Elephant Trunk...

[ricksastro]

In keeping with my round-about of equipment, I've replaced most of my setup again     Here's the latest:
http://www.ricksastro.com/Equipment.htm


Here's my first complete image with my new setup...I always wanted to do a nicely framed version of the elephant trunk and this combination allowed me to get it from my head and onto an image.  I've always had a fondness for the subtle shading and softness a straight LRGB gives on this one rather than narrowband...besides, I can always do narrowband when the moon's out


Details and highest res shot here:
http://www.ricksastro.com/DSOs/vdb142_st10.shtml


Rick Krejci

[Jake]

Ok - please tell me that you have at least looked through that TEC 140...   

That's a really great scope - your current setup looks really "together".   

Last weekend, there was a man up at Mt. Pinos who had an 8" TEC apo, setup with
binoviewers.  I looked through it - talk about a deep space refractor!  I have never
seen an 8" apo before that - let alone looked through one.  It was an experience.   

It made me a believer in the TEC optics - I am sure you will love that scope!

Your image looks wonderful - there is alot of color depth you have captured, especially
in the dark areas of the nebula.    It really has a natural look - your processing is superb!
I notice a small diffraction spike on the brighter stars - any idea what is causing that?

Fantastic job!

-Jake

[lindendave]

You really put my dial-up to the test,  but well worth the wait Rick! Nice to see you even updated your equipment page too! 

[twilbur]

Wow Rick! Super ultra deep, detailed, just fantastic! I could look at that shot all day.

[ricksastro]

One of the many motivating factors in getting the TEC140 was Mars later this year...viewing, not just imaging.  In fact, the larger scopes are better for imaging, but for visual, I'm looking forward to a little refractor heaven...plus the thing just looks so darned cool 

The XME version of the 10D gets those little vertical diffraction nubs...part of the microlensing thing.  It gave me hesitation on getting the XME vs. the XE for that reason.  Steve's new camera is the microlensing one as well. 

But the sensitivity gain is great!   Here's a mere 6x5 minute set of exposures (30 minutes) of the Pelican area at f/7:
http://www.ricksastro.com/temp/pelha50tecf7.jpg

I was pretty amazed!  I am starting to now be a believer that aperture (or more accurately, collecting area) plays a more important role than f ratio in Signal/Noise of an object.  The mainstream thinking is that aperture is king in resolving stars (point sources) and f ratio is king for extended objects (nebula, galaxies).

But this seems mostly to be a film notion.   As long as your exposures are long enough to overcome read noise, if you look at a given area of the sky at the same magnification taken with scopes with the same aperture, but different f/ratios (and thus focal length), they will all look very similar.  Note that the images taken with the faster, lower focal length scopes will be wider field/less magnified and will look less noisy at the native image scale.  But if you shrunk an area from the slower scope to the same image scale as the faster scope, they would have similar S/N.  Same if you zoom in the image of a faster scope to higher magnification.

My case in point is the E160 vs. the TEC140.  The E160 has a large central obstruction which reduces light gathering, so the 2 scopes probably have similar ight gathering area.  But the E160 is f/3.3 and the TEC140 is f/7.  With conventional thinking, it should take me over 4x as long to get a similar depth and smoothness with the TEC.  But even with the St2000, which has 1/2 the Ha sensitivity of the SXV-H9, I was getting great depth (take my crescent http://www.ricksastro.com/DSOs/NGC6888_TEC.shtml and http://www.ricksastro.com/DSOs/Crescent.shtml as an example).  The SXV version was 1/2 the exposure, but was taken with a doubly sensitive camera.  The TEC version is smoother and deeper to my eyes, and it should be 1/4 the S/N of the E160 version.

The full (and much better explained with a better example) is found on Stan Moore's great page:
http://home.earthlink.net/~stanleymm/f_ratio_myth.htm

The advantage to going with a faster scope is the wider field of view, not necessarily the faster accumulation of photons for a given feature.


Rick

[twilbur]

Rick,

Thanks for the excellent viewpoint on this subject. It's also been a matter of recent discussion on the Cloudy Night's forums, how do focal length, aperture and exposure time interrelate, with respect to imaging point sources and extended objects. The term "light recording power" being used as defined by the equation LRP=r^2/f^2 (r is radius of aperture, f is f ratio). But then there's the effect of length of exposure to account for and with CCD imaging the sensitivity of the camera, size of pixels and size of chip play into it too.

Your example is an interesting one, you have 2 scopes about the same aperture, but one is fast (160) and one is slower (Tec). You took an image on the faster scope with a faster camera with 1/2 the exposure time and the other with a less sensitive camera on a slower scope at twice the time. According to the formula it should have taken 4x as much time to reach the same SN level, and yet, twice the time was enough to produce an image with less SN. Of course seeing could have had an impact too.

Thanks for the link to Stan's article, I need to read it, I'm still trying to get my head around the conclusions we can draw based on the characteristics of the various components.

BTW - Nice Pelican! You should grab some color before the moon comes back!

[Jake]

Wow Rick, lots of stuff here!

That will be a real treat to observe Mars through the refractor this year.
I agree that it looks way cool - there is something very pleasing about a nice
refractor (both to look through, and look at) - it looks like a "proper" telescope!   

That theory is really interesting about focal ratio vs. S/N.  I have always understood
the traditional film model.  This is one of the reasons I stay with my 8" f/5 - good
photographic speed, and reasonable image scale and resolution at 1000mm FL.  I have always tried
to choose my imaging targets based on my image scale. 

I read your explaination, and the article.  I can't argue with the results, and the theory
is starting to make sense to me - it is totally a new idea to me, and is fascinating.  One thing
I am still trying to understand is the statement Stan Moore makes:

"Focal length (and thus f-ratio) has absolutely no effect on the number of photons collected and delivered."

While this statement is absolutely true, the part that sticks with me is that although the same number of photons
are collected for a given aperture regardless of focal length (and f-ratio), isn't it also true that by increasing the focal
length of a given aperture, the distribution of photons per given area decreases?  In other words, by increasing image
scale per a given aperture (i.e. increasing focal length), isn't that effectively decreasing the amount of photons hitting each pixel per second, because although the same amount of photons are collected per an aperture, a longer (slower) focal length is distrubuting those
photons over a wider area?

I am not doubting the theory at all, just trying to understand it better.  It is good news for me,
as I purchased a 4" apo a few months back (and boy do I love it! ) that is f/9.  I bought it
to use as a visual instrument, thinking it would probably be to slow to use with my camera. 
I would love to use it for imaging - it would give me a slightly wider fov than the 8", and would
inherently give me a much flatter field, so that I wouldn't have to worry about things like my recent
brain-fart of forgetting the paracorr.      Do you think that I could obtain some nice images with
my apo and digital rebel?

Also, your pelican looks reallly great! 

-Jake

[ricksastro]

Sorry for the delayed response...I was at cub scout camp with my youngest...

Anyway, I'm still working on this whole thing myself. 

The way I understand it is that a fixed (by aperture) number of photons per unit sky area (let's say square arcminute) are being collected and delivered.  As the focal length decreases (f/ratio gets faster), there are more overall photons being collected (wider area being covered), but for that square arcminute, the same number of photons are delivered. 

If your "perfect" camera had 0 read noise and 100% QE, you really should choose the focal length that best matches your camera resolution and desired target.  A short focal length will be wider, but may be undersampled.  But if you want an image of that wider target, your only alternatives are to use the shorter focal length and give up resolution or use a larger camera chip (with the same size pixels).

When the reality of read noise comes into play, it starts to make a difference.  You need an exposure time long enough to overcome the read noise by a large margin (say 10x).  What you say is correct...you are spreading that square arcminute of the sky over more pixels, so you have to expose longer to overcome read noise, and yes, this amount of time will be determined by f/ratio.  So and f/10 scope may have to use 10 minute exposures, while the f/5 would only need 2.5 minutes.  But since these exposure times were chosen such that the background is much larger than read noise (10x), then the 10 minute exposure would have 9 minutes of beyond read noise recording and the 2.5 minute exposure would have about 2.25 minutes.  if the camera is 50% efficient (QE of 50%), the 10 minute exposure would contain 4.5 minutes of exposure of the "perfect" camera.  The 2.5 minute about 1.125 minutes equivalent of a "perfect" camera.  So if you stacked 4x2.5 minute exposures from the f/5, the same number of pixels in that sky area would be recorded as the 10 minutes at f/10 coverage of that same area.  Now, if you took a single 10 minute image at f/5, you will gain a small advantage...since the additional 7.5 minutes would not be trying to overcome read noise, you would have 9.75 minutes beyond read noise recording, or 4.875 min equivalent to the perfect camera...about a 8% gain over the  f/10. 

Note that you'd get almost the identical photons collected for that area by binning the camera at f/10, since really the only advantage of binning is reducing read noise.  Also not that, as exposured get longet, the difference is less.  If you are using short exposures, f/ratio makes a much larger difference.

If you view the a images --at the same magnification--, the f/10 image will cover a smaller are but with possibly greater resolution.  The f/5 image will be larger, but the look af that single area of the sky will look quite similar in both.

So, if you're looking for optimum resolution, choose a focal length which gives a arcsec/pixel value in the range of .75-1.25, and then get as big an aperture as possible to collect as many photons as possible.    If you want wider sky area coverage, choose your focal length for that and get the largest aperture you can. 

[Ron]

Rick,

That's a terrific image, another one of those worth framing images

As far as the technical discussion it's pretty much over my head but still interesting.   You always seem to come up some of the best links to help a person in understanding the technicles.

One Question?  After reading through all of the discussions, I didn't see any mention of PEC or tracking.  How would these factors enter in to the choice of camera and telescope for a specific target or object?

Thanks,

Ron

[dsnay]

I'm with Ron on this one.
Great image!
Way too much information in the discussion for me to understand at this point. I'll get there, but not yet.
I've still got too much to learn from the DSI Pro I've recently picked up.

Anyway, it never ceases to amaze me how easy folks like you and Steve make these wonderful images.
The smoothness is just incredible!

thanks for sharing,
Dave

[ricksastro]

One Question?  After reading through all of the discussions, I didn't see any mention of PEC or tracking.  How would these factors enter in to the choice of camera and telescope for a specific target or object?

If you're limited to 30 seconds or a minute of tracking, f/ratio will indeed make a difference...you have to get anough photons converted to electrons per pixel to overcome the read noise, and a faster f ratio will concentrate more photons per pixel. 

The summary of all this gobbeldygook is, if you can take long exposures (5-15 minutes), f/ratio isn't as much the determining factor for Signal/Noise as aperture.  As exposures get shorter, the f/ratio advantage comes into play.

Cameras with low read noise (SXV-H9, ST2000) can get nice S/N with shorter exposures as well, whereas cameras with large read noise (ST7, ST8) require longer exposures to get smooth results.  When I was using the SXV-H9 with the E160 f3.3, I was getting great results with short (1 minute) exposures. 

And with narrowband imaging (Ha, OIII), since it takes much longer for the sky background to get above the read noise level, this effect is magnified and faster focal ratios are more important.  With a reasonably fast (f/6 or f/7) scope, a camera like the ST8 will  still require exposures in the 15 minute range to get smooth Ha results.   At f3 you could drop to 4 minutes or so.  With the SXVH9 and the e160, even 1-2 minute exposures were sufficient.

So, Ron, you are right to bring up the mount capabilities into the equation.  Local sky pollution comes into play as well.  The more you understand the factors and how they play together, the more you can make good decisions on equipment and  methods that fit your situation.

Rick

[Ron]

Thanks very much Rick,

You have given a lot of information here that will take me awhile to digest.  It does give me the valuable information to not make a hasty decision when it comes to buying equiptment, all the factors enter in to the final equation.

Ron

[twilbur]

Rick, my thanks too for helping to shed some light on this subject. I'm still not totally clear on it but I think your explanation of reducing readout noise makes sense.

I'm having difficulty understanding statements like:

f/ratio isn't as much the determining factor for Signal/Noise as aperture

Since increasing aperture for a given focal length reduces f/ratio. It still seems to me that you'll get better S/N using a faster f-ratio for a given FL (per unit time). Perhaps part of the problem here is that I don't know how to quantify / measure S/N in an image. It seems really subjective to me.

[ricksastro]

I'm having difficulty understanding statements like:

f/ratio isn't as much the determining factor for Signal/Noise as aperture

Since increasing aperture for a given focal length reduces f/ratio. It still seems to me that you'll get better S/N using a faster f-ratio for a given FL (per unit time). Perhaps part of the problem here is that I don't know how to quantify / measure S/N in an image. It seems really subjective to me.

You're exacty right.  If you fix the focal length, aperture and S/N and f/ratio will go hand in hand.

But it you keep the Aperture constant and vary the focal length and f/ratio (focal length longer = slower f/ratio), that's where all of this comes in to play.  Stan Moore's example shows this as clearly as any example I've seen.  Using the CN212 which is 8" aperture but has the option to be either f/3.9 or f/12.4 (and thus a fl of 820mm or 2630 respectively).  As you can see in Stan's pic, AT THE SAME VIEWING SCALE, the images look very similar.   

You could argue (successfully) that if you looked at the images at native resolution, the f/3.9 will look smoother...it will.    But the point is, if you shrunk the f12.4 image to the same image scale, it will also look similarly smooth.   Or if you blew up the f3.9 image to the f12.4 scale it would look similar.

So why if this important...why don't you just image at f3.9 and capture a larger area?   This also buys you the flexibility to take shorter images.   It will also be less demanding on your mount.  What's the downside?   

Resolution.   At 820mm, you may be undersampled for many cameras and will be missing some detail, particularly if seeing is good.   


When you look at it all, while the points are true, for the vast majority of the imagers out there it's not overly relevant.  Most imagers are better off sacrificing resolution for the flexibility to get good results from short exposures.   And if you're already well sampled (say < 1 arcsec/pixel), going with a longer focal length will not buy you any more resolution and will get you a smaller FOV.  And if you don't have an awesome mount (or an AO unit), guiding difficulties will limit resolution if seeing doesn't get you.

Bottom line, if you match your arcsec/pixel to your typical seeing and mount capabilities (1.25-2.25 arcsec/pixel)and then get the largest aperture possible, you will have an optimized setup for most of the situations.  Add a widefield lens or scope to the mix, and you cover most of the rest.   Another valid philosophy is to optimize your scope so that your target will fill the FOV the way you want. 

90% of this discussion comes into play if you want to go to the edge of resolution on the very best nights and have a very nice mount.   In this case, for the smaller targets, don't be afraid to shoot with longer focal lengths...you'll get better results on great nights, and no worse on medium seeing nights. 


Rick

[twilbur]

Rick, your input on this has really been good to help solidify it in my mind.

Bottom line, if you match your arcsec/pixel to your typical seeing and mount capabilities (1.25-2.25 arcsec/pixel)and then get the largest aperture possible, you will have an optimized setup for most of the situations.

I came to the same conclusion. And, as you have already mentioned, camera noise and chip area play into it too.

A big variable that makes CCD imaging so much different than film is that different cameras have different pixel sizes and it's the pixel size and focal length that determine image scale, which in turn has a huge impact over the exposure times you can use. In my opinion, those lucky enough to have huge budgets and great seeing can say focal ratio isn't important and can spend $30K+ on Paramounts and RCs. Those of us trying to get good results with a limited budget still need to consider it as a factor, but not necessarily a factor that's as important as it was when using film.

Thanks for your lucid explanations on this subject Rick.