Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
08-28-2012, 10:33 AM - 1 Like   #31
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,400
QuoteOriginally posted by bdery Quote
Yes, the sensor influences DOF. Your lens has a fixed resolution, that you can or cannot really exploit depending on your sensor (if your sensor can provide more resolution than your lens, you will see the resolution limit. If the sensor has less resolution than what your lens can provide, your picture will be limited by the resolution of the sensor).

Before we go any further, it is important to define the circle of confusion. It goes like this : take a dimensionless point. The image of the point formed by any lens will be a small circle, not a point. But that circle can be small enough that it will be perceived as a point nonetheless.That's when the sensor comes into play. If the circle is equal or small than a pixel on your sensor, it will be recorded as a point. If it's larger, it will be recorded as a circle. The largest circle that will be seen as a point is called the circle of confusion.

Still here?

Now, if the image of a point looks like a point, you will perceive the image as sharp. If it looks like a circle, you will perceive the image as... less sharp. Now, asyou move away from the point of sharpest focus, the image of your point will look more and more like a fuzzy circle. But as long as that circle is smaller than the circle of confusion, for all practical matters the image will still look sharp, because the circle will still look like a point.

That's the point.

So DOF is linked to the lens properties, of course, but also to the sensor.

I hope it helps.
you have gone too far here on the idea. whether it is a pixel or a grain of silver, we never discuss depth of field relitive to the maximum resolving power of the medium, either sensor or film, because you can't resolve anything when image artifacts occupy such a small part of the frame.

Normally when discussing depth of field and circle of confusion we are discussing things that occupy more than one pixel, we are discussing large objects occupying much of the frame..

08-28-2012, 11:02 AM   #32
Pentaxian




Join Date: Nov 2011
Photos: Gallery | Albums
Posts: 3,735
QuoteOriginally posted by Lowell Goudge Quote
you have gone too far here on the idea. whether it is a pixel or a grain of silver, we never discuss depth of field relitive to the maximum resolving power of the medium, either sensor or film
The Online Depth of Field Calculator
asks you to input the camera or film format
if you don't know the specific circle of confusion.
08-28-2012, 11:30 AM   #33
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,400
QuoteOriginally posted by lytrytyr Quote
The Online Depth of Field Calculator
asks you to input the camera or film format
if you don't know the specific circle of confusion.
but that has nothing to do with the pixel size, only that all sensors are not created eqwual. generally speaking the ASP-C format is 16x24mm but different sensors could be 15.7 x 23,7 for example, and still called ASP-C. that is why they ask for the sensor, it is to get to the overall image size recorded, not to look at pixels.

note that the circle of confusion is really 1/100 of an inch on an 8x10 enlargement off 35mm print / slide film. everything equates back to the 1/100 of an inch so if you change the sensor size . format, then to get a full frame up to 8 x 10 requires higher magnifiction and therefore a smaller CofC on the sensor.,
08-28-2012, 11:57 AM   #34
New Member




Join Date: Aug 2012
Photos: Albums
Posts: 5
Nice related article from The Online Photographer:

The Online Photographer: So You Say You Want a 'Sharp' Lens

08-28-2012, 12:35 PM   #35
Pentaxian
Just1MoreDave's Avatar

Join Date: Feb 2007
Location: Aurora, CO
Photos: Gallery | Albums
Posts: 4,862
QuoteOriginally posted by slip Quote
to test this theory... If someone with a lot of primes/zooms take a photo of the same subject at the same focal length (lets say 50mm) and post them and let's see if you can tell the difference between the photos (don't forget to put in a couple of el cheapo's)
take out the exif.... would be very interesting to compare


thanks

randy
Here are two I took to demonstrate something from the Sears lens I referred to earlier. The first one is the Sears 28/2.8, the second is a reasonably sharp Tokina 28/2.8 EL model, both at f2.8. My idea was they were both inexpensive 3rd party primes with similar color and contrast, but at different ends of the sharpness spectrum - the Sears has significant coma.





But as I have said elsewhere, I like the Sears for everything except sharpness, and have photos that I like from it. It proves to me that sharpness isn't everything. But it sure won't make me sell the sharper 28mm lenses I have.
08-28-2012, 01:29 PM - 1 Like   #36
Pentaxian




Join Date: Feb 2010
Location: Eureka, CA
Photos: Gallery | Albums
Posts: 1,977
QuoteOriginally posted by altopiet Quote
Do I understand what you say correctly, that the lens resolution will differ between 12mp K-x and 16mp k-5 sensors, which will influence image sharpness taken with the same lens on both cameras?
Yes, that's true, but it's not that significant. A bad or mediocre lens will still be a bad or mediocre lens, even if it does score higher on resolution tests on a higher MP camera.

QuoteOriginally posted by altopiet Quote
If that is so, is there any simple way to find out at what mp a certain lens will reach optimal sharpness, for instance my 1961 Tak 105/2.8, FA 28-80mm, DA 55-300mm
You could test each lens on different MP cameras. However, just as measured resolution tends to be over-rated in lenses, measured resolution in sensors is even more over-rated. If you have a double-digit MP sensor, you have more than enough MPs for most purposes. Additional MPs just lead to slower post-processing and greater storage issues. The differences in terms of visual sharpness between my 10 MP K200D and my 16 MP are, with most of my lenses very subtle, even at 100% resolution. The only lens where the gain appears significant is the DA 10-17; but I'm not altogether sure that it isn't merely an issue of better AF with the K-5.
08-28-2012, 04:31 PM   #37
Pentaxian
tromboads's Avatar

Join Date: Apr 2012
Location: Melbs
Photos: Gallery | Albums
Posts: 1,148
QuoteOriginally posted by Just1MoreDave Quote
Here are two I took to demonstrate something from the Sears lens I referred to earlier....
And at the end of the day, It's still a boring photo of some boring plants.

That extra level of detail (even as blindly obvious as that) didn't make the sharper image any more interesting to look at. So again I say, sharpness shmarpness
08-28-2012, 05:25 PM   #38
Pentaxian




Join Date: Jun 2008
Location: Flyover America
Posts: 4,475
QuoteOriginally posted by Just1MoreDave Quote
The first one is the Sears 28/2.8,
Just going by this one pic...

...it does point out the other side of the argument. There are limits and this lens is beyond my limits. For me it would be just too soft to use for any purpose. I would think even the kit lens would do better than this.

08-28-2012, 08:08 PM   #39
Pentaxian




Join Date: Feb 2010
Location: Eureka, CA
Photos: Gallery | Albums
Posts: 1,977
QuoteOriginally posted by wildman Quote
I would think even the kit lens would do better than this.
The kit lens would do better, a lot better in fact. I suspect that Sears lens is a bad copy. Even a cheap zoom lens would do better than that (at least at that focal range).
08-29-2012, 05:44 AM - 1 Like   #40
Pentaxian
bdery's Avatar

Join Date: Mar 2008
Location: Quebec city, Canada
Photos: Gallery
Posts: 5,830
QuoteOriginally posted by altopiet Quote
Do I understand what you say correctly, that the lens resolution will differ between 12mp K-x and 16mp k-5 sensors
The lens resolution will not change but the lower resolution sensor might not be able to show you the full details of the image created by the lens.

Just look at the resolution figures at Photozone, they are retesting manny lenses with the K-5 sensor, previously they had tested the lenses with the K10D sensor.

QuoteOriginally posted by altopiet Quote
which will influence image sharpness taken with the same lens on both cameras?
That part is correct.

QuoteOriginally posted by altopiet Quote
is there any simple way to find out at what mp a certain lens will reach optimal sharpness

snip

If there was a simple way for me to know that, it would prevent me from wasting money, on especially old lenses, knowing that a certain lens can not be used to it's optimal capacity on my 12mp camera.
There is an "absolute" measurement for resolution for a lens, it's called MTF (modulation transfer function). It's the capability of a lens to resolve a transition between fully black and fully white. It's a pretty universal measurement, so much so that you can ask your Pentax camera to always select the sharpest setting when in P mode.

However, MTF is NOT a complete measure of sharpness. Sharpness (our perception of resolution) also is influenced by contrast, saturation, DOF, bokeh, etc. It's not a simple thing.

For older lenses, if you can find an evaluation of MTF you can compare them with newer lenses. Older lenses will probably be quite sharp as a rule, it's relatively easy to get good sharpness. But aberrations (chromatic aberration, coma, halos, for instance) will occcur much more often. The best example I know is the Vivitar series 1 70-210, a legendary lens. I have owned version 1, it was an amazingly sharp lens, with great colour and smooth bokeh, but it showed a lot of chromatic aberration.

I hope I made things a little clearer

QuoteOriginally posted by Lowell Goudge Quote
Normally when discussing depth of field and circle of confusion we are discussing things that occupy more than one pixel, we are discussing large objects occupying much of the frame..
That's the point. When the CoC is larger than one pixel, you can see its effect, and the perceived resolution is lower.
08-29-2012, 06:41 AM   #41
Site Supporter
altopiet's Avatar

Join Date: Sep 2010
Location: The Great Karoo, South Africa
Photos: Gallery | Albums
Posts: 1,927
Thanks northcoastgreg and bdery, I get it now
08-29-2012, 06:45 AM   #42
Pentaxian




Join Date: Jun 2008
Location: Flyover America
Posts: 4,475
QuoteOriginally posted by northcoastgreg Quote
I suspect that Sears lens is a bad copy.
I wouldn't be so sure. I have a no name zoom I bought back in my Nikon film days and its even worse than your Sears.

Thinking further about the Rockwell post. I think all he was saying is that there is certain practical limits set by the way we actually use a lens in the real world and that most modern lens' more or less meet this limit especially with respect "sharpness" - whatever that means.

A case in point:
I have a FA35 F/2. For pure raw optical quality it meets my upper limits of optical performance especially sharpness for my purposes. In other words if all the lenses I have and all the lens' I ever will have is as sharp as the FA35 I would be a happy camper. Gong beyond this upper limit of optical performance would be empty optical performance with no real world practical advantage for me. The only possible exception I can think of may be my Sigma 105mm macro when using it specifically for macro work only.

FA35 at F/2, ISO 3200, available light only, handheld full frame:

Last edited by wildman; 09-09-2012 at 06:00 AM.
08-29-2012, 07:44 AM   #43
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,400
QuoteOriginally posted by Lowell Goudge Quote
Normally when discussing depth of field and circle of confusion we are discussing things that occupy more than one pixel, we are discussing large objects occupying much of the frame..
QuoteOriginally posted by bdery Quote
That's the point. When the CoC is larger than one pixel, you can see its effect, and the perceived resolution is lower.
I think you are still missing something.
If you start back at either the circle of confusion for a printed photo, this is 1/100 of an inch using the old standard of how suitable sharpness was defined either for out of focus points of light, or for camera shake and the "rule of thumb" for minimum shutter speed to eliminate shake. This was considered an acceptable amount of blur on an 8 x 10 inch print. This leads to a circle of confusion on the sensor of 20 microns (if you mathematically rescale things back to the sensor) and is also consistent with the on line DOF calculators for the APS-C sensor

this means that a suitably "sharp" detail edge has a width of a minimum of 3 pixels on my *istD and considerably more on any other DSLR since.

We are NEVER discussing depth of field in the same consideration as a single pixel. sharpness at the pixel level is not something that is measureable except at the exact point of focus, because no where else, an you ever have an "infinitely thin" line than the point of focus.
08-29-2012, 09:31 AM   #44
Pentaxian




Join Date: Feb 2010
Location: Eureka, CA
Photos: Gallery | Albums
Posts: 1,977
QuoteOriginally posted by wildman Quote
Thinking further about the Rockwell post. I think all he was saying is that there is certain practical limits set by the way we actually use a lens in the real world and that most modern lens' more or less meet this limit especially with respect "sharpness" - whatever that means.
That's part of what he is saying; and although there is an element of truth in his premises, his conclusion is exaggerated and misleading. Yes, photographers nowadays obssess too much about measured resolution, whether in terms of lens or camera performance. But that doesn't mean that most photographers can't produce better images with better glass. Any good photographer can take (on average) better images with lenses that are faster and/or have better color rendition than the typical kit lens. And even a bad photographer can accidentally take a compelling shot; in which case, better glass will lead to a better image. I just sold a 20 x 30 inch print taken with the FA 24-90. I would not have made that sale if the image had been taken with the kit lens. The 24-90 has the best color rendition of any zoom lens I've ever used. In landscape photography, that can be a critical X factor. People respond to images with bright, rich, vivid distinctive color. Post-processing can help a little bit, but generally speaking, potential buyers don't go for over-saturated, over-manipulated images.

I think Rockwell's argument is more compelling when it is applied to a higher order of lenses: for example, to the differences between mid-range and "professional" offerings. Lenses such as the DA 35/2.4, DA 50/1.8, FA 35/2, although not quite as superlative as the DA 35/2.8 Limited, the DA* 55/1.4, or the FA 31, are still good enough for the needs of most photographers. A very talented photographer can produce images from an old M 50/1.7 lens costing $40 that look nearly as good as anything produced by a professional photographer using pro-level glass. But that's partly due to the fact that that old 50 is a pretty good lens in it's own right, significantly better than most kit glass.
08-29-2012, 11:03 AM   #45
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,400
on the whole, while I understand what northcoastgreg is saying, I do have some comments, which I have inserted below

QuoteOriginally posted by northcoastgreg Quote
That's part of what he is saying; and although there is an element of truth in his premises, his conclusion is exaggerated and misleading. Yes, photographers nowadays obssess too much about measured resolution, whether in terms of lens or camera performance. But that doesn't mean that most photographers can't produce better images with better glass. Any good photographer can take (on average) better images with lenses that are faster and/or have better color rendition than the typical kit lens.
i would have to disagree with faster here, faster lenses are not always as sharp and unless you are shooting at an aperture that is not available with the slower lens, open aperture speed is not an image changer once stopped down. Color rendition is a lens artifact, not a design criteria, some lenses are better or produce more pleasing colors than others, true, but it is not a question of simply paying more $ for better color
QuoteQuote:
And even a bad photographer can accidentally take a compelling shot; in which case, better glass will lead to a better image.
not necessairly, only if the better glass is used in a way that the acceptable glass cannot be
QuoteQuote:
I just sold a 20 x 30 inch print taken with the FA 24-90. I would not have made that sale if the image had been taken with the kit lens. The 24-90 has the best color rendition of any zoom lens I've ever used. In landscape photography, that can be a critical X factor. People respond to images with bright, rich, vivid distinctive color. Post-processing can help a little bit, but generally speaking, potential buyers don't go for over-saturated, over-manipulated images.
agreed, but other lenses may also be able to take that shot, just not the present kit lens, with the same colors
QuoteQuote:

I think Rockwell's argument is more compelling when it is applied to a higher order of lenses: for example, to the differences between mid-range and "professional" offerings. Lenses such as the DA 35/2.4, DA 50/1.8, FA 35/2, although not quite as superlative as the DA 35/2.8 Limited, the DA* 55/1.4, or the FA 31, are still good enough for the needs of most photographers.
this is quite true, once you move away from true consumer lenses, the gap of performance per $ spent is extremely narrow, and middle range lenses can often produce excellent results, that is why they are so popular
QuoteQuote:
A very talented photographer can produce images from an old M 50/1.7 lens costing $40 that look nearly as good as anything produced by a professional photographer using pro-level glass. But that's partly due to the fact that that old 50 is a pretty good lens in it's own right, significantly better than most kit glass.
that old 50/1.7 was already the middle upgrade of the kit lens when film cameras were sold. the basic kit lens was a 50/2, the 50/1.7 was an upgrade, and the 50/1.4 was the pro lens of the day. in many ways, the 50/1.7 outperforms the 50/1.4 because the larger glass introduces problems with respect to sharpness etc due to the larger diameters of the elements and thicker (by necessity to maintain the same surface profiles over larger diameters) glass. People bought 1.4 lenses for wide open shooting, not for ultimate sharpness. in this instance using a 50/1.4 (or 1.2 even) at F2 was not as good as a 50/1.7 at F2.
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!
k-mount, lens, pentax lens, sharpness, slr lens
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Novice lens/exposure/sharpness question anselesn Troubleshooting and Beginner Help 12 08-22-2012 11:26 AM
Sharpness, Fine Sharpness, Extra Sharpness Taviali Pentax K-5 8 03-31-2012 12:56 PM
How Do I Read/Use The Numbers/Colors On My Sears 60-300mm Lens? TechGuru Pentax SLR Lens Discussion 5 12-07-2011 05:21 PM
Super A unable to read the minimal aperture from A lens ligi Troubleshooting and Beginner Help 5 01-12-2011 02:07 PM
Fine sharpness and sharpness move together on K20D 1.01 morfic Pentax DSLR Discussion 2 07-11-2008 09:18 PM



All times are GMT -7. The time now is 09:36 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