i1d3 Teardown

[Pascal de Bruijn]

Hi,

I'm aware this is quasi-offtopic, but I thought some of you might enjoy it:

Regards,
Pascal de Bruijn

[Richard Hughes]

Very much, thanks. Choosing a 18F PIC was interesting (same family as
the ColorHug, and those data pins at the end sure look like an ICSP
programming port :) Using glass filters probably contributed to the
cost, but otherwise the board is very simple indeed. I'm kinda
confused about that giant lens tho, surely the beam needs to be
collimated from a narrow angle, although I don't know how the diffuser
would affect things.

Richard.

[Nikolay Pokhilchenko]

...Using glass filters probably contributed to the
cost, but otherwise the board is very simple indeed. I'm kinda
confused about that giant lens tho, surely the beam needs to be
collimated from a narrow angle, although I don't know how the diffuser
would affect things. I think the lens is needed for enlarging the instrument aperture to increase the light flow. The last is improving the sensors signal/noise ratio.

The glass filters seems to be interesting: they are semi-transparent mirrors with the coatings that reflects unwanted wavelength and passes the needed. Surely the filter coatings are multi-layer to achieve desired spectral transparency shape.

[Chris Lilley]

Hello Nikolay,

...Using glass filters probably contributed to the
cost, but otherwise the board is very simple indeed. I'm kinda
confused about that giant lens tho, surely the beam needs to be
collimated from a narrow angle, although I don't know how the diffuser
would affect things.
I think the lens is needed for enlarging the instrument aperture to
increase the light flow. The last is improving the sensors signal/noise ratio.

I suspect you are right. In photography, the f stop is the focal
length of the lens divided by the entrance pupil. The latter is the
apparent size of the front of the aperture as seen through the optics,
and its size (and position) can be quite different from the physical
size and position of the iris diaphragm.

I suspect a similar principle is in use here. Certainly it gives
accurate, fast results even on dark patches.

The glass filters seems to be interesting: they are
semi-transparent mirrors with the coatings that reflects unwanted
wavelength and passes the needed. Surely the filter coatings are
multi-layer to achieve desired spectral transparency shape.

That would also seem reasonable.

In passing I loved the comments made during the video about this
device incorporating a license key and being mainly designed to stop
people using their proprietary software because "software is more
important than keeping your customers happy".

Incidentally the proprietary software for I1D3 fights with the same
vendor proprietary software for the ColorMunki Photo :)

On my personal desktop machine I uninstalled the XRite drivers and use the
Argyll ones; on my more recent work laptop I have never installed the
XRite drivers or software; always used Argyll with better and more
flexible results.

[Graeme Gill]

Post by Richard Hughes
I'm kinda
confused about that giant lens tho, surely the beam needs to be
collimated from a narrow angle, although I don't know how the diffuser
would affect things.

The main lens focuses collimated light onto the diffuser, the second lens focuses
an image of the diffuser onto sensors. The diffuser ensures that the three
sensors see a uniform sample color, as well as eliminating any polarization.

Post by Richard Hughes
The glass filters seems to be interesting: they are semi-transparent mirrors with the
coatings that reflects unwanted wavelength and passes the needed. Surely the filter
coatings are multi-layer to achieve desired spectral transparency shape.

They're thin film dichroic type filters of course. See this article for an
idea of the type of thing X-Rite and others might have done to design such filters:

<http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.133.5284&rep=rep1&type=pdf>

Graeme Gill.