Spectroscope

[Nate]

@photoniI found this useful site you might be interested in. Just click on the part you want to zoom in on goes from around 360nm-950nm

https://fermi.jhuapl.edu/liege/s00_0000.html

[ulf]

Peaks in flash tube light are generated by the gas of rather low pressure in the flash tube.

Unfortunately (for wavelength identification) those peaks are few and mostly in the NIR area. 

Fortunately for the image quality the spectrum is rather continuous and spike free.

[dabateman]

3 hours ago, photoni said:

what do you mean ?

 

I was hoping to have some known "references" for example those peaks in IR what are they?

Get a Mercury bulb, like a compact fluorescent bulb and shine it on your bar setup. You will see the Mercury lines, the strong 365nm seems to even show up in some compact fluorescent lights.

Here they are listed for you in Angstroms, just move the decimal point over to the left one digit for nanometers.  So 4046.563 Angstroms is 404.6563nm.

https://physics.nist.gov/PhysRefData/Handbook/Tables/mercurytable2.htm

 

You should see 365, 398, 404, 435, 546, 567, and 614nm.

[photoni]

On 7/13/2022 at 11:18 AM, dabateman said:

Get a Mercury bulb, like a compact fluorescent bulb and shine it on your bar setup. You will see the Mercury lines, the strong 365nm seems to even show up in some compact fluorescent lights.

Here they are listed for you in Angstroms, just move the decimal point over to the left one digit for nanometers.  So 4046.563 Angstroms is 404.6563nm.

https://physics.nist.gov/PhysRefData/Handbook/Tables/mercurytable2.htm

 

You should see 365, 398, 404, 435, 546, 567, and 614nm.

on the site I found this visual  reference

 

does a normal neon fluorescent light 20w cold light 6000 ° K work? .

do you know the type of emission of the tanning lamps (neon fluorescent tubes - low pressure)?

 

[Andy Perrin]

Photoni, stars like the sun have an absorption spectrum that depends on their composition. Different chemical elements contribute different absorption lines. In the case of the sun, the lines are named Fraunhofer lines. 
 

Lamps that have a low pressure electrified gas inside like mercury lamps have an emission spectrum instead. Rather than black lines on a colorful background, they glow only at a few bright lines. The location of the lines depends on the chemical composition of the gas. It will not match the sun’s lines unless the gases are the same as the sun.
 

We do not know what is in your flashlight so it is impossible to know what kind of lines to expect. 
 

Neon lines can probably be googled

[photoni]

@Andy Perrin  thank you for your patience, I understood these things :)

 

I'm looking for an empirical method to "see" ... not to measure "nm"

I changed the chrome reflex bar, it used to measure 12mm in diameter now 3mm

I used the A7 f.s. with Soligor KA at f: 11

- 1000W studio flash light

- Modeling light 3200 ° K

this confirms that the flash light is not perfectly constant in the VIS, and is very jagged in the IR

Since neither of them have "recognizable" lines I used Jason's histogram as a reference for the ZWB1
it seems to me that the Fraunhofer scale also coincides.

For the first time I saw green after yellow
if I put TSN575 that cuts red and IR ... it disappears :)

 

Thanks

Antonio

 

P.S. it is impossible to have a constant set with sunlight to compare my large format lenses.

I only use flash light for my collodion photos

 

 

 

.

 

[photoni]

I have not found answers to the "lines" of the flash light

I found this neon fluorescent light chart
I used the diffraction grating with normal warm fluorescent light.
but I have no precise reference points with that flash light
but it seems acceptable.

[dabateman]

That looks great. You also can resolve the 543 and 546 doublet. So that that gives you some indication into your resolution. 

 

[photoni]

On 7/24/2022 at 1:30 PM, dabateman said:

That looks great. You also can resolve the 543 and 546 doublet. So that that gives you some indication into your resolution. 

Thanks !

 

a help
do you have any idea why the graph is not all in register?
I tried to improve the set
I rebuilt the frame of the diffraction grating, more rigid and flat
I used a smaller Ø2mm steel pin
I also tried with Pinhole (I thought it was a deformation of the lens)

I would like to have a precise reference of the neon lines; what are those two on the right and the one on the left?

Thank you

Antonio

[colinbm]

There are many instructions on using Diffraction Gratings.
To calibrate your images repeatably you need to include the light source slit or pin so that you can measure reliably the wavelengths.
The spectrum from a diffraction grating is nearly linear for your purposes. You just need the source & a known wavelength, then you can measure the rest.

 

https://astro3d.org.au/education-and-outreach/diy-smartphone-spectroscope/

 

Unfortunately each order, left & right from the source overlaps the next order etc.
That is why you are seeing these extra lines, colours.

[photoni]

7 hours ago, colinbm said:

Unfortunately each order, left & right from the source overlaps the next order etc.
That is why you are seeing these extra lines, colours.

Colin ... I thought it was a defect in the diffraction grating
in the visible they do not overlap :-)

 

the problem remains that if I keep the two central references (546 and 611 nm) good, the lateral ones are wrong

if I keep the lateral ones good (405 and 692 nm) the central ones are wrong.

 

if I include the light source, do I have to consider it zero nm?

 

Some of you recommend low pressure mercury lamps for precise calibration
do you have an advice on the web?
(I don't want to buy dangerous or expensive lights)

 

[Stefano]

Diffraction gratings are not linear, probably that's the reason you can't align the lines.

[photoni]

38 minutes ago, Stefano said:

Diffraction gratings are not linear, probably that's the reason you can't align the lines.

Stefano

if it were a linear perspective it would be easy With photoshop

I have tried in many ways to make them coincide together but it doesn't work

if it's logarithmic I can't

A.

[colinbm]

This looks pretty linear to me......
https://www.teachersource.com/product/hand-held-spectroscope

[colinbm]

This will give you the mercury emission lines from UV 365nm to IR.
https://www.ebay.com.au/itm/292083711654?_trkparms=amclksrc%3DITM%26aid%3D1110006%26algo%3DHOMESPLICE.SIM%26ao%3D1%26asc%3D240678%26meid%3Dc9079c4e57114467991c4e28b64f930e%26pid%3D101195%26rk%3D1%26rkt%3D12%26sd%3D165587008521%26itm%3D292083711654%26pmt%3D1%26noa%3D0%26pg%3D2047675%26algv%3DSimplAMLv9PairwiseWebMskuAspectsV202110NoVariantSeedKnnRecallV1WithNode2Vec%26brand%3DUnbranded&_trksid=p2047675.c101195.m1851&amdata=cksum%3A292083711654c9079c4e57114467991c4e28b64f930e|enc%3AAQAHAAABMFVvFeUTQhL4KP79koL%2B4Ydgokeb2CQrAyC%2BOwbdpQetV2qAC7C3Vu0XLY04JD2l71qihjhRi6NHnF8%2BhgXS2aYrSe2gbn68aOiuHVBnxF63KgNvZhJA7t%2FeEhOtCdlYa07HAg%2ForK6yg3NfGr3s%2BAKP1BiX4a%2FHz%2Bst%2BuSx27oicAIPBZ0vUdIeUSQrffa79ucvR3%2BZpQHhY%2BTVh2kAZBE3XzGBCDNrhRbEh5zvn4jesFLYmxbAC7D6E%2FzLdn6wWc0ZXORVGjpE7EnuQfB2wk%2BLgchBrEyPK5MgKL6VDvekNHGsnmpDQXVWIORCr5ufvPAGRtLg8%2F%2BH6jqKrG78kPtl38LtH9XQhCN%2Fbx0SCp7YBY2cslqKIhHwOtVp9Gvf49XDD7s403uOo4viF5X7g4I%3D|ampid%3APL_CLK|clp%3A2047675

[Andy Perrin]

4 hours ago, photoni said:

Stefano

if it were a linear perspective it would be easy With photoshop

I have tried in many ways to make them coincide together but it doesn't work

if it's logarithmic I can't

A.

It’s neither linear not logarithmic. We worked this out already back here:

https://www.ultravioletphotography.com/content/index.php?/topic/4148-camera-sensitivity-limits-measurement-and-possible-way-to-use-your-camera-as-a-spectrometer/#comment-39429

 

[Stefano]

This is one of my old posts on the topic: https://www.ultravioletphotography.com/content/index.php?/topic/3617-infrared-with-a-solar-panel/page/2/#elControls_31757_menu

 

The distance of the diffracted light from the middle is approximately linear for wavelengths significantly shorter than the pitch of the diffraction grating (for 1000 lines/mm that's about 500 nm or less), or for narrow bands (like 700-710 nm). It even shoots to infinity for wavelengths equal to the pitch.

 

Edit: see Andy's post just behind mine.

 

How do spectrometers take that into account?

[ulf]

4 hours ago, Stefano said:

This is one of my old posts on the topic: https://www.ultravioletphotography.com/content/index.php?/topic/3617-infrared-with-a-solar-panel/page/2/#elControls_31757_menu

 

The distance of the diffracted light from the middle is approximately linear for wavelengths significantly shorter than the pitch of the diffraction grating (for 1000 lines/mm that's about 500 nm or less), or for narrow bands (like 700-710 nm). It even shoots to infinity for wavelengths equal to the pitch.

 

Edit: see Andy's post just behind mine.

 

How do spectrometers take that into account?

The spectrometers (off array type with a linear sensor) are built around a small optical bench with mirror(s) that are not flat:

https://bwtek.com/spectrometer-part-4-the-optical-bench/

That is to get the projection from the input slot at the sensor as optimal as possible for the wavelength calibration.

My guess is that this is a similar problem as designing a camera lens with low distortion.

 

The output from the linear array is then calibrated, using a special calibration lamp with several well defined emission peaks.

I use a lamp with mercury and argon lines, with a function similar to this:

https://www.oceaninsight.com/products/light-sources/calibration-sources/hg-2/?qty=1.

 

Even with the effort by the design of the shape of the mirrors in the optical bench the input is quite nonlinear.

The calibration generates a set of coefficients used to correct the wavelengths scale with a higher order polynomial.

 

[photoni]

morning test

Soligor KA 35mm - 1" - @f11 - 3200iso - A7fs - neon light

white balance on teflon ball (bottom right)

 

I kept lines 543 and 546 good
it seems to me that line 365 matches,
maybe the other gases are different from the graph?

 

 

With Nemo zwb2 ...

[ulf]

I think the peaks from your lamp is correct and the deviations are from geometric nonlinearity from your setup, including projection errors induced by the lens.

Sometimes some peaks are more or less high from such light sources.

The wavelengths of the peaks are super stable, to many decimal places.

[photoni]

1 hour ago, ulf said:

I think the peaks from your lamp is correct and the deviations are from geometric nonlinearity from your setup, including projection errors induced by the lens.

Sometimes some peaks are more or less high from such light sources.

The wavelengths of the peaks are super stable, to many decimal places.

 

above I added a test made with the Nemo flashlight with its black filter zwb2

superimposed on the previous image confirms the wavelength

the pinhole photo also confirms the displacement

it is a pity that there are no possible correction parameters

I will have to settle for the "colors" if I see the green is 340nm

... and this was enough for me to test my objectives. Thank you

[ulf]

One thing to add here is that the peak-wavelength of an UV-LED is not as well defined as the gas emission peaks.

All "365nm" LEDs I have seen had their peaks shifted more or less toward 370nm.

The wavelength is also dependant on drive current and chip-temperature.

53 minutes ago, photoni said:

 

above I added a test made with the Nemo flashlight with its black filter zwb2

superimposed on the previous image confirms the wavelength

the pinhole photo also confirms the displacement

it is a pity that there are no possible correction parameters

I will have to settle for the "colours" if I see the green is 340nm

... and this was enough for me to test my objectives. Thank you

Could it not be possible for you to use two light sources, the Neon light to get calibration data and the flash for the measurement, without moving around the setup too much.

That is how I always do for wavelength critical measurements with my spectrometer.

First the reference lamp for calibration, and then a continuous light source for the measurement

 

The Nemo-peak you have added above is shifted even further towards the longer wavelengths because of the quickly increased sensor sensitivity toward the VIS range. The peak looks correctly shifted just as expected compared to the Neon tube Hg peak at 365.0153nm 

This shifting phenomena is true for all wider peaks of either light sources or filters.

 

An easy to see example of this peak shifting is to look at transmission by a UG11. 

Alone it has an UV-peak around 336nm, but in combination with a BG39 (2+2mm), the left slope of this filter shifts the resulting UV-peak to 364nm.

[photoni]

On 8/1/2022 at 1:57 AM, Andy Perrin said:

It’s neither linear not logarithmic. We worked this out already back here:

https://www.ultravioletphotography.com/content/index.php?/topic/4148-camera-sensitivity-limits-measurement-and-possible-way-to-use-your-camera-as-a-spectrometer/#comment-39429

 

to much difficult

 

 

On 8/1/2022 at 1:57 AM, Stefano said:

This is one of my old posts on the topic: https://www.ultravioletphotography.com/content/index.php?/topic/3617-infrared-with-a-solar-panel/page/2/#elControls_31757_menu

.

 

to much difficult

 

On 8/1/2022 at 1:25 AM, colinbm said:

This will give you the mercury emission lines from UV 365nm to IR.
https://www.ebay.com.au

 

I bought a neon Wood 26W - 220v - E27

 

 

On 8/1/2022 at 7:16 AM, ulf said:

The spectrometers (off array type with a linear sensor) are built around a small optical bench with mirror(s) that are not flat:

https://bwtek.com/spectrometer-part-4-the-optical-bench/

That is to get the projection from the input slot at the sensor as optimal as possible for the wavelength calibration.

My guess is that this is a similar problem as designing a camera lens with low distortion.

 

made me realize that it is a spherical deformation.

I tried to correct the image with photoshop photo filters, but I couldn't.

 

Grazie mille a tutti

Thank you all @Andy Perrin @Stefano @colinbm @ulf

Antonio

[colinbm]

Neon is a gas. Neon lights are usually yellow-red.
The fluorescent tube lights & these Compact Fluorescent Lights CFL have Mercury vapour when on.

[ulf]

I built my wavelength calibration lamp around something like this:

https://www.ebay.com/itm/255009309321?var=554940242348

Warning This is  emitting hazardous UV-C and should not be used without proper protection!!!

My lamp is completely enclosed in a metal box and the only output is via a SMA fibre connector.

 

As I wanted a simple design it has a DC input connector for power instead of an internal battery.

The power switch is momentary.

That is useful as the relation between the levels of the different peaks vary over time and also give a bit of extra safety.

It must be pushed for the lamp to be on.

 

A lamp like this will likely not be especially useful for photography as the main peak is far into UV-C.

It is also creating ozone and need good ventilation. Both Ozone and UV-C are bad for your health.

 

When I disconnect the fibre after a calibration I can smell a faint smell of ozone.