Physics Application

[Alan Sailer]

So I apologize for this post if it is too broad.

I am doing a demonstration of the classic quantum mechanics phenomenon of the double slit single photon experiment. The set-up is working and I am seeing what I would expect.

I want to use SharpCap to show the single photons adding up to create an interference pattern. I have an .avi video file that shows the "random"
photons. I want to make a movie showing in "real" time the photon hits adding up to show a clear pattern. I'm running into problems doing this.

First, I was intending to use the virtual camera feature of SharpCap to access a folder containing frames from the .avi video file. I tried using VLC to do this .avi to .png conversion but could not get any output from the program. I then tried some sketchy free websites to convert the .avi file into .jpg. This data ran on SharpCap but I can't see any of the photon hits when I live stack the .jpg files. I suspect the .jpgs are not high enough quality to show the rather dim hits.

In a final note, an astronomy program like SharpCap would seem to be the ideal tool to do this. The video camera I am using to image the photons has hot pixels which I should be able to subtract (dark frame). the live stack should show me the patterns as they slowly appear.

As a side note I did use another stacking program (that does not have live stacking). The final image file clearly shows the interference pattern. But that program does not have the live stacking feature that I want.

Any clues or help are welcome.

Cheers.

[admin]

Hi Alan,

A few things come to mind that might help...

In live stacking, make sure you have turned off alignment, as it will otherwise be looking for stars to align on and fail to add the frames as it will not find any. With alignment turned off, it will just stack the frames - that might help.

Jpeg is not a good format for faint data, as it uses lossy compression, and it's quite likely that the compression is throwing away some of (or most of) the information you are looking for - an area slightly brighter than the surroundings against a nearly black background will be likely to be wiped out by jpeg compression. Png is a better option - that will make bigger files, but uses lossless compression.

In recent versions of SharpCap (ie 4.1), you can load an existing SER file (not AVI sadly) into 'Test Camera 2 (High Speed)' via the 'Browse' button in the Testing controls. The camera will then play the SER file repeatedly. You can then turn up the exposure and turn down the gain to slow down the playback of the video without changing the overall brightness - (ie 10x higher exposure, set gain 10x lower). Then you should be able to turn on live stacking I think. You should be able to use a tool like PIPP to convert your existing AVI to SER format. Unfortunately the PIPP website has gone offline, but there are reliable download links on this forum post : viewtopic.php?p=38025

Hope that helps,

Robin

[Alan Sailer]

Eobin,

Thanks. It gives me other stuff to try. I'm also thinking about buying the frame grabber described in the users manual. Another long shot is to see if I can find an astrophotographer with a compatible camera who wants to help.

Another question. Is there a cheap 1/3 format CCD astro camera out there that will run on SharpCAp?

Cheers.

[admin]

Hi Alan,

if I'm reading the sensor size table right then 1/3" is roughly a 6mm by 4.8mm sensor. That's a bit bigger than is typical for entry level astronomy cameras, but something using the sony IMX290 sensor would have a sensor area of about 5.5mm by 3.1mm (1920x1080 with 2.9 micron pixels). Such cameras can be purchased for about $130/£130 or so (ie https://www.svbony.com/sv305-astronomy-camera).

Note that almost all modern astronomy cameras are CMOS rather than CCD - for the last decade or more, vast amounts of effort have gone into improving CMOS cameras to support digital cameras, mobile phones, etc and practically no research has gone into CCD. CMOS now matches or out-performs CCD for most use cases.

cheers,

Robin

[Alan Sailer]

Thanks again. I ordered the frame grabber so that will be the next try. Wish me luck.

Cheers.

[Alan Sailer]

I got the frame grabber yesterday and got it running with SharpCap. I ran the single photon set-up and got a stack going.

I wasn't seeing what I hoped for and I think I know why. After looking into stacking it appears to me that stacks are not what I need.

Correct me if I am wrong. Stacking is a process that tends to eliminate noise and amplify signal. Noise, being random will sometimes light up a pixel and sometimes not. A star (or other non-noise light sources) will always light up a pixel. Stacking tends to cancel out the random variations and add up the non-variable pixels ie the stars and nebulae.

My data, since it just shows up as random hits looks like noise to the stacking process. Rather than stacking I need something that integrates successive frames and gradually shows an apparent random series of hits slowly adding up to a pattern.

It's possible that SharpCAp has a feature that allows integration of frames. If so, please let me know.

Cheers.

[admin]

Hi Alan,

internally, the stacking procedure adds the values of each frame to the data representing the progressing stack and then produces an output that is scaled based on the maximum value in the stack data so far, so in theory your approach should work given enough frames - if all of the other pixels produce a value of zero, but one area sometimes produces a value of (say) 15, that area in the stack will built up to higher values.

Of course, hoping that all the other pixels will be zero is wishful thinking - there will be random noise and the stacking procedure will slowly reduce the value of any truly random noise in the stack relative to any true signal - four times as long in stacking will increase the signal by a factor of 4, but the noise by a factor of only two.

What you may need to do is to apply a very strong stretch to the stacked data using the live stacking histogram. Move the middle yellow dashed line to the left - the horizontal position of that line determines the brightness level in the raw stacked data that will appear as 50% brightness in the viewed image. You can use the options for a logarithmic scale on the horizontal histogram axis and the horizontal zoom to give better control of this and allow for more extreme adjustments. You can also drag the left hand yellow dashed line to the right somewhat to ensure a steep curve gradient (very high gain) in the right area of the data.

Hope this helps,

Robin

[Alan Sailer]

Robin,

Thanks again.

I tried the suggestions you gave (never would have found them) and it helped very much. I'd say I'm getting 90% of what I wanted. I can turn down the photon flux until it is just a very light speckle and then stack and get a recognizable interference pattern is about a minute. Just like the original demo shows.

I may mess with the dark frame feature to see if I can get rid of a bunch of hot pixels. But that would just be an appearance thing.

Cheers.

[Alan Sailer]

Robin,

Here is a short video showing the use I put your program to. It's been a long path to get here. The hardware went together over a few weeks of trial and error. The software aspect was driving me nuts. With your help I am sane again

https://www.flickr.com/photos/8763834@N02/53334653284

Cheers.

[admin]

Hi Alan,

glad it all worked - looks like a very good signal in the video

Back when I did my Physics undergrad (late 80s, early 90s), getting your optics practical results usually involved spending time in the dark room developing the film with the distinctive smell of developer and fixer. Not an imaging sensor in sight back then (at least not in the undergrad labs).

cheers,

Robin