: Gallery \| Albums Posts: 8,817	I am currently renting the 10-17mm for the third time, it's a great lens, and if you are reasonably careful when composing the shot, you can avoid the CA

01-22-2016, 08:32 AM	#65
northcoastgreg Pentaxian Join Date: Feb 2010 Location: Northern Michigan Photos: Gallery \| Albums Posts: 6,174	Originally posted by normhead Having light scattered on your image where it's not supposed to be doesn't affect micro-contrast? I'm not sure what is meant by "having light scattered on your image." Are you sure you're not confusing chromatic aberration with spherical aberration? Spherical aberration will most definitely affect micro-contrast, as is easily seen in any image from a soft-focus lens. Originally posted by bobbotron This thread... is derailing Well, not necessarily. Because the issue of whether CA correction in post has any affect on image quality does actually have quite a bit to do with the topic of best bang for buck lenses, and here's why: CA correction in post is much cheaper than optical correction of CAs. So conceivably we can better glass (in terms of resolution, contrast, flare control, etc.) for less money if we are willing to accept software correction of CAs. There are two broad attitudes toward this issue: that of the purists, and that of the pragmatists. The purists are offended at the very idea of any kind of software correction for lens' faults, regardless of whether it has any impact on "real world" image quality. The pragmatists only care about what actual images look like, and they'll gladly accept software corrections provided they don't have any real impact on image quality. Now while I wouldn't necessarily say that one attitude is "wrong" and the other "right" (it's all a matter of opinion), I would suggest that if you really want to get the best bang for you buck with lenses, you have no choice but to be a pragmatist. For those looking for the best bang for the buck lenses, the only question they should ask regarding CAs are: "Do they clean up easily in post?" If the answer is yes, there's no need to bother one's head about them.

01-22-2016, 09:15 AM	#66
normhead Pentaxian Join Date: Jun 2007 Location: Near Algonquin Park Photos: Gallery \| Albums Posts: 40,451	Yet we seem to have no real world examples to support either case. Whether or not images clean up in post is irrelevant if you aren't proficient with the software that cleans it up. It's hardly cost efficient to buy a lens based on the need to buy software to clean it up, unless, where relevant you include that in the cost of the lens. So are we talking lens and software combinations bang for buck... or just the lens. I don't remember software being part of the package.

01-22-2016, 10:34 AM	#67
kh1234567890 Veteran Member Join Date: Sep 2010 Location: Manchester, UK Photos: Gallery Posts: 2,653	Originally posted by normhead Whether or not images clean up in post is irrelevant if you aren't proficient with the software that cleans it up. It's hardly cost efficient to buy a lens based on the need to buy software to clean it up, unless, where relevant you include that in the cost of the lens. So are we talking lens and software combinations bang for buck... or just the lens. I don't remember software being part of the package. The extra cost is effectively zero since the CA and distortion corrections are built into the camera firmware (as long as you shoot JPEG, use Pentax lenses and are ok with the slightly slower saving time) or can be done in the free software (PDCU) that comes with the camera (if you shoot RAW). Even if you use paid for software (eg Lightroom etc) it is not exactly hard to tick the box that turns the corrections on.