IR Comparison 590nm to 1,000nm (Converted and Unmodified)

KBeezie

Mu-43 All-Pro
Joined
Sep 15, 2012
Messages
1,393
Location
Grand Rapids, Mi
Real Name
Karl Blessing
Since I had most of my filters out with me at one of the parks, I found a bench and set up two cameras on the tripod. I did not have my Olympus E-M5 (mk1) with me for the unmodified comparison (which I think has a similar sensor to the E-M1 Mk1). So for the unmodified set I used my Pentax Q10 which seems to be about a stop more sensitive to IR than the E-M5 (but not as effective for faux color as my E-M5).

A note about any filter below 700nm; on an unmodified sensor that still has the "Hot Mirror" (blocks Infrared above 700nm and Ultraviolet light below 400nm typically), colored filters or "contrast filters" as they are typically called for black and white photography will usually be pretty useless. They let in too much visible light to allow for any infrared light to register on the sensor.

One thing I noticed about the Pentax Q10 is that even up to 1,000nm it will still yield an image, a crappy image, but still a recognizable image.

Both cameras were shot in manual mode, and in raw mode. On the E-M1 the white balance was kept the same thru all the shots, with some changes in adobe camera raw for highlights/shadows, and modified the red/blue/green curves as needed in the 'point curve' (the higher IR filters pushes more of the blue curve with the initial white balance, whereas below 850nm the red tends to be higher). Once loaded in photoshop, every image has the exact same post-processing treatment done to it, with the top of the silo dome setting the color balance using the level's adjustment layer (middle eye dropper clicked onto the white of the silo dome).

Today was pretty much 'mostly cloudy', but not quite overcast which can have a small impact on the infrared light. I'm hoping tomorrow will be sunnier because then I can do the same comparison again, but this time around with the same lens on the E-M5.

One thing to take note of between the converted and unmodified results is the noise and general quality. But it's also why I want to do it again with the E-M5, since the Q-10 is a 1/2.3" sensor that is Back Side Illuminated (which usually gives comparable results to my E-M5 when the ISO is kept below 400, and the aperture below 5.6).

The Filters

B+W 091 "Dark Red" (630nm)
Transmission starts at about 600nm, tops out at 92% by 650nm. This is not an infrared filter, but rather a color filter for black and white photography. I also have the B+W 090 (590nm, roughly orange/light-red) which likewise transmits well into the infrared spectrum allowing for full spectrum and converted IR (of a wider nanometer) cameras to have a strong faux color. I don't even bother to pack the 090 because my E-M1 already has a 590nm glass directly in front of the sensor. Sometimes called an R29 in other brands, with a 090 being an R25 in other brands.

B+W 092 "Deep Red" (695nm)
Blocks visible light up to 650nm, 50% transmission just below 700nm, 730nm to 2,000nm is above 90%. Usually the lowest nanometer filter that can be used on an unmodified camera for faux-color infrared. Any lower than that (such as with the B+W 091 or 090 contrast filters), and the visible light is too strong compared to the infrared light to yield any usable result.

Probably my favorite to use normally for 'faux' color without being too overpowering on the visible light side. And the only one I been using on my E-M5 ditching the R72 until I got the converted E-M1.

Hoya R72 (720nm)
starts around 720nm, reaching about 95% at 750nm and dropping off around 870nm until 1,000nm. 720nm is often called "Standard Infrared".

Typically very muted false color ability and the typical infrared filter you would use on an unmodified camera to get most of the infrared light the sensor can see. Still sees some visible light, but the transmission is cut low enough to not overtake the IR source.

Tiffen 87 (~ Wratten 87C)
Transmissions start around 720nm, 830nm at 50%, 1,100nm at 80%, peaks at 85% around 1,600nm which is well beyond a converted sensor's 1,200nm threshold. (sensor sensitivity drops sharply around 1,100nm).

Unlike most other Wratten 87C filters out there, it does not have a steady transmission curve, it's kind of a bumpy hill. Transmission is relatively low compared to other brands like the B+W 093 or the Hoya R90, but it's also a little cheaper (some new-old stock versions can be found at half the price of the competitors). The copy I have is likely a 1980s new-old-stock, the coating it seems to use is a bit speckled and they're not known to age gracefully they also exhibit a weird behavior such as transmitting white for florescent tubes when they're supposed to be red.

Schott RG850 2mm thickness (850nm)
Cut and mounted in a 52mm circular filter by UVIROptics, cuts all light below 850nm, transmission increases sharply from there to about 98% at 900nm, and bout 99% from 910nm onward. Much sharper/cleaner than the Tiffen 87 with a much better transmission curve.

This one is going to quickly be my go-to black-n-white infrared filter especially since the transmission is high enough for normal handheld usage with the converted camera, but narrow enough into the infrared field to make the bayer dyes transparent (making the red, green and blue photosites closer to equal exposure from the infrared light without influence from the visible light blocked).

Schott RG1000 2mm thickness (1000nm)
Also cut/mounted by UVIROptics into a 52mm threaded circular filter. Cuts all light below 800nm, transmission increases slowly from there to about 50% at 968nm, reaching about 92% at 1,200nm. Very similar visual results to an RG850, but with more exposure time required (roughly 3 to 4 stops more than an RG850 on a converted camera). Can be useful to get longer exposure results, but for handheld usage and higher overall transmission past 850nm the RG850 is a much better choice.

Photos

For the Imgur Album for the E-M1 results : Infrared Filter Comparison (Converted E-M1) All ISO 200, f/4
For the Imgur Album for the Q10 Results (which also include normal visible, X4 CPL, and X4 10-stop ND) : Filter (IR/etc) Comparison on Pentax Q10 + 01 Prime

Olympus E-M1 (converted), Pentax-M 50mm f/1.4
All shots at ISO 200 and f/4 only exposure time changes between shots. Exposure changed to keep +1EV on the meter reading since the sensor reads red higher than the other channels for most of the situations, you have to compensate or you end up with some underexposure for most the image, especially if you're trying to maximize the data captured by the raw file.

Attached photos are smaller size than shown at Imgur

Unfiltered (590nm IR on the sensor) - 1/1,250th
WdsTxkNh.jpg
Subscribe to see EXIF info for this image (if available)


B+W 091 (630nm) - 1/1,000th
yvJ4P3wh.jpg
Subscribe to see EXIF info for this image (if available)


B+W 092 (695nm) - 1/1,000th
LCiOHyUh.jpg
Subscribe to see EXIF info for this image (if available)


Hoya R72 (720nm) - 1/1,000th
xd021Hlh.jpg
Subscribe to see EXIF info for this image (if available)


Tiffen 87 (~830nm) - 1/500th
UWBTQ1xh.jpg
Subscribe to see EXIF info for this image (if available)


Schott RG850 2mm - 1/250th
3bFm1GFh.jpg
Subscribe to see EXIF info for this image (if available)


Schott RG1000 2mm - 1/50th
Cugd9V7h.jpg
Subscribe to see EXIF info for this image (if available)


Pentax Q10, 01 Standard Prime (47mm equivalent)
All shots are at ISO 100, f/2.8, except the very last one bumped to ISO 125 to keep +0EV on the meter reading.

Unfiltered, 1/1,250th ( note the huge mass of mosquitoes near the left tree :eek: )
i6J7XUih.jpg
Subscribe to see EXIF info for this image (if available)


Breakthrough X4 Circular Polarizer, 1/500th
6E1CzGah.jpg
Subscribe to see EXIF info for this image (if available)


Breakthrough X4 10-stop Neutral Density, 1.3 seconds
VrgCo5Rh.jpg
Subscribe to see EXIF info for this image (if available)


B+W 091 (630nm), see note at top regarding Black and White contrast filters on non-converted cameras. 1/60th
OlTFqYSh.jpg
Subscribe to see EXIF info for this image (if available)


B+W 092 (695nm), 1 second
fm20FdGh.jpg
Subscribe to see EXIF info for this image (if available)


Hoya R72 (720nm), 2 seconds
HSrlraxh.jpg
Subscribe to see EXIF info for this image (if available)


Tiffen 87 (~830nm), 8 seconds
Py5LJeyh.jpg
Subscribe to see EXIF info for this image (if available)


Schott RG850 2mm, 8 seconds
dIupuoZh.jpg
Subscribe to see EXIF info for this image (if available)


Schott RG1000 2mm, 30 seconds (maximum time for camera), ISO bumped to 125
qV0MtB6h.jpg
Subscribe to see EXIF info for this image (if available)


Older Results from E-M5

Some older pictures from the unmodified E-M5 using the filters above for comparison until I can do an actual side by side.

All with a 1951 Canon Serenar 35mm f/2.8, all the images shot in RAW and have the exact same post-processing steps (no channel swaps on the normal visible light shot).

Unfiltered, ISO 200, f/4, 1/160th
Hme2EPsh.jpg
Subscribe to see EXIF info for this image (if available)


B+W 092, ISO 200, f/4, 4 seconds
5O025snh.jpg
Subscribe to see EXIF info for this image (if available)


Hoya R72, ISO 200, f/4, 8 seconds
A7M05bth.jpg
Subscribe to see EXIF info for this image (if available)


A different approach to the B+W 092 image with some more dramatic changes in the point curve in adobe camera raw before opening in photoshop.
zgfWZmgh.jpg
Subscribe to see EXIF info for this image (if available)
 
Last edited:

KBeezie

Mu-43 All-Pro
Joined
Sep 15, 2012
Messages
1,393
Location
Grand Rapids, Mi
Real Name
Karl Blessing
A few RG850 shots from the same afternoon.

JYxoMqV.jpg
Subscribe to see EXIF info for this image (if available)


pWQ5LGw.jpg
Subscribe to see EXIF info for this image (if available)


Shot thru the glass since the door to the observation deck at top was locked today.
RC06R96.jpg
Subscribe to see EXIF info for this image (if available)


Nvc1l06.jpg
Subscribe to see EXIF info for this image (if available)


aDQvXW7.jpg
Subscribe to see EXIF info for this image (if available)


MVyW9X2.jpg
Subscribe to see EXIF info for this image (if available)


And all the rest here are multiple-frame composites (resulting in a large composite varying from 65 to 125 megapixels)

KZ2mk0Z.jpg
Subscribe to see EXIF info for this image (if available)


0qMc24l.jpg
Subscribe to see EXIF info for this image (if available)


ENlsHKP.jpg
Subscribe to see EXIF info for this image (if available)
 

Latest posts

Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Forum GIFs powered by GIPHY: https://giphy.com/
Copyright © Amin Forums, LLC
Top Bottom