Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
09-13-2014, 03:47 PM - 1 Like   #16
Site Supporter
cheekygeek's Avatar

Join Date: Jul 2008
Location: Alda, Nebraska USA
Photos: Gallery | Albums
Posts: 1,254
You *want* your sensor to "out-resolve the lens", people. Don't say it
like it's a bad thing. It is the reason that we can now do without an anti-alias filter
that blurred the output of the lens (what a dumb concept, however it was
necessary to avoid moire with lower resolution (Read: "larger pixel pitch") sensors.

I think that people are forgetting that if the sensor DOES NOT
out-resolve the lens you theoretically have moire. This is why AA
filters were necessary in the past, but as pixel pitch shrinks it
becomes unnecessary. It is not that the problem of moire has
disappeared, it is that the sensor "over samples" the image and
eliminates most moire. This isn't a function of sensor size, but
photosite size, which is why the Q can get away with no AA filter.

This page is a nice demo (albeit at lower MP sizes than we are talking
about): What Does 'Diffraction-Limited' Mean?
You get the picture. (Also take note of his "Conclusions") This
example takes big jumps (1MP, 4MP, 16MP) but smaller jumps are
obviously still going the same direction, if not as noticeably.


The point is that as sensors get better you have more "headroom" to
improve you photography by *taking advantage* of better lenses. But
there will probably be a point (particularly if you print large) at
which you start to realize that some of your lenses are no longer
passing muster.

09-13-2014, 06:34 PM   #17
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,114
QuoteOriginally posted by cheekygeek Quote
But
there will probably be a point (particularly if you print large) at
which you start to realize that some of your lenses are no longer
passing muster.
As if any of us can really afford to print big in the first place!


Steve
09-16-2014, 10:19 AM   #18
Site Supporter
LaurenOE's Avatar

Join Date: Sep 2009
Location: San Francisco Bay Area
Photos: Albums
Posts: 3,482
The K3 can not only out-resolve some lower quality lenses, but is also unforgiving of lens "slop".
09-16-2014, 11:51 PM   #19
Veteran Member
awaldram's Avatar

Join Date: May 2008
Location: Hampshire
Photos: Gallery
Posts: 720
QuoteOriginally posted by jatrax Quote
Of course it is a complete sentence. And I also used the word minutia which counts double. And that's just a straw man argument, it does not matter whether the sentence is complete. The minutia is the big thing here. You probably don't even have a dictionary.
Since when has minutia ever been big ?

As to the OP the worst that will happen if a sensor out-resolves the lens is the need for an AA filter will be ameliorated .

Now the use of ameliorated must be worth three and as its used in the common wrong context where alleviated would be more accurate I get 4

09-17-2014, 08:10 AM   #20
Pentaxian
yorik's Avatar

Join Date: Jan 2012
Location: Scotts Valley, CA
Photos: Gallery
Posts: 806
QuoteOriginally posted by RobA_Oz Quote
I wonder why we still miss Kodachrome ...
Blame Paul Simon.
09-17-2014, 01:02 PM   #21
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,114
QuoteOriginally posted by awaldram Quote
has minutia
...have minutia...

Correctness is incredibly important.


Steve
09-23-2014, 03:28 AM   #22
Veteran Member
awaldram's Avatar

Join Date: May 2008
Location: Hampshire
Photos: Gallery
Posts: 720
QuoteOriginally posted by stevebrot Quote
...have minutia...

Correctness is incredibly important.


Steve
So true
09-25-2014, 01:11 PM   #23
Site Supporter




Join Date: Jul 2011
Posts: 1,007
QuoteOriginally posted by cheekygeek Quote
You *want* your sensor to "out-resolve the lens", people. Don't say it
like it's a bad thing. It is the reason that we can now do without an anti-alias filter
that blurred the output of the lens (what a dumb concept, however it was
necessary to avoid moire with lower resolution (Read: "larger pixel pitch") sensors.

I think that people are forgetting that if the sensor DOES NOT
out-resolve the lens you theoretically have moire. This is why AA
filters were necessary in the past, but as pixel pitch shrinks it
becomes unnecessary. It is not that the problem of moire has
disappeared, it is that the sensor "over samples" the image and
eliminates most moire. This isn't a function of sensor size, but
photosite size, which is why the Q can get away with no AA filter.

This page is a nice demo (albeit at lower MP sizes than we are talking
about): What Does 'Diffraction-Limited' Mean?
You get the picture. (Also take note of his "Conclusions") This
example takes big jumps (1MP, 4MP, 16MP) but smaller jumps are
obviously still going the same direction, if not as noticeably.


The point is that as sensors get better you have more "headroom" to
improve you photography by *taking advantage* of better lenses. But
there will probably be a point (particularly if you print large) at
which you start to realize that some of your lenses are no longer
passing muster.
Good to see that someone here understands basic concepts.

Actually we should be happy in case the camera outresolves the lens. Only in this case you can clearly use the whole potential of the lens and sample it correctly on the sensor. The sensor in the camera should outresolve by a factor of at least 1.4 to avoid moiree - undersampling. From another standpoint outresolving implies wasted pixels, but donīt worry about that. You should also be aware that most scenes that you capture will not even present enough details to require high resolution - the discrete pixels on the sensor are outresolved most of the time. Put in yet another way - the K3 will not in all cases unleash unknown potentials of your lenses. Yet another prespective yet, would argue that pixel pitch is getting smaller and that shear optical resolution is influenced or overshadowed by camera shake, mirror virbration, ... Nikon had a lot of trouble with the D800 in this regard, at the micrometer level many factor are important for image quality.

09-26-2014, 04:25 AM   #24
Pentaxian




Join Date: Dec 2013
Photos: Gallery
Posts: 541
Let us put it this way : high resolution of K-3 sensor makes all the characteristic features of the lens more noticeable. For example the SMC version of DA 21 Lim is soft in the corners and quite good in the center. K-3 compared to, say K-01, will highlight the centre sharpness and highlight the corner softness. No magic here.
On the other hand FA 50 1.4 produces even better images on K-3 than it did on K-01 (although you have to stop it past F 2.0 or so).
09-26-2014, 05:32 AM - 1 Like   #25
Pentaxian
noelpolar's Avatar

Join Date: Oct 2013
Location: Goolwa, SA
Posts: 3,066
QuoteOriginally posted by LaurenOE Quote
The K3 can not only out-resolve some lower quality lenses, but is also unforgiving of lens "slop".
By the term "lens slop" are you reffering to me?....
09-26-2014, 07:50 AM   #26
Site Supporter




Join Date: Nov 2012
Location: North Wales
Photos: Gallery | Albums
Posts: 1,476
QuoteOriginally posted by zapp Quote
Actually we should be happy in case the camera outresolves the lens. Only in this case you can clearly use the whole potential of the lens and sample it correctly on the sensor. The sensor in the camera should outresolve by a factor of at least 1.4 to avoid moiree - undersampling.
Mr Mars takes the opposite view:

"There is no point in a sensor that has greater resolving power than the lens!"

However he doesn't mention moire and the increased Pixel density of eg K3 rendering the AA filter redundant. I'm not actually qualified to scrutinise his technical arguments, I htink many would tend to agree with his point about the pursuit of more MPx being rather distinct from the pursuit of better images. I think his quoted point is touched on by zapp = wasted pixels.
09-26-2014, 08:37 AM   #27
Site Supporter
lm4187's Avatar

Join Date: Jul 2009
Location: Germantown MD
Posts: 83
Just out of curiosity--does anyone have an example of a lens that is out-resolved by the K-3? So far what I have seen are a lot of opinions and theories, but no concrete evidence.

I've used DxOMark measurements to compare many lenses, from mediocre to superb, on different cameras: usually comparing the K20D, K-5 and K-3. In all cases I have seen so far, the resolution (DxO's 'sharpness') ranks as K-3 > K-5 > K20D. I would think that a less than stellar lens would show little or no change as the resolution of the camera increased. I haven't seen that, yet, but am interested in finding instances where the camera truly outresolves the lens.

Perhaps the Q-cameras with K-mount lenses?

Cheers,
Lou
09-26-2014, 09:04 AM   #28
Site Supporter




Join Date: Nov 2012
Location: North Wales
Photos: Gallery | Albums
Posts: 1,476
Actually I had a lens that my K-r out-resolved - a ludicrously bad hanimex 28-200mm. Basically the images became acceptable once well downsized. Focussing with magnified live view, particularly at 28mm, was impossible - the displayed image was too "mushy"
09-26-2014, 09:15 AM   #29
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,114
QuoteOriginally posted by lm4187 Quote
Just out of curiosity--does anyone have an example of a lens that is out-resolved by the K-3?
Yep! Though the same lens was out-resolved by my K10D and 35mm film. The problem is in drawing the line between theoretic resolution at the sensor and actual (measured) resolution of the optical system with lens mounted. In simplest of terms, if the two are the same the lens is either equal to or better than the sensor. In the real world it is not that simple. A good sensor tends to bring out the best in all but the nastiest of lenses and excellent lenses approach the limits of the sensor.

The question posed in the original post is actually sort of silly. In practice, with very few exceptions, every lens on my shelf performs better on the K-3 than on the K10D...yes, even the Soviet stuff. The interesting question might be whether the modest performers on the K-3 are doing better on that camera than my excellent lenses did on the K10D. Too bad I sold that camera. It would have been an interesting test, but alas, the K10D is sold.


Steve

---------- Post added 09-26-14 at 09:28 AM ----------

QuoteOriginally posted by lm4187 Quote
I've used DxOMark measurements to compare many lenses
It is actually pretty easy to do basic resolution testing yourself. Bob Atkins provides a usable target and instructions on his Web site. If you don't have a good enough printer or are very picky, you can also purchase a large resolution test target from Edmund Optics. And, of course, there is always Imatest ($$ - $$$).

Camera Lens Testing- Sharpness, Chromatic Aberration and Distortion - Bob Atkins Photography

http://www.edmundoptics.com/testing-targets/test-targets/resolution-test-tar...wer-chart/1665

imatest | Image Quality Testing Software & Test Charts


Steve

Last edited by stevebrot; 09-26-2014 at 09:30 AM.
09-26-2014, 11:59 AM   #30
Site Supporter
lm4187's Avatar

Join Date: Jul 2009
Location: Germantown MD
Posts: 83
QuoteOriginally posted by stevebrot Quote
Yep! Though the same lens was out-resolved by my K10D and 35mm film. The problem is in drawing the line between theoretic resolution at the sensor and actual (measured) resolution of the optical system with lens mounted. In simplest of terms, if the two are the same the lens is either equal to or better than the sensor. In the real world it is not that simple. A good sensor tends to bring out the best in all but the nastiest of lenses and excellent lenses approach the limits of the sensor.

The question posed in the original post is actually sort of silly. In practice, with very few exceptions, every lens on my shelf performs better on the K-3 than on the K10D...yes, even the Soviet stuff. The interesting question might be whether the modest performers on the K-3 are doing better on that camera than my excellent lenses did on the K10D. Too bad I sold that camera. It would have been an interesting test, but alas, the K10D is sold.


Steve

---------- Post added 09-26-14 at 09:28 AM ----------



It is actually pretty easy to do basic resolution testing yourself. Bob Atkins provides a usable target and instructions on his Web site. If you don't have a good enough printer or are very picky, you can also purchase a large resolution test target from Edmund Optics. And, of course, there is always Imatest ($$ - $$$).

Camera Lens Testing- Sharpness, Chromatic Aberration and Distortion - Bob Atkins Photography

http://www.edmundoptics.com/testing-targets/test-targets/resolution-test-tar...wer-chart/1665

imatest | Image Quality Testing Software & Test Charts


Steve
Actually, I have the Edmunds chart hanging on a wall, but I find that it's easier and quicker to tell the bad lenses from a few quick snaps. After hanging the chart and making a few measurements, I found I was too lazy to be as meticulous as needed.

Interestingly, my Sigma 28-200mm was obviously a loser from day 1, like the Hanimex mentioned above. [Same lens? ] It was so bad the first time I used it, that I didn't even consider it to be out-resolved by my cameras!

Sometimes a bad lens is just a bad lens.

I agree that most lenses are enhanced by the better sensor of the K-3. I think that worries about the K-3 (and similar sensors in other brands, to be fair) out-resolving most lenses is overblown, even for the 'older technology' MF lenses.

Cheers,
Lou
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
100mm, bob, dslr, experience, f4, imatest, issue, k-3, k-3 & lenses, k3, lens, lenses, pentax k-3, resolution, sensor, series, steve, test
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Pentax K-3 - Not that good with third-party lenses?! vishalv Troubleshooting and Beginner Help 34 08-18-2014 11:38 AM
Has It Been Confirmed That The K-3 Will Accept UHS-I...? tabl10s Pentax K-3 11 10-20-2013 12:21 PM
Will the K-3 automagically kill purple fringing and the like? Sagitta Pentax K-3 4 10-08-2013 03:53 PM
Will the VF-3 View Finder work on the K-01? Jerome Welch Pentax K-01 34 08-08-2013 02:56 AM
New lenses that will fit my k1000 Xerocs Troubleshooting and Beginner Help 3 08-05-2013 05:11 PM



All times are GMT -7. The time now is 01:53 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top