jdowning - 2-13-2013 at 01:05 PM
A well established technique used by museums to examine paintings for invisible details such as repairs or signatures otherwise obscured by layers of
paint is to photograph the painting under reflected infra red light. The same technique is used to examine documents that may be faded or damaged.
Curious to know if this technique might be applicable to the forensic examination of old oud labels otherwise difficult or impossible to read due to
faded writing or text obscured by dirt, glue or repair labels, I thought that it might be of general interest to run some trials to explore the
possibilities using low cost tools and equipment readily available to everyone.
Having no first hand experience with IR photography I have no idea if these trials will be successful but I shall go ahead anyway - step by step - as
the journey may turn out to be as interesting as arriving (or not) at the final destination.
More to follow.
jdowning - 2-14-2013 at 06:43 AM
The wavelength of radiant energy in the visible spectrum (the colours of a rainbow - red, orange, yellow, green, blue, indigo, violet)) ranges from
400 nanometers (violet) to 700 nanometers (red). The electromagnetic spectrum extends far beyond visible light, however, of primary interest here is
'near' infra red (just beyond visible red) with wavelengths from 700 nm to around 1000nm. At the other end of the visible spectrum is ultraviolet
(beyond visible violet) with wavelengths below 400 nm - also useful for forensic photography but not of direct interest for this investigation.
Prior to the introduction of digital cameras - reflected IR photography was a tricky procedure undertaken using IR sensitive film and suitable opaque
filters. The process was complicated by the need for long exposure times and required specialised facilities to develop the film.
The procedure has been made easier with the development of digital photography as the sensors used in digital cameras are IR sensitive to the point
where IR blocking filters (IR cut filters) are necessary for a camera to operate over the full range of the visible light spectrum (Full Spectrum)
when used for 'normal' photography.
In digital 'night vision' cameras (that can literally 'see in the dark' using invisible (to the human eye) IR illumination, the IR cut filter in the
camera lens is removed allowing the camera sensor to detect any IR reflected light. When used in daylight conditions (rich in both IR and UV energy)
and the camera is fitted with an 'over the lens' IR pass filter that blocks most visible light, these cameras can 'see through' fabrics and other
materials under the right circumstances. Whether or not they can see through a paper label remains to be seen.
So the first step in this investigation is to obtain a 'night vision' camera. I do not have one of these cameras but I do have some junk 'full
spectrum' digital cameras to play around with so will see what it takes to convert one of these to an IR camera.
Here is a brief article on the subject that gives an experienced account of the difficulties associated with IR photography. Even with the advent of
specialised IR digital cameras it would seem that it is not an easy matter to obtain good results due to high levels of 'chroma noise' inherent in the
digital images due to the use of dark IR pass filters and the correct choice of filter for a specific application.
http://www.mjbphotonetwork.com/2012/06/forensic-photography-reflect...
Danielo - 2-14-2013 at 12:28 PM
Hi John,
I think this would be complicated. Small B&W CCTV cameras sensitive to near-IR (700-1000 nm) are easy to find and relatively cheap (around $100).
However, in order to separate the different layers of text, you need to select specific wavelengths with band-pass filters. This is more costly !
Dan
jdowning - 2-14-2013 at 01:13 PM
Thanks Dan. This confirms the conclusions by the author of the article previously posted - and with narrow band pass filters costing around $100 or so
each this would certainly not be a low cost option.
But - with potential failure in mind as a real prospect for this investigation - lets move on regardless. I need to find out the hard way - always a
good way to learn!
jdowning - 2-16-2013 at 01:03 PM
By low cost I mean next to nothing - at least for the first part of this investigation - only using stuff that I have to hand. This is just to see if
IR imaging will work at a very basic level before getting more sophisticated.
The digital camera is a mini key ring 'spy' camera - a novelty item that I bought some years ago out of curiosity and now a little 'beat up'. It has a
0.3 Megapixel CMOS sensor with maximum image resolution of 640 X 480 pixels and fixed focus lens focussing from about 0.5 meter to infinity. Exposure,
white balance and focus are supposed to be automatic. It has no LCD screen to view the image but can be USB connected to a PC as a web camera to view
images through the camera. It can take both still and video images (of sorts).
No self respecting photographer would even consider this camera seriously so if some results can be obtained from this then things can only get
better!
These cameras can still be purchased for about $10 on ebay (or much less from China direct).
Image quality for medium and long range subjects is pretty terrible because of the low sensor resolution and the poor quality wide angle lens - see
attached image (slightly cropped) of my tractor taken from a distance of about 10 meters in contrasty sunlit snow conditions. It can be seen that the
image is only in focus (approximately) in a small area at the centre. Not very promising it would seem!
However, the intent of this project is to examine oud labels or other documents in detail (close up or macro) - not at long distance. So lets see next
how the camera can perform in macro mode.
[file]25868[/file] [file]25870[/file]
jdowning - 2-16-2013 at 04:09 PM
The minimum focal length of the fixed focus lens of the camera is about 0.5 metre - far too long for macro images. The solution for this camera is
simply to place the camera over a 10X magnification stamp loup used here as a 'close up' lens. The lens of the loupe itself is not very good optically
showing some image blurring around the edges. However, as the camera lens only uses a small area of the centre of the loup lens, the overall macro
image is of surprisingly good resolution and quality.
The uncropped image of the stamp is shown here in full spectrum colour. The actual size of the small print 'CHRISTMAS NOËL' on the stamp is about 1
mm in height.
Next to convert the camera to Infra Red operation.
[file]25872[/file] [file]25874[/file]
jdowning - 2-17-2013 at 11:43 AM
Taking the camera apart to access the lens/sensor module is fairly straightforward for a simple camera like this. The back is held in place by a
single screw hidden under the plastic cover plate over the lens/mode LCD - the plate (glued in place) is removed by prying with a thin blade.
With the back removed, the circuit board, held by two screws, can be released.
The lens/sensor module is soldered to the circuit board so cannot be removed. The lens housing is screwed into the module and locked in place with
some kind of glue. The housing can be unscrewed with a little bit of force (to break the glue bond) using needle nose pliers and, at the same time,
holding on to the lower half of the module to avoid damaging the circuit board connections.
With the lens housing unscrewed the red coloured IR 'cut' or blocking filter is found under the lens glued in place with a soft adhesive. The filter
is easily removed by prying with a small screwdriver and is discarded. Any loose debris is then blown out of the module with compressed air and the
lens housing screwed back in place.
The camera is now capable of night vision and the focussing distance can be adjusted by screwing the lens in or out as required - even dispensing with
the need for a close up lens.
Some fine adjustment of the lens focus is necessary due to the longer wavelength of infra red light (lens screwed out a little). This is best done by
connecting the camera to a PC and viewing the image through the camera on the computer monitor as focus adjustments are made by rotating the lens
housing.
jdowning - 2-17-2013 at 12:12 PM
Here is an image in daylight conditions of the postage stamp with the camera IR cut filter removed and focus adjusted for use with the stamp loupe
close up lens. The focus is not quite as sharp as it might be and requires some further adjustment but it can be seen from the 'faded' colours that
the camera sensor is now picking up a mixture of full spectrum wavelength light and infra red - the sensor now being fully sensitive to IR with the
cut filter removed.
To test for night vision capability a source of 'invisible' IR illumination will be required. The most convenient, low heat, IR light source might be
an infra red LED flashlight - low cost - available on ebay - but not locally available so will have to make one with materials I have to hand. More on
that later.
Next to test the camera in full IR mode using IR pass filters.
jdowning - 2-18-2013 at 11:59 AM
Commercially available IR pass photographic filters are expensive - costing around $100 to $300 each - so have no place in this project.
Two 'found' plastic film materials that can be used to make IR pass filters - i.e. a filter that will pass infra red but block most visible light
are:
1) Computer floppy discs (3.5 or 5.25 inch). Just beak open the outside casing to reveal the plastic disc inside.
2) Exposed colour film that has been processed - the dark red blank material typically found at the end of a roll of processed film.
Preliminary testing was done using a piece of blank exposed 35 mm colour film temporarily taped in place over the camera lens.
The density (darkness) of the filter may be increased by using several layers of film - presumably the darker the filter the greater the IR wavelength
(and reduction in the intensity of light reaching the camera sensor = longer exposure time).
The attached image of the stamp is taken in daylight conditions with the camera lens cut filter removed and with the IR pass filter over the camera
lens - so is pretty much a full infra red image with little if any visible light reaching the camera sensor.
This is the effect sought after by photographers using near infra red photography - mostly for making dramatic landscape images it would seem.
Under daylight conditions there would appear to be plenty of illumination for near IR macro imaging despite the small aperture of the camera lens
(almost a 'pin hole').
Next to remove the IR pass filter and test the imaging capability in total darkness using IR illumination.
[file]25914[/file] [file]25916[/file] [file]25918[/file] [file]25920[/file]
jdowning - 2-19-2013 at 11:53 AM
The final camera setup trial is to test the 'night vision' capability i.e. the ability to record an image in total darkness.
The IR light source required to illuminate the subject was made from an old incandescent flashlight (difficult to find them these days and the modern
LED flashlights are no good for this purpose as they emit little IR). The lens of the flashlight was covered with a single piece of floppy disc film
as an IR pass filter. It can be seen that there is still some light from the visible end of the spectrum passing through the filter. Another layer of
film would likely block all visible light but - at the same time - reduce light intensity.
When pointing the flashlight directly at the camera (with the IR cut filter removed) the camera sensor does not 'see' the IR pass filter on the
flashlight only the intense IR wavelength light passing through. This is a good way to judge the effectiveness of material to be used for IR pass
filters.
With the camera set up for night vision (no IR cut and no IR pass filter) and in total darkness, the image of the stamp can clearly be seen under IR
illumination.
All a bit rough and ready but it does demonstrate that 'night vision' is also possible with this basic setup.
Now to test if 'seen in a different light' hidden or obscure details of an oud label can be revealed using this simple apparatus.
jdowning - 2-20-2013 at 12:13 PM
It is interesting to compare images taken by the minicamera unmodified (daylight, full visible spectrum or F.S), modified to infra red, daylight
illumination (i.e. with the camera IR cut removed and replaced with an IR pass filter) and I.R. in daylight with additional incandescent side
illumination. This demonstrates a big improvement in the resolution of the camera when the sensor only receives reflected infra red light. It also
confirms the need for strong, uniform lighting for macro photography.
The '2011' numbers are actually about 0.3 mm high on the stamp.
By way of comparison an image taken with a Canon A470, 7.1 Megapixel camera in full macro mode, daylight conditions is attached. Clearly - and not
surprisingly - the higher resolution and better lens system of the Canon camera makes a big difference. Note that the image has been reduced in size
from its original dimension of 3072 X 2304 pixels to match the dimensions of the Minicamera images (665 X 482 pixels)- so there is some loss of
resolution as a result.
[file]25954[/file] [file]25956[/file] [file]25958[/file] [file]25960[/file]
jdowning - 2-23-2013 at 11:24 AM
Although the results obtained from the little minicamera - even with only a 0.3 Megapixel sensor - are encouraging, it is difficult to use with
precision without an LCD screen to frame and focus the image.
The local hardware store had a low cost fixed focus camera with LCD screen and 5.1 Megapixel sensor on sale for under $20 so I have decided to modify
this camera for further IR trials.
The camera is a Vivitar Vivicam 5022 - made to look like a more expensive camera costing in the $100 to $200 range - it is really only suitable for
taking images in daylight conditions but should be ideal for a straightforward IR conversion.
It is powered by 3 AAA batteries and can be USB connected to a PC for downloading images (but cannot operate as a webcam)
The conversion will entail opening the lens module to remove the IR cut filter and modifying the lenas housing so that the camera focus can be
manually adjusted for close up work.
[file]25998[/file] [file]26000[/file]
jdowning - 2-23-2013 at 11:58 AM
The camera front panel is held in place by five screws located on the side. With the front panel removed the lens module is readily accessible held in
place by two springs. When converting cameras with electronic flash be careful not to touch the high voltage capacitor to avoid getting an electric
shock.
With the retaining spring released the lens housing can be lifted out to reveal the CMOS sensor. To avoid dirt or dust getting on to the sensor it is
best - at this stage - to place the camera inside a sealed plasic bag.
The lens housing is screwed into a plastic mounting fixed with a couple of dabs of glue at the factory focus setting. This glue can be carefully cut
away with a scalpel to allow the lens housing to rotate (use needle nose pliers initially to gently break the residual glue seal)
The IR cut filter on this camera is placed in front of the lens so the lens must first be removed from the housing. It is held in place with a press
fit metal retaining ring - locked in position with three spots of glue. The glue is cut away with a scalpel to free the ring that can then be pryed
out (being very careful not to scratch the plastic lens). The lens can then be popped out to reveal the IR cut filter that is not glued in place so is
easily removed. The lens is then reassembled in the housing and the retaining ring pressed back in place.
To allow access to the lens for manual focussing the plastic cover on the front of the camera was removed.
[file]26004[/file] [file]26006[/file] [file]26008[/file] [file]26010[/file]
jdowning - 2-23-2013 at 12:33 PM
With the IR cut filter removed the camera sensor 'sees' both visible and IR light.
The range of close up focus by manual adjustment of the camera lens (screw out for closer focus and in for longer distance) is demonstrated by the two
attached, uncropped (but reduced in size) images of the postage stamp.
Both of these images are hand held as the camera has an 'anti-shake' function - which should be very handy for macro work where the camera must be
held perfectly still, preferably mounted on a tripod.
Note the improved resolution at extreme close up due to the 5.1 Megapixel sensor.
The camera also has a 4X digital zoom function that may also be useful for macro work (but pretty useless otherwise) - but this feature has yet to be
fully tested.
Next to test the full IR capablilities of the converted Vivicam 5022.
[file]26018[/file] [file]26020[/file]
jdowning - 2-24-2013 at 05:36 AM
The ability of infra red cameras to 'see through' two or more layers of material (paper or fabric) is well known (notoriously so in some
quarters!).
This ability may have an application in interpretation of old oud labels that are difficult to read due to fading of the print or manuscript
signatures or where the original text is obscured by paper, cloth or glue overlay.
The objective here is to test if the 'see through' effect can be replicated with a basic low cost converted digital camera operating in macro, close
up mode under reflected light conditions. If proven successful this may be a useful low cost tool for the oud historian.
The following trial images of fabric and paper overlaid on a test panel, printed on heavy card, measuring 67 X 43 mm have all been taken, hand held,
in identical full reflected daylight conditions (no supplementary or back illumination) using a Canon A470 digital camera for full visible light
spectrum photos and the converted Vivitar 5022 camera for full infra red.
For full IR a small disc, 5 mm diameter, of exposed colour film has been placed over the front of the Vivitar camera lens (held in place with a couple
of spots of glue). This is the near IR pass filter that blocks most of the visible light from reaching the camera sensor.
The first test compares the Canon full spectrum image with the Vivitar IR image with a heavy paper overlay in place over the test panel. The 'see
through' effect of the infra red image can be clearly seen.
Note that the paper is not pressed in close contact with test panel so the print of the test panel is somewhat less distinct that it might otherwise
be.
The same applies to a patterned cotton cloth overlay - just loosely placed over the test panel. Again the see through effect of the IR image is here
clearly demonstrated.
To follow, two more sets of trial images one using an opaque black silk overlay and the other several overlays of hand made paper.
[file]26026[/file] [file]26028[/file] [file]26030[/file] [file]26032[/file] [file]26034[/file]
jdowning - 2-24-2013 at 06:10 AM
The black silk fabric overlay is closely woven and opaque under reflected full spectrum light whereas the print on the test panel is visible in
reflected near IR.
Repeating the test with thin hand made paper overlays - one paper layer is part translucent so the test panel can be seen by both comeras. However,
when a second layer of paper is added the test panel is no longer visible under reflected full spectrum lighting but is still visible in IR. The panel
is still faintly visible under three layers of paper in IR.
Note that all test images posted have been reduced in size from original but have otherwise not been edited.
[file]26036[/file] [file]26038[/file] [file]26040[/file] [file]26042[/file] [file]26044[/file] [file]26046[/file] [file]26048[/file]
jdowning - 2-26-2013 at 11:43 AM
As a final test - to see if glue makes any difference to the 'see through' effect - test strips of paper, card and fabric were glued over the test
panel and photographed in full spectrum light and infra red.
The sample strips are as follows:
A) hand made paper - one layer.
B) hand made paper - two layers.
C) heavy card 0.3 mm thick.
D) patterned cotton cloth.
E) black silk.
F) light card 0.16 mm thick.
Lighting was natural daylight on an overcast afternoon. Bright sunlit conditions might have provided even clearer results for the infra red image.
It can be seen that - except for the single layer of hand made paper all of the other strips are opaque in reflected full spectrum light.
For the infra red image, the see through effect is only slight for the light card and almost non existant for the heavy card (although still faintly
visible in the original image).
So, these basic tests confirm that there may be some potential for reading hidden texts on oud labels using infra red imaging although much will
depend on the composition, colour and thickness of the overlay materials as well as the ink composition of the underlying text. Changing the
wavelength of the IR pass filter on the camera lens as Dan previously suggested might be necessary for optimum results - but this has not been tested
here.
Different wavelength filters might be made up from several layers of exposed colour film - darker = longer wavelength?
All digital cameras - including camcorders, webcams, cell phones etc can be converted to IR by replacing the IR cut filter with an IR pass filter.
However, the simplest of cameras may be the best way to go to provide reasonable macro results at low cost. A camera suitable for straightforward
modification as previously described should have:
1) No moving parts - i.e. fixed focus, no optical zoom facility.
2) An LCD screen for convenient framing and focussing of the subject matter.
3) USB connectivity for loading images to a PC.
The greater the resolution of the camera sensor the better as well as other 'nice to have' stuff like an anti-shake facility.
The low end Vivitar range are a good example of cameras that are relatively easy to modify to give acceptable results - but other makes may also be
suitable.
Note that modifying the camera lens from fixed to manual focus allows the camera to be used for both extreme close up as well as longer distance work
- even if the camera has not been converted to IR.
[file]26082[/file] [file]26084[/file] [file]26086[/file]
jdowning - 3-4-2013 at 03:57 PM
One last observation for general information.
The Vivitar camera comes with a small built in electronic flash unit that can be useful for IR illumination at close range - the flash being rich in
IR wavelength light.
It works quite well in extreme close up where the camera lens is only a few mm from the subject so illumination by other means is difficult. The flash
is intense enough to be reflected sideways over the subject area for illumination. As the subject cannot be seen through the camera LCD screen for
precise focussing it is usually necessary to take a few shots and pick the one that is in best focus - not a problem with digital photography.
Here are two hand held extreme macro images of the postage stamp using the camera flash unit - one with the IR pass filter in place (full IR) the
other with the IR filter removed (a mix of full spectrum and IR).
The little dots resolved in the image are the individual screen dots used to print the stamp design in colour.
[file]26161[/file] [file]26163[/file]
juju - 3-19-2013 at 04:57 AM
Hi jdowning,
I'm working with thermal imager for 5 years, I mainly use this technique as an undestructive way to study heat transfer in buildings. We also use
infra red to detect and prevent damages from humidity too. Last year I participated to a national congress about infra-red investigation. Someone
explain the application of infra-red to detect defaults in historical paintings. Our national museum (the "Louvre") use this technique in France for
the restauration of paintings. I can find the article for you if you want to know more about this technique.
I made for you 2 pictures of my oud in infra-red. The first thing you have to know about infra-red is that the reflexivity of material is changing in
that frequency. I took a picture of the bowl of my oud, the varnish is creating a perfect mirror, here it is impossible to investigate the wood. A
good way to have a better look is to put off the rosette and the strings (but I didn't do it !) and to take a picture inside the oud.
I took another picture of the face of the oud, which show the structure of the oud. Unfortunately I don't have any knowledge about lutherie and oud
construction, I won't be able to help you about the real interest of infra-red observation for the oud.
best,
Julien
[file]26403[/file] [file]26405[/file] [file]26407[/file] [file]26409[/file]
jdowning - 3-19-2013 at 05:16 AM
My eldest son has an interest (among many others technical or otherwise) in microscopy and has carried out some trials using near IR transmitted light
to illuminate prepared mounted specimens. His IR pass filter, mounted over the microscope lamp, was made from 3 layers of exposed and developed colour
film (as used in my reflected light experiments with the modified digital cameras). His microscope digital camera is sensitive to IR light so likely
does not have an IR cut filter in the optical path as most digital cameras do.
The results are interesting enough to be posted here for anyone interested in using IR for analysing documents and labels etc. This might have a
useful application if a manuscript on paper or vellum can safely be illuminated from the back with IR and photographed from the front with an IR
sensitive digital camera. Heat is a potential problem with illumination rich in IR so heat filters or electronic flash may be required to avoid any
damage to sensitive subject matter.
For information here are two images of the head of a caterpillar at 25X magnification one in normal visible light illumination the other in near IR
that demonstrates very well the 'see through' capabilities of IR. The chitin (hard shell like material) in the caterpillar head that is opaque under
visible light is invisible under near IR wavelength light.
I shall be running further tests with transmitted IR rather than reflected IR to see what more can be achieved with a digital camera converted to IR
as far as reading damaged or obscured documentation is concerned - time permitting!
A couple of off topic observations concerned with strings that may be og general interest.
The head of the caterpillar is from a 19th C prepared microscope slide . The species of caterpillar is unknown but it shows the two ducts leading to
the spinarettes that extrude the double silk filament that the caterpillar uses to spin its cocoon. It might even be the Bombyx Mori species whose
silk filament is used for silk string making?
Interestingly, chitin derived from shellfish waste can be processed into a fibre that has many practical uses - including dissolvable (in the body)
medical sutures that can replace the silk and gut sutures used in times past. I wonder if the stuff might also make good musical instrument strings
(non historical of course)?
[file]26411[/file] [file]26413[/file]
jdowning - 3-19-2013 at 05:43 AM
Thank you juju. Yes I would be interested in the report on the techniques used for investigation of paintings under IR wavelength light. Can you post
it here as a PDF file or is the file too large?
I guess that for your work in detecting heat loss in buildings you are making images that are deeper into the IR spectrum than I am using for near
Infra Red imaging (i.e. my wavelengths are restricted to around 760 to 900 nm or so). This means that your camera is measuring the surface
temperatures of your oud and surroundings and so does not otherwise effectively 'see through' the wood? Mind you I doubt if near IR wavelength
illumination would be capable of seeing through the sound board of an oud either unless perhaps a very powerful source of illumination is employed (at
risk of setting fire to the instrument!) - but I do not expect it to.
So far this technique (low cost near IR) looks as though it may have a useful application for forensic examination of documents using either reflected
or transmitted light, For any document (including an oud label) the subject matter would have to be viewed directly, close at hand and not at a
distance through a layer of wood or other material.
I have yet to test transmitted IR illumination but will do so to see if it might also be possible to illuminate a label through an oud bowl (i.e.
through the wooden ribs) - the label, of course, being glued in intimate contact with the wood. I am not hopeful that it will work but "nothing
ventured nothing gained".
jdowning - 3-19-2013 at 12:07 PM
Curious to see if IR would 'see through' wood, I ran a quick test this morning on a lute and guitar sound board using reflected and transmitted light.
Wood thickness for the lute soundboard is about 1.8 mm for the lute and about 3 mm for the guitar
Using reflected light, daylight and electronic flash using the converted Vivitar camera with IR pass filter. Result - no success.
With transmitted light - the light source behind the subject - it is a different story.
Test 1 - an unfinished lute soundboard was placed against a window (daylight illumination - rich in IR). Photographed in visible light the bracing is
barely visible. Using the converted Vivitar camera (IR cut filter removed) without the IR pass filter (i.e. both visible and IR seen by the CMOS
sensor of the camera the sound board appears to be almost translucent with bracing clearly visible. Adding an IR pass filter did not improve matters -
just a darker image due to less light intensity.
Test 2 - a low intensity incandescent light source was placed inside the body of a classical guitar for illumination. Photographed in visible light
nothing is visible but the fan bracing is clearly visible photographed in visible and IR light spectrum.
No attempt has been made to improve on focussing the images as this is just a quick test to verify the possiblities.
Next to check the local Radio Shack store for miniature 12v incandescent bulbs - perhaps small enough to pass through a lute or oud rosette in order
to provide IR rich illumination to enable digital IR imaging of the bracing? High intensity (and high heat) illumination is not required for this
application - even the night light illumination is excessive.
Perhaps this might be another useful low cost application for the luthier researcher?
[file]26417[/file] [file]26419[/file] [file]26421[/file] [file]26425[/file] [file]26427[/file]
juju - 3-20-2013 at 02:32 AM
You can find the article here http://we.tl/759y4uXIys
What interests you starts at page 97, unfortunately it's written in french.
Those investigations are made in infra red (8 to 12 µm).
Best,
Julien.
jdowning - 3-20-2013 at 02:47 AM
Thank you juju - a good opportunity to 'brush up' on my French.
jdowning - 3-21-2013 at 10:21 AM
The smallest miniature lamps that I could find in a local store measure 3 mm (1/8 inch) in diameter. These will easily pass through the rosettes of
most lutes and ouds.
Light output is quite low (mini krypton flashlight bulb, 0.6 candle power at 0.3Amp, 2.5 volt) but is still sufficient to be detected by the modified
camera in IR + visible wavelength light - even through the thicker guitar sound board. Nothing, of course, is visible through the sound boards without
the camera imaging.
Next to wire up the lamp for interior illumination and run some tests to inspect and photograph the bracing layout of several lutes and other
instruments with rosette covered sound holes.
[file]26450[/file] [file]26452[/file] [file]26454[/file]
jdowning - 3-23-2013 at 12:07 PM
The tiny 'krypton' filled lamp (full of infra red wavelength light) has been soldered to a longer lead and connected to a battery supply (2 AA Ni-Cad
rechargeable batteries in series, total voltage 2.4 volts. Standard Alkaline batteries at 3+ volts would quickly 'burn out' the lamp). Low voltage,
low heat means a safe application.
The lamp was tested on three of my instruments - lute, baroque guitar and vihuela - each with a sound hole covered with a rosette. The lamp was fed
through each rosette to illuminate the instrument interior and the soundboards photographed with the modified Vivitar camera (visible + IR). Although
the illuminated area and light output is small (but the lamp can be easily moved around to compensate if necessary) the bracing geometry can clearly
be seen (viewed by the camera but not by the naked eye). Note that the illumination is more uniform in the lute where the bowl acts as a curved
reflector to focus the light compared to the flat backed guitar and vihuela.
Clear focussing of the image requires improvement (with perhaps a different camera with longer focus than macro to cover a full sound board) but,
nevertheless, I suspect that it will not be possible to bring the hidden bracing into sharp focus through the sound board wood (a similar problem is
found with Xray imaging). Nevertheless this non destructive imaging technique may provide useful information for the luthier at little cost for the
tool.
For those instruments played with the support of a shoulder strap access for the light source may alternatively be provided by removal of the strap
button.
[file]26472[/file] [file]26474[/file] [file]26476[/file] [file]26478[/file] [file]26480[/file]
jdowning - 4-6-2013 at 03:53 PM
Moving a bit off topic - but related - posted here only for information for those interested in exploring the scope of infra red imaging.
The modified Vivitar camera - as previously reported - works quite well at close up and macro imaging despite the basic simplicity of its lens.
The modified camera did not work well for long distance (landscape) images that were out of focus. The lens module has, therefore, been further
modified to allow in focus landscape IR imaging.
The IR cut filter on this camera was located in front of the lens and so acted as a spacer about 1 mm thick. Removal of the IR cut filter therefore
resulted in the lens being moved that distance closer to the CMOS sensor. As it happens, there was then insufficient adjustment in the lens module to
manually correct the long distance focus of the lens.
The lens module was, therefore, then dis-assembled again and a leather washer - about 1 mm thick - placed in front of the lens to compensate for the
missing IR cut filter.
This modification worked reasonably well although could not overcome the relatively poor quality of the simple lens to provide high resolution images
in long distance focus.
I was recently sent (for free!) a high quality precision optical grade IR filter that passes infra red wavelength light above 750 nm. This happened to
conveniently fit the front plastic lens cover of the Vivitar camera to give the camera a (quite inappropriate!) high tech IR appearance.
The attached landscape images taken with the IR pass filter attached - edited to black and white to remove the original red colour of the IR images -
gives a sense of what to expect with IR landscape digital photography - dramatic dark skies and evergreen vegetation that is white (note the shrubs
around the door).
The mounds of white stuff on the ground is snow - bad winter this year.
[file]26589[/file] [file]26591[/file] [file]26593[/file] [file]26595[/file] [file]26597[/file] [file]26599[/file] [file]26601[/file]
jdowning - 5-17-2013 at 11:38 AM
The hidden detail on the test card previously reported on this thread was revealed quite well with near infra red imaging (using the modified Vivitar
digital camera) under reflected daylight illumination.
The test has been repeated this time under transmitted light conditions - that is with the card illuminated from the back and the card photographed
from the front with near IR imaging.
The test card is 0.4 mm thick printed on one side with the print covered by strips of paper and fabric glued in place to obscure the printed detail
under visible wavelength light. For this test the card was placed in front of a desk lamp fitted with a standard 60 watt incandescent light bulb -
incandescent light being rich in IR wavelength light.
The test card is so thick that no visible spectrum light passes through.
The view through the modified camera is a different story, however as is evident in image Fig. 1. The near IR wavelength light passes easily through
the thick card allowing the hidden printed detail on the card to be clearly revealed - a considerable improvement over the images recorded under
reflected IR light conditions.
Taking the test further - to get some idea of the material thickness that near IR wavelength light will penetrate under transmitted illumination
conditions - the test card was clipped to a piece of quarter sawn Sitka Spruce sound board material 2.5 mm thick. The attached image Fig.2 shows that
even with the wood thickness plus the card thickness, the results are still slightly better than the reflected light images previously posted in
revealing the hidden detail on the card (the focus of this image is slightly off so could be improved).
The images have been converted to 'grayscale' to remove the red tint inherent in the IR images.
No doubt, increasing the intensity of the incandescent illumination (i.e. using a higher wattage light bulb) would result in an even greater depth of
penetration of the IR wavelength component? However, higher intensity incandescent light comes with greater heat output that might cause problems when
illuminating heat sensitive materials - so some caution is necessary.
Alternatively low voltage, high intensity, incandescent flashlight lamps (e.g 'Krypton' bulbs) or infra red LED's would be a safer, cooler
alternative.
[file]26796[/file] [file]26798[/file]