Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

USB camera for Lunar imaging


furrysocks2

Recommended Posts

I've been enjoying imaging the moon with my 8.5" newtonian, most recently with a PS3 Eye camera. This has a 1/4" Omnivision sensor, 640x480 at 60fps, giving 1/8 degree (width) fov. Without tracking, I have to crop the videos but I'm pleased with the results so far.

moonps3eye.png

 

I'm looking at trying to get a wider field of view and more megapixels, and have found a C mount 5MP 1/2.5" Aptina microscope camera on aliexpress at about £35, giving my 1/5 degree fov - 60% more than the 1/4".

https://www.aliexpress.com/item/5-0MP-USB-Cmos-Camera-Electronic-Digital-Eyepiece-Microscope-Free-Driver-Measurement-Software-High-Resolution-for/32741153822.html

It appears to be able to do:

  • 2592*1944 (5MP) @ 14 fps
  • 1920*1080 (2MP) @ 30 fps (taken from another listing)
  • [insert 60fps resolution here?]
  • 640x480 (0.3MP) @ 124 fps

 

I know I should probably bite the bullet any buy a "proper" astro camera and I could double those rates and have higher sensitivity but as a cheap, dedicated lunar camera, any thoughts or experiences with similar?

Link to comment
Share on other sites

  • Replies 26
  • Created
  • Last Reply

One thing I noticed from the link is that is has a 2.2um pixel size. This means you will be oversampling a lot. Not sure how and how much this matters for lunar imaging though.

Another thin to think about with a non astro camera is if you have enough control over gain. Not sure what the requirements are for a microscope, but the link sad nothing about min, max gain or other settings.

Link to comment
Share on other sites

13 minutes ago, Magnus_e said:

One thing I noticed from the link is that is has a 2.2um pixel size. This means you will be oversampling a lot.

Is this arc-seconds per pixel? Full res would be 0.28"/px, more with a focal reducer or reduced resolution. One of the reduced resolutions is a crop, the others I assume use binning.

Link to comment
Share on other sites

Well the sampling rate is given by

Arcsec/pixel = (CCD Pixel Size / Telescope Focal Length ) * 206

So if I just guess on a fl of 1200? I did see you had a large newt.

(2.2 / 1200) * 206 = 0.378.

From what I've seen on different post about this a sampling rate between 1 and 3 is good. 3 might be a little much, but 0.378 is very little.

You can add your actual fl to the equation and see what you get.

 

Worth noting that sampling rate is not all that should be considered, but it's nice to bee within a standard range.

Link to comment
Share on other sites

That's a very low sampling rate, and I don't think it would work well.

3 minutes ago, furrysocks2 said:

From newt-web:

The theoretical resolution (Dawes limit) for a 215 mm objective is 0.54 arc seconds.

That theoretical limit based on aperture is very generic i think.

You mus also considure fl and seeing. On the best dark sites you can possibly get a seeing around 1-2, but 3 would be more normal seeing.

I have asked some questions about this before, but I cannot say I understand it all, so cannot give any more tips than that.

Just remember that theoretical limits are very theoretical, and usually not achievable.

Link to comment
Share on other sites

Just now, Magnus_e said:

That's a very low sampling rate, and I don't think it would work well.

That theoretical limit based on aperture is very generic i think.

You mus also considure fl and seeing. On the best dark sites you can possibly get a seeing around 1-2, but 3 would be more normal seeing.

I have asked some questions about this before, but I cannot say I understand it all, so cannot give any more tips than that.

Just remember that theoretical limits are very theoretical, and usually not achievable.

Thanks, Magnus.

Something to chew on.

Link to comment
Share on other sites

Been googling on oversampling, copying here for reference - note that I cannot verify anything here.

 

http://www.alpo-astronomy.org/publications/Presentations/Owens - Planetary Imaging.pdf - Larry Owens (2006-ish?)

Quote
  • Get the "sampling" right
    • Ideal - half the Dawes Limit per pixel
    • Reduce, depending on atmospheric stability
Quote
  • Resolution is not the important setting
  • Sampling is really what matters
    • Nyquist sampling rule: 2-3 pixels over the highest possible resolution
    • Use the Dawes Limit of your scope
    • Raise the focal length so that 1 pixel covers half of the Dawes Limit
    • Somewhat dependent on seeing though - Experimentation
Quote
  • Compression issue
    • Above a certain frame rate the camera performs compression
    • Effects of compression not visible on video
    • Causes severe processing artefacts

 

http://www.astropix.com/wp/2011/03/23/image-scale/ - Jerry Lodriguss (2011)

Quote

To be correctly sampled you want your pixels to have an image scale of about 2x to 3x better than the smallest detail you hope to record.

...

To correctly sample 1 arc second detail, you really need an image scale of about 0.33 arc seconds per pixel.

 

Other stuff:

          Linked.JPG

 

Link to comment
Share on other sites

56 minutes ago, furrysocks2 said:

"If you are doing planetary imaging then there is benefit from a greater degree of oversampling or density of pixels as you are going to pick up better random moments of excellent seeing in very short exposures and weed them out for stacking." - Satchmo

 

This would suggest that oversampling would be good in your case.

 

To correctly sample 1 arc second detail, you really need an image scale of about 0.33 arc seconds per pixel.

This is the one I'm struggling with.

Lets say a bright point in the distance is one arc second and the seeing is exactly one arc second. You have a setup that gives one arc second sampling rate per pixel. At this point (in my mind) you are optimally sampling that one arc second. If you change your setup so one pixel samples 0.33 arc second then there still is a physical limitation in the seeing that makes sampling less than one arc second redundant. (It's not the first time I see this quote so its probably correct and I'm not getting it).

Considure the example above. with the seeing and distance, an that you are viewing the point with one eye. if your eye has a sampling rate of one arc second then you are perfectly able to see the point. If your eye then gets 3 times more sensitive (a rate of 0.33 arc seconds), then you are exactly as able to see the point as the seeing is one arc seconds. If you then become an alien with three 0.33 arc second eyes, then you are still exactly as able to see the point as before. Your brain would probably be reassured that the signal would have been validated three times over, but you could not resolve more.

 

I guess that the point at top of the random moments of excellent seeing is why a sampling rate of more than the physical seeing can be used to resolve more detail, but it's all a little over my paygrade :)

 

Looking at the quotes you added, the microscope camera might be well suited. I'll leave it up to someone else to come along and explain it all.

anyways, good luck in finding a new camera :)

 

Link to comment
Share on other sites

17 minutes ago, Mick J said:

Interesting cameras furrysocks2, ...

Which, the PS3 Eye or the microscope ones?

 

17 minutes ago, Mick J said:

... did you think of cooling the PS Eye.

Yes - it crossed my mind - I'm sure it's been done.

They're not perfect but with large pixels, ability to output raw, up to 10s exposure, high frame rates, etc...  at £1.50 - good to learn about camera limitations. ;)

Link to comment
Share on other sites

  • 4 weeks later...

Still waiting on delivery.

The one I got (cylindrical body) is now nearly double the price at £63 having paid £33, the square one now at £40. I don't know if it sends uncompressed frames at any resolution but if it does, the point at which it starts compressing may be the limiting factor. 

Having continued to take lunar shots at 0.3 and 1.3MP, I'm hankering after that slight increase in FoV and resolution.

Just have to wait...

Link to comment
Share on other sites

Aye, well...

 

Listing qutoed "YW500C" which a web search shows does indeed have specs "15fps @ 2592*1944;30 fps @1920*1080" as the seller quoted.

Delivered was "SX-EP500B" which a web search shows is "2fps @2592 x 1944, 4fps @2048 x 1536, 5fps @1600 x 1200, 7.5fps @1280x1024", or similar - much like what I now have.

 

Indeed the photos on the sellers site show the latter. Description was inaccurate.

Disappointed.

Link to comment
Share on other sites

Oh dear ! what a  a ,,  something unprintable :(

And I earlier* had the bright idea that you might have been able to find a generic uptodate driver with a google on 'microscope eyepiece driver' other than the default plug&play. Indeed there are many but that wont solve this now :(:evil4:

* would have suggested earlier as well but had a sudden break in the clouds !

 

Link to comment
Share on other sites

  • 2 weeks later...

Disputed based on the received item not having the framerates advertised. Rather than adding return postage into the equation for a full refund, proposed a partial refund for what I paid minus the cost of return postage, keeping the item.

Refund approved, just waiting to actually get the money back.

 

Still toying with the idea of taking a punt on the other one, although I gather the higher framerates are MJPEG and I don't know how they stack, not well I expect.

Link to comment
Share on other sites

On 13/01/2017 at 20:53, Magnus_e said:

Lets say a bright point in the distance is one arc second and the seeing is exactly one arc second. You have a setup that gives one arc second sampling rate per pixel. At this point (in my mind) you are optimally sampling that one arc second.

Igniring seeing, I think the idea of sampling at 0.3 of the Dawes limit is that the smallest details your scope can pull in will occupy three pixels in the image, and so are actually visible to whoever views the image. If sampling at exactly the Dawes limit the smallest details will sit on one pixel and will be difficult to spot in the final image.

Link to comment
Share on other sites

43 minutes ago, CraigT82 said:

Igniring seeing, I think the idea of sampling at 0.3 of the Dawes limit is that the smallest details your scope can pull in will occupy three pixels in the image, and so are actually visible to whoever views the image. If sampling at exactly the Dawes limit the smallest details will sit on one pixel and will be difficult to spot in the final image.

That's how I understand it. The smallest detail resolvable by the telescopes needs at least a 2x2 pixel grid to be imaged,  3x3 is about the max usable for imaging. Greater than 3x3 (4x4) doesn't resolve more details.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.