Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-15-2012, 07:39 PM   #16
Veteran Member
creampuff's Avatar

Join Date: May 2007
Location: Singapore
Photos: Gallery
Posts: 3,955
Can you give specific details of the camera, lens used, filter and exposure times? There are so many variables why your shots aren't sharp (eg. long exposure, subject movement, camera shake, etc.) Also your image conversion is not mentioned. Strangely, for an IR shot, none of the leaves of your trees and grass are light toned.

I have been shooting IR images for quite a while and I've not had the sharpness loss you've encountered.





07-16-2012, 09:34 AM - 1 Like   #17
Inactive Account




Join Date: Dec 2008
Location: Ames, Iowa, USA
Photos: Albums
Posts: 2,965
Here's a time-of-day discussion: Best Time of Day/Lighting Conditions to Shoot Infrared - Photo.net B&W Photo - Film & Processing Forum by Helen Bach , May 15, 2007; 12:41 p.m.

Broadening of the in-focus diffraction disk in proportion to IR wavelength is certainly an issue and diffraction softening will be a problem at around half the f-stop for normal light. This is a general problem and the longer the wavelength of the radiation, the greater the broadening.

There are three main actors in this play: the sun's light, the lens focus of the particular light and the sensor's response to it. First, a typical sensor's response:


Your IR filter probably lets radiation from about 700nm through. The Red component of the Bayer filter responds to all the wavelengths above 700nm while the green & blue components don't begin to respond until about 800nm is reached.

Now look at the typical solar radiation vs time of day:


The cyan curve shows the sun's brightness at various wavelengths at sunup (sundown) while the blue curve shows brightness at about 10:40am. Clearly the fraction of 700-800 nm illumination is much greater at mid morning than when the sun is low in the sky. Such 700-800nm light will be "seen" by the red Bayer sensor but not by the blue or green sensors.

I expect that focus is much worse in the 800+nm illumination band than in the 700-800nm band for two reasons: first is the lens was designed to work for a mean around 550 nm & the greater the deviation from that wavelength the worse its sharpness will be, second, the limiting blur for a lens is proportional to the wavelength used.

If this these phenomena are causing your problems (and they are certainly contributing to time-of-day variations,) using an external filter that cuts off IR above about 800nm would be helpful. Similarly, you might just use the red portion of the spectrum for the image to be displayed.

Dave in Iowa

Last edited by newarts; 07-16-2012 at 09:41 AM.
07-16-2012, 09:54 AM   #18
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster
Well, you IR shooters seem like first rate science togs to me.

I'm only into the look, not the science. But a 'little' science is good to know.

THANKS!!
07-16-2012, 10:54 AM   #19
Inactive Account




Join Date: Dec 2008
Location: Ames, Iowa, USA
Photos: Albums
Posts: 2,965
The science details may be as they are but there is no obvious difference in the Red, Green, and Blue layers in the original posting:


Perhaps the HDR and other processing munged things enough to make all color channels be essentially the same.

Dave in Iowa

07-16-2012, 11:07 AM   #20
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster
Nah, I don't know Dave. I just did a HDR first time with IR. Sure, HDR may have made it less sharp for somne reason. But I noticed the same focus issues with non HDR IR.

IR just does not seem as sharp as non IR to me. Do you all feel that IR is as sharp as non IR?

I did some IR color slides about 40 years ago. They seemed pretty good with sharpness. But never made prints from em.

Will have to get the original to post here later today.

Last edited by slackercruster; 07-16-2012 at 11:25 AM.
07-16-2012, 11:24 AM   #21
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster


Original JPEG just reduced....fuzzy!

Maybe a little sharper than the 3 combo HDR, but nothing exciting any way your slice it.



Last edited by slackercruster; 07-16-2012 at 11:29 AM.
07-16-2012, 01:52 PM   #22
Inactive Account




Join Date: Dec 2008
Location: Ames, Iowa, USA
Photos: Albums
Posts: 2,965
Here's a further analysis of your more original photo. Separated into its three color channels:


Red is overexposed and has a large background haze component (might there have been a haze due to time of day?). This haze is likely the culprit in perceiving a lack of sharpness to the image. I don't know the origin of the haze. Green shows the best contrast (hence sharpness) but I wonder why the tree to the left of the church isn't whiter. The Blue image looks pretty good too.

I subtracted a constant background from the red image (90 of 255) then increased its contrast a bit to compensate. The channels were re-merged then converted to B&W. Here's the result


I think it looks a bit sharper.

You might check the red channel in other shots - maybe they all have a haze?

The simplest physical way to solve the problem might be to find a filter that blocks radiation up to about 750nm or to just not use the red channel at all. I use ImageJ freeware to separate and manipulate the individual color channels.

I hope this helps us in understanding what might be going on. I did not know about the different mix of longer and shorter IR wavelengths depending on time of day - but it makes sense.

I'm also impressed by the effect of the red over-exposure on background haze - maybe it is a result of the pre-shot white balance set?

Dave in Iowa
07-16-2012, 02:03 PM   #23
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster
Thanks for the outstanding rundown Dave!

I have not looked at the WB, will have to check it out.

07-16-2012, 02:30 PM   #24
Veteran Member




Join Date: Sep 2006
Location: West Chester, PA
Photos: Gallery | Albums
Posts: 1,420
I too have been shooting with an IR-converted 6mp Pentax, and my pictures are sharper than before the conversion.

Some thoughts:
Not all lenses do equally well in IR. Some suffer badly from haze (an IR "hot spot"), which is reflections from the glass. (Optical designers don't worry much about invisible reflections!)
Are you stopping down? For these types of landscapes I'm usually at f/11. (Note - stopping down lenses may show the "hot spots" more clearly.)
Some lenses are sharper in the center than on the edges. My Tamron SP 14mm f/2.8 gets weaker in the corners, so if it is critical I take several shots and make a pano. (Again, designers don't care how sharp the corners are for invisible light!)

A suggestion is to do focus bracketing, where you take a shot with the lens focused closely and take consecutive shots as you focus towards infinity. If none of them are showing 'peak' focus, and it is the same with all lenses, it is possible that your IR filter is softening things or something else (not the lens) is going on.

An example with my Tamron SP 14mm/2.8 - notice the edges are less sharp than the center, even reduced to this magnification:


My FA*24 is sharper corner-to-corner:


I've got more samples here:
Sean Carpenter – Photographs and Work

Edit to add: Shoot raw.
07-16-2012, 02:43 PM   #25
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster
Thanks Sean. Nice site you got!!

You find RAW comes in sharper?

I used a Sigma 8-16 zoom for the shots in question. Prior to that used a 15 Pentax and Samyang 8mm fisheye. Didn't pay much attention to stops. Will try some other lenses and stops.

From what I recall, the 15 was a little sharper. Esp in the early AM when it was cooler and the bright sun had just come out.

Last edited by slackercruster; 07-17-2012 at 06:32 AM.
07-16-2012, 03:02 PM   #26
Veteran Member
demp10's Avatar

Join Date: Jun 2011
Location: Atlanta
Photos: Albums
Posts: 602
QuoteOriginally posted by Asha Quote
I don't actually know how far into the IR the Pentax sensor sensitivity is. I just want to add a caution that people emit heat, which is basically what IR is. So, the mixing of reflected visible light with the emitted IR light from people's bodies will definitely not look as sharp as a "normal" visible spectrum photo.
There is a wide spectrum of IR from near-IR (0.7 m 1.4 m) used in IR photography, to thermal-IR (> 8 m ) which is actually emitted heat and used in thermal imagers.

Body head (thermal IR) should not be an issue in this type of IR photography. It will not even pass through the typical lens glass. To capture thermal IR you need special sensors and lenses.
07-16-2012, 06:11 PM   #27
Senior Member




Join Date: Mar 2012
Photos: Albums
Posts: 100
QuoteOriginally posted by demp10 Quote
There is a wide spectrum of IR from near-IR (0.7 m – 1.4 m) used in IR photography, to thermal-IR (> 8 m ) which is actually emitted heat and used in thermal imagers.

Body head (thermal IR) should not be an issue in this type of IR photography. It will not even pass through the typical lens glass. To capture thermal IR you need special sensors and lenses.
Visible light has a "thermal" signature, but since the sensors we typically use (eyes, typical cameras) see the visible, the thermal is not perceived. Lens glass does let NIR through, but you are right, many glasses have absorption in midwave and longwave. Even the atmosphere has absorption bands (H2O, CO2, for example).

And incidentally, NIR is most certainly emissive! Think of the Planck blackbody function--just because a lens does not let something 310 Kelvin through the system does not mean there is no heat being transmitted!

Last edited by Asha; 07-16-2012 at 06:19 PM.
07-16-2012, 06:43 PM   #28
Senior Member




Join Date: Mar 2012
Photos: Albums
Posts: 100
QuoteOriginally posted by carpents Quote
Not all lenses do equally well in IR. Some suffer badly from haze (an IR "hot spot"), which is reflections from the glass. (Optical designers don't worry much about invisible reflections!)
Good point, and there are actually a couple things to consider.

"Ghosting", would be present for both IR and visible if the design is prone to it. Ghosting is when reflected light inside an optical system converges, forming a ghosted spot of light. This would occur from light coming into the system from the scene being photographed. So, if you have visible light ghosting in a lens, you'll most likely get IR ghosting too. Since IR conversion involves removing a filter from the detector, then you have effectively altered the design, and you might get ghosting that wasn't present before the conversion.

Also, when the detector runs continuously and when you use the LCD a lot, heat builds up around the detector. Since the converted camera is more sensitive to IR, this heat will cause fogging inside the camera cavity. This would not likely cause ghosting since rays are not directed and are bouncing around in all directions, but it would definitely cause a reduction in contrast and likely lower the sharpness too. Plus, the sensor noise increases, which compounds the problem.
07-17-2012, 06:31 AM   #29
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster
I checked the WB. It was set to auto. Should I be using a custom WB for IR?

I tried the cam with monochrome and it still shot in color. Maybe it was disabled in the conversion.

Speaking of conversions.

Where do you guys get your color cams converted at?

Seems like lots of converters don't like to do Pentax for some reason. I bought mine for a few hundred $ already done from a private party just to get my feet wet with IR. Did not want to spend lots of $ to see what IR was about.

Thanks for ALL the help guys!
07-17-2012, 06:59 AM   #30
Veteran Member




Join Date: Jan 2012
Location: NE, USA
Posts: 1,302
Original Poster
What about using a polarizer filter? Will that help sharpness?
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!
ir, pentax help, pix
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Ron Paul: I helped a fuzzy dude cut a piece of fruit boriscleto General Talk 5 01-08-2012 09:26 AM
Macro Wittle Fuzzy Spider (And His Prey) joe.penn Post Your Photos! 8 09-19-2011 07:39 AM
Macro Some bugs and a fuzzy flower the swede Post Your Photos! 3 08-06-2010 02:04 AM
Fuzzy Corkscrews... heliphoto Post Your Photos! 13 09-11-2008 09:56 PM
a few fuzzy portraits hamidlmt Post Your Photos! 6 07-31-2007 03:23 AM



All times are GMT -7. The time now is 07:24 AM. | 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