||11 months ago|
|pictures||11 months ago|
|.gitignore||11 months ago|
|license.md||11 months ago|
|readme.md||11 months ago|
Extending a compound microscope: DIY multichannel microphotography
Microscope photography is all about constructing an optical circuit to record the state of pixels. These signals, selected for against various optical filters, capture different biological processes.
This document describes my extended microscopy setup. It's capable of 6 light modes including UV and infrared. The base microscope is a Nikon Labophot 2: featureless quality.
|Overview of the 6 light modes: basic configuration and example use case.|
Each mode depends on the configuration of circuit elements:
Normal. Settings persist unless otherwise noted.
- 20W stock halogen illuminator
- 650nm infrared lowpass filter (for camera mod)
- Linear polarizer just before the camera mirrors
Polarized. To visualize torsion and crystals.
- 55mm rotating linear polarizer above illuminator
Darkfield. To visualize transparency.
- Opaque black circle above condenser
- Oblique: crescent not circle
Phase Contrast. Coming Soon!
Fluorescent. To visualize most green dyes.
- 460nm blue LED illuminator
- 530/30nm green bandpass filter
IR Spectro. To visualize chlorophyll and blood.
- 660/850nm red/infrared LED illuminator
- Replace 650nm lowpass with 750nm infrared highpass filter
Phosphorescent. To visualize marine samples.
- 365/395nm UV-A blacklight LED illuminator
- Wratten No. 12 ("minus blue") 500nm yellow highpass filter
Anyway, I'll just go through the optics path one component at a time. Starting from the top down.
Top down index
todo: Describe controlling the camera with Lightroom. Basic raw editing/exporting in Photoshop. Stripping EXIF data and lossless compression.
Any prosumer dSLR released in the last 10 years can match or exceed a fancy lab camera. Even my Nikon D60 has an accurate 10.2 megapixel sensor with good dynamic range.
I regret the lack of Live View and computer tethering abilities. So I may upgrade to a Canon EOS 40D+ with both features.
Modifying the camera
Astrophotography hobbyists often remove the internal CMOS filter to capture infrared nebula clouds. If you think about it, astrophotography is very similar to microphotography in important ways.
An internal filter artificially limits the CCD sensor's bandwidth to the visible spectrum. Without it, photos have a conspicuous red tint instead of being true color.
I correct the problem with an external infrared filter. Otherwise, Image => Auto Color in Photoshop also fixes it.
BLAH BLAH BLAH
[b]Before going further, please understand that you're voiding any warranty and making your camera dependent on external infrared filtering and/or post-processing to take "normal" pictures without a conspicuous red tint.[/b]
The exact instructions vary by camera (search Youtube for your camera "astrobiology mod"). The point is that you must manipulate the sensor and the precision optical electronics around it.
I removed the case screws and the topmost display PCB.
The space between the camera and the adapter is a 40mm diameter circle about 8mm deep. It's the perfect place to stash discs of precision, often mismatched dichroic glass.
I bought quality used filters with known curves whenever possible.
todo: Describe making fittings for the mismatched optics inside the camera adapter. Also, sourcing appropriate optics.
I'm using an AmScope dSLR adapter on a standard eyepiece. The camera adapter approximates the magnification of a normal eyepiece. Depending on the sensor size, this is 12-15x for my adapter.
todo: calculate FOV/magification from the camera specs
I taped a circle of polarizing film inside the adapter. This protects the lens from foreign objects and it rectifies the light hitting the sensor. The semi-permanent polarizer also reduces reflections and increases apparent contrast.
Each eyepiece tube allows a different view:  650 nm infrared cutoff filter  normal unfiltered view
The infrared filter is required for true-color photos on a modified camera. By replacing the internal filter with regular glass, the CCD picks up infrared and near UV.
This tints all "normal" photos red so I found two 20 mm < 650 nm band pass filter disks. They fit inside the microscope eyepiece tube and behind the normal camera lens.
Effectively, the filtered view becomes the "normal" one and the unfiltered view becomes the fluorescence one. The two should look almost identical to the naked eye.
When shopping for a microscope you should look for a complete set of apochromatic PLAN objectives. These are the best optical standards and the objective is the cornerstone of magnification. Unless you already have good objectives or want to buy a set, avoid scopes with bad objectives.
My ideal turret would have a phase contrast condenser and these objectives:  10x normal  40x normal  40x phase contrast  100x oil [*] 100x oil/phase contrast
This is where the specimen is. The mounting medium is important.
I prefer glycerol whenever possible because its refractive index most closesy matches the 1.515 glass standard. The closer the separate components' refractive indices, the clearer the picture will appear.
Pure glycerol is slippery and bubbly, difficult to handle. But it lends an oil-like effect to all objectives, clear and bright. Under oil immersion, is matches the oil itself and enhances the image.
The condenser focuses the light source onto the stage. Darkfield, oblique, and phase contrast microscopy happens here.
It's possible to do one of three things:  buy a used darkfield/phase contrast condenser ($250)  buy plastic inserts for the stock condenser ($75) [*] make your own stock condenser filters
Proper phase contrast microscopy uses specific diameter disks for each objective. The difficulty cost increases unsustainably because you also need to buy matching objectives.
todo: make my own to rest on top condenser lens. oblique and darkfield, maybe rhiemann when using them, open the condenser aperture all the way
todo: Briefly discuss Osram Bellaphot and the clean warmth of halogen. Link to diybio/light-source.
This guide will teach you how to extend the professional imaging qualities of a lab surplus compound microscope such as a Nikon Labophot.
To start, here's the basic optics of a normal compound microscope.
A good imaging setup is more complicated but most of the mods are simple and obvious. Adding a dSLR with adapter is one example, and polarizing film is another.
Anyway, your magication factor is the product of the components in the optical chain. For example, my max effective magnification is 1200-1500x because of these factors:  100x objective with immersion oil  12-15x camera lens (depends on CCD)
So for every nominal objective magnification, I can expect 12-15x more magnification. I'm going from light to computer and will discuss details in the appropriate sections.
== Basic considerations ==
Solid equipment will deliver the best results. It's possible to use a cheap USB microscope, but ideally the setup would last for life if upgraded, i.e., swapping out the dSLR with advances in the secondhand market.
The optical integrity of your microscope is key. It's worth paying extra for a professionally refurbished one in known working condition, lest you find yourself surrounded by half-broken scopes.
A rugged, well-built classroom scope such as a Nikon Labophot can be had for about $350. Look for listings that specify a complete set of PLAN objectives or prepare to replace the missing ones. Apochromatic PLAN are considered the best re: color retention and vignetting.
== Lamp == Most scopes use a halogen bulb. This provides clear, warm light.
I was able to find new old stock Osran Bellaphot bulbs cheap. These will be good replacements when the current bulb burns out.
A removeable high-power LED array is a necessary upgrade. LEDs with multiple colors (red, green, blue, and white) helps with fluorescence microscopy. They replace the exitation filter and mercury lamp assembly, which is expensive and hazardous.
Also, cold white LEDs can help with infrared microscopy. This is important with modified cameras because halogen is mostly infrared. Therefore halogen will contaminate the field and ruin IR images.
Here are some examples of good premade LED arrays, albeit expensive. I can't stress enough the importance quality LEDs and known spectral curves.
LED Engin LuxiGen LZP-00M RGBW LED https://www.mouser.com/new/ledengin/led-engin-luxigen-lzp-00m-led/
LED Engin LZ4 LuxiGen™ UV LED Emitters https://www.mouser.com/new/ledengin/led-engin-lz4-led-emitters/
LED Engin LZ1 IR LEDs https://www.mouser.com/new/ledengin/led-engin-lz1-ir-leds/
== Linear polarizer ==
The Labophot 2 light source is about 55 mm so it can take camera lens filters. To make polarized light images, I place a 55 mm linear polarizer on top of the light source. I can freely rotate the bottom polarizer relative to the other polarizer inside the camera adapter.
Note that both polarizers must be linear. Unlike radial or elliptical polarization, linear lets us control the degree.
Every element occludes the core optics and the trick is to use quality filters. Unfortunately, special microscope filters
== Equipment ==
=== Microscope ===
For normal use, you should find clear and bright halogen bulbs. I happened to find 4 new old stock Osram Bellaphot bulbs for cheap. They have a poor lifetime and eventually need to be replaced.
== Upgrading the scope ==
=== Polarized light ===
I cut a 40 mm circle in a polarizing sheet and placed it inside the camera adapter at the lend interface. This is a good place
== Wratten No. 12 ==
Minus blue filter; complements #32 minus-green and #44A minus-red. Used with Ektachrome or Aerochrome Infrared films to obtain false-color results. Used in ophthalmology and optometry in conjunction with a slit-lamp and a cobalt blue light to improve contrast when assessing the health of the cornea and the fit of contact lenses. Longpass filter blocking visible wavelengths below 500 nm
== Infrared ==
Wratten No. 88
Passes infrared, blocks visible wavelengths below 700 nm.