Focussing Monochrom with filters

[pico]

Advertisement (gone after registration)

Not at all the same. The only comparison would be to Hasselblad’s multi-shot backs where the sensor can be shifted in both dimensions in pixel-pitch steps. Four shots deliver a full set of RGB data at the nominal sensor resolution, rendering the demosaicing step superfluous.

 

Here is where a hot pixel might be useful. For registering numerous shots one can use the absolute location of the hot pixel and then in Photoshop (for example), use the Filter-Other-Offset to shift the frame to a standard location. (The Info window can give you the absolute location of the hot pixel.)

 

Lindolfi: Well, I just built a different setup: a microscope with a micrometer looking at the aerial image

 

Very cool.

 

That reminds me that I have a heavy micrometer controlled aerial focusing telescope which can focus from several inches away. God, it's been sitting here for decades. I'll have to play with it.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Lindolfi]

Here are the first results: The object distance was 1.0 meter. The two lenses tested were the 50/1.4 Summilux ASPH and the 90/2.8 Elmarit (early). The focal shift from a green to a red filter of the 50 mm was less than 0.05 mm, which is close to the repeatability of the measurement. The focal shift from green to red of the 90/2.8 was about 0.2 mm.

 

The focal depth of the 50/1.4 at f/1.4 at 1 meter is about 0.022 mm given the resolution of the M Monochrom, so there may or may not be a need for correction.

 

The focal depth of the 90/2.8 at f/2.8 at 1 meter is about 0.05 mm at the M Monochrom resolution, so a correction is certainly needed there when using filters, since the focus shift is four times the focus depth.

 

Interesting, since this makes it even more worthwhile to invest in modern lenses with the Monochrom if resolution is really important and red filters are applied.

 

Curious to see what data Michael comes up with in his article in LFI...

[Hosermage]

The focal depth of the 50/1.4 at f/1.4 at 1 meter is about 0.022 mm

 

Are you serious? 0.022 millimeters? It's a wonder if anyone ever catch things in perfect focus.

[Lindolfi]

Focal depth is fhe tolerance at the focal plane (sensor or film side), which is different from depth of field (tolerance at object side), which at 1 meter for a 50/1.4 at f/1.4 is about 7 mm at the resolution of the M Monochrom.

[luigi bertolotti]

Yes Luigi, that is possible. The three exposures should be perfectly registered (no camera motion down to less than a pixel) and nothing should move in the scene. Could be fun and like most fun things of little practical value

 

No doubt... using a MM like a COLOR SCANNING BACK is someway surrealistic... ... but I see that others have got this idea...

[batmobile]

I tried the following with my monochrom using a distant subject about 40m from my balcony, and inspected the files closely:

 

Yellow filter: As near as dammit to no filter in terms of resolution at the intended point of focus.

Orange-Red: Very slight reduction in resolution at point of focus. Noticeable when repeated.

Deep Red: Fuzzy image. Very significant focus shift.

 

I shot at F4 to ensure that focus shift would show up more clearly than F11.

 

I then casually focus bracketed with the deep red and could not get it much better so need to experiment further to see if it will be possible to use such filters on the camera.

[jaapv]

Advertisement (gone after registration)

Use a DOF mark two stops down as a focus marker for a starting point. (or the red R if your lens has one.)

[batmobile]

Two stops down in which direction?

[jaapv]

To the right.

[batmobile]

OK, I have done some testing with:

 

MM

35 Summarit at f2.8.

B&W #91 Deep Red MRC filter.

Focus distance was about 20m, using sharp leaves on a plant for easy reference. Being on the first floor helped as I could see where focus was falling across the garden.

 

Compared to MM without a filter, the shots with the 091 filter were VERY soft at the point of focus without focus adjustment. Completely unusable. Maybe at F16 they would work, but who wants such a crude workaround.

 

I bracketed focus in much smaller increments than my first lazy test by reducing the distance the lens was focused at. I did this, because I could see in the shots with the 091 filter attached, that the far distance was much sharper than the unfiltered shots. This showed that the 091 filter was causing the camera to focus much further away than the rangefinder suggested.

 

I found my les needed to have the focus distance reduced by about 2/3 of a stop i.e. I focused the lens and looked at where the 'wide open' indicator pointed on the distance scale. I then reduced the focus distance until that mark on the distance scale fell between the central wide open mark and the F4 mark, but leaning more towards the latter. With no clear reference on the distance scale and it all being very 'eyeball' I repeated it several times to confirm. Going one full stop, i.e. to rest the F4 mark at the distance of the object, resulted in clearly inferior results. At a hair under a full stop and 1/2 a stop, the image at 1:1 was a tiny bit less sharp than when about 2/3 of a stop was used, so this is clearly where one needs to be for perfect results.

 

Please note that the above was at f2.8, so any error will reduce quickly as you stop down and you are not likely to be shooting landscapes with a 091 filter at f2.8.... But the bottom line is, if in doubt, bring the focus closer in a bit more than needed rather than less, as more DOF lies behind the point of focus than in front anyway.

 

If you are vaguely in the ballpark of 1/2 a stop to a stop closer in I can confirm that by f5.6 you will see no difference anyway, so by f8-11 for landscapes this is not something that has to be all that precise.

 

Problem solved. Its only for the 091 filter that any of this matters. DOF takes care of the tiny shift with yellows and oranges anyway, although a red filter, wide open with a long fast lens would likely be an issue. For my uses of the MM (documentary and street with nothing more than a yellow or perhaps orange) and landscapes (maybe then rarely using the 091) there is no issue that is not easily sorted out.

 

Hope this helps.

 

PS you also need to raise the exposure by about 2/3 to 1 stop when using TTL metering and the 091 filter and the meter is evidently more sensitive than the sensor with this filtration. One would be likely to bracket anyway for landscape work with such a strong filter, but this will get you right into the right area.

 

As a dedicated B&W shooter, I love this camera! I shoot lots of digital for more commercial/non-personal work and those cameras are just tools. This is the first digital camera I actually enjoy using. Who wants an M240?