Schott IR-Pass Filters: 15 Views of the White-balanced Color Checker Card

[Andrea B.]

See also: Schott IR-Pass Filters: 15 Views of the Raw Color

 

Edits: I rewrote the intro here so that when the series is complete, it will all hang together better.

Title also slightly edited.

---

 

Experiment: Investigate the white balance of 15 IR-pass filters.

 

Gear: D610-mod + UV-Nikkor 105/4.5 and 15 IR-pass filters.

 

Filters: I bought a nice set of 16 Schott 2mm thick IR longpass filters from Uviroptics on Ebay. I did not use the RG-9 IR bandpass filter in this series. The filters are labeled as follows:

• Orange + Red + IR Group
  • OG515
  • OG530
  • OG550
  • OG570
  • OG590

  [*]Red + IR Group

  • RG610
  • RG630
  • RG645
  • RG665
  • RG695
  • RG715

  [*]IR Only Group

  • RG780
  • RG830
  • RG850
  • RG1000

.

Question: I wonder why is there no IR filter between RG715 and RG830? RG780?

 

Subject: Color Checker Card together with a 99% and a 75% reflective Labsphere Spectralon white standards. In this topic only the CC Card is used. Here's the OG515 raw color file so you can see what I was working with.

 

 

Processing:

Each IR file was converted in Capture NX2. After setting black point and white point, the white balance was made by drawing the largest possible WB marquee over the 99% reflective Labsphere Spectralon standard. Cropping, labeling and the composite were made in PS Elements 11.

 

This compsite is 964 x 1204 px. The forum software resizes it to a max of 1200 px. (Qué tan estúpido es eso?)

 

 

 

---

So I got to wondering how other conversions might look. For the first photo in the series made under the OG515 filter (top right in composite), I made additional conversions using RGB Render Auto from Raw Digger, Dark Table, and 3 from Photo Ninja. I made a few minor white/black point, minor contrast and minor saturation adjustments after the initial conversions so that the photos would be somewhat alike in those areas. For example, the RD raw colours have a very low saturation so I boosted sat there.

 

You will right away note the color differences across the conversions. This kind of thing drives me rather crazy sometimes because it makes me want to ask the slightly silly question: which conversion has the most accurate false colours? Should that question even be asked given that we are dealing with false colours? Prolly not!

 

For the standardization of my own work, I have chosen to go with a Photo Ninja conversion that uses a color profile made with for modded camera in use. But YMMV and all that. The point is: don't let conversion differences bug you too much in the false colour area. And be aware that in these kinds of filter comparison tests you can always turn the false colour wheel to see a different false colour palette.

 

 

Nikon Capture NX2

 

Photo Ninja: D610 Color Profile

The visible color profile was made under a Baader UV/IR-cut filter.

 

Dark Table

I don't know quite enough yet about DT to get the best edits following conversion. Working on it.

 

Photo Ninja: Daylight

 

Raw Digger: RGB Render with Auto Option

 

Photo Ninja: No Profile

I assume that this is PN's idea of the raw colours? Don't really know.

[ulf]

Question: I wonder why is there no IR filter between RG-715 and RG-830?

 

¿Que?

• 
  
  • RG715
    

  [*]IR Only Group

  • RG780 <====
    
  • RG830
    

[Andrea B.]

oh la! I should have written this instead: why is there no IR filter between 715nm and 780nm. Seems like a large gap.

[Cadmium]

There is no Hoya filter between R-72 and R-760 either.

Here is the progression if you include both Schott and Hoya brands together:

(by the way, Schott glass types have no "-" in their names, Hoya has the "-" in their names)

 

695nm longpass: Schott RG695 @3mm

700nm longpass: Hoya R-70 @2.5mm

715nm longpass: Schott RG715 @3mm

720nm bandpass: Schott RG9@3mm

720nm longpass: Hoya R-72 @2.5mm

760nm longpass: Hoya IR-76 @2.5mm

780nm longpass: Schott RG780 @3mm

800nm longpass: Hoya IR-80 @2.5mm

830nm longpass: Schott RG830 @3mm, and Hoya IR-83 @2.5mm

830nm bandpass: Hoya RT-830 @2.5mm

850nm longpass: Schott RG850 @3mm, and Hoya IR-85 @2.5mm

960nm longpass: Hoya RM-90 @2.5mm

970nm longpass: Schott RG1000 @2mm

1000nm longpass: Schott RG1000 @3mm, and Hoya RM-100 @2.5mm

[Cadmium]

By the way, I really like your longpass color checker tests. Really shows where the Bayer false colors equalize, right about 830nm, 850nm.

I always use 850nm personally, it seems to look more monochrome to me.

[Andrea B.]

It is interesting where there are "breaks". Between 570 and 590 the magenta goes and the greens start to show up in greenish blue. Then after 665 there are no more greenish blues. At 780 there's not much blue left, just a tint.

 

I took the post down temporarily to remove the dashes, but it was too much work. So I'll have to leave them. Perhaps the dashes make the labels a bit easier to read. :D

[bobfriedman]

i use the dashes.. i never really thought about it until Steve pointed that out.

[Cadmium]

It is interesting where there are "breaks". Between 570 and 590 the magenta goes and the greens start to show up in greenish blue. Then after 665 there are no more greenish blues. At 780 there's not much blue left, just a tint.

 

Yes, very interesting. These are very well done tests.

I took the liberty of swapping the channels around on the set.

 

This first one shows an EIR Photoshop action applied to the first three.

I am putting at the top so it is easier to compare to the top row of the larger one.

 

The standard red/blue channel swap shows which ones have red after the swap.

Some color is still visible in the RG780 shot.

I see little difference in the RG830, RG850, and RG1000, except I see some difference in the whites of the RG830 compared to the RG850 and RG1000.

I can't say I see any differe3nce between the RG850 and RG1000, in these tests at at least.

[ulf]

Yes, very interesting. These are very well done tests.

I took the liberty of swapping the channels around on the set.

 

This first one shows an EIR Photoshop action applied to the first three.

I am putting at the top so it is easier to compare to the top row of the larger one.

 

Thanks Steve. That was an interesting addition to the post.

 

The recommended filter for analog EIR-film was a Tiffen 12 filter.

If has a transition wavelength close to 530nm if I remember correctly.

The transition is also less steep than the filters from Schott and Hoya.

[JMC]

Love the set of image of the colour charts - so clearly shows the transition to monochrome at the longer IR wavelengths.

 

I started the see the effect with my work on the Bayer filter measurement;

 

http://www.ultravioletphotography.com/content/index.php/topic/2764-bayer-filter-transmission-in-uv-and-ir/page__view__findpost__p__21122

 

In the 3rd graph on the first post, the blue and green transmission is starting to rise rapidly up towards the red at 800nm which is as high as I can measure. Once they are all the same then the final image will be obviously be monochrome. Given the rate of rise in my graph it would make sense that 'monochrome' IR would happen at around 830nm to 850nm and above.

[Andrea B.]

This next topic illustrates what Cadmium was trying to show, I think. (Added Later: well maybe not....)

Schott IR-Pass Filters: 15 Views of the Raw Color

 

 

I slightly changed the name of this original topic so that the final series names are similar.

 

Stay tuned for Schott IR-Pass Filters: 15 Views of a Seascape which is currently in progress.

[Cadmium]

Thanks Steve. That was an interesting addition to the post.

 

The recommended filter for analog EIR-film was a Tiffen 12 filter.

If has a transition wavelength close to 530nm if I remember correctly.

The transition is also less steep than the filters from Schott and Hoya.

 

Here is the rough comparison graph I made a while back showing Tiffen #12 and others in that range.

#12 seems to cross 520nm at 50%, and there doesn't seem to be any exact equivalent, but I use various filters for that in that range.

[Cadmium]

i use the dashes.. i never really thought about it until Steve pointed that out.

 

Sometimes you feel like a dash, sometimes you don't. Hoya's got dashes, Schott don't.