2 replies to this topic

[plofstof]

I have been doing DSO for quite some time but only recently started playing with deeper focal lengths, after I got an EdgeHD 11"

 

The Rayleigh limit of the EdgeHD is advertised by Celestron as 0.49" and the Dawes limit as 0.42"

As far as I can tell Rayleigh and Dawes try to establish the same thing.  Rayleigh through a mathematical definition (which I read through and makes sense to me) and Dawes apparently tried the same thing through subjective observation and experimentation.

 

So. I attach two framings of M104 from Telescopius, one at a focal length of 1960mm (that is, using my focal reducer) and another one at the native focal length of 2800mm.

As you can see in the images, the 1960mm resolves to a pixel scale of 0.4"/px on my sensor, and the native 2800mm focal length would translate to 0.28"/px which is well below the OTA's Rayleigh or Dawes spec.

 

Now, just to simplify the present discussion let us assume I can achieve perfect guiding with an RMS below 0.28" (I know I probably can't! I just don't want to distract from my primary question)

 

Then, I would then like to paraphrase my understanding of Rayleigh/Dawes's implications as follows:

 

Because the OTA cannot resolve finer details than 0.49"/px, it would be a waste of my time and effort to try and use the full focal length of the OTA (and not the focal reducer) for DSO purposes.  Sure, I will end up projecting a larger image on my sensor, and I will think I am capturing more pixels on the target, but the actual information represented in those pixels will not be more "useful" than an image shot through the reducer, as the scope is incapable of resolving details at a finer level than ~0.49"/px

 

Put differently, if shot an image with the reducer at 1960mm and resampled it to increase it in size by a factor of 2800/1960=~143%, then I would look at about the same image I would get from a perfectly guided 2800mm session.  Which is to say that, with my sensor, an image taken at 2800mm FL will always be oversampled.

 

Is this the correct way to talk about it?  Please critique my understanding...

If the above is a correct formulation, it would probably never make sense to use the EdgeHD without the focal reducer for DSO purposes on my IMX571 sensor (which seems to be the consensus out there, I am just checking as to why)

 

Thanks!!

 

 

Attached Thumbnails

• 
• 

[SilverLitz]

Your image scale needs to be multiple times finer than your systems resolution capability to be properly sampled.  This is why audio recordings have a sampling rate of 44.1Khz for CD and higher for higher quality recordings, even though human hearing is thought to not exceed 20Khz (for younger listeners).  Nyquist requirement is for sampling to exceed at least 2x the highest frequency to accurately reconstruct a 2D audio waveform.  Imaging is 3D, so a higher multiple is needed, and we think in image scale (arcsec/px) which is ~ the reciprocal of frequency.  

 

The OTA's resolution limit is just one of several "filters" that limit resolution, with additional limits being atmospheric "seeing" and mount tracking accuracy.  Generally seeing has the greatest limit.  A good rule of thumb is critical sampling being Seeing/3.   So with 1" seeing conditions, you should realize increasing resolution down to ~ 0.3"/px (either native or drizzle).  My resolution champ being my SV152 (LZOS lens cell, 1216mm @ f/8) using 2x drizzle combined with my QHY268M for 0.64"/px (0.32"/px drizzled).  My local seeing estimates by Meteoblue are typically 1.5" (0.7" to 2.2" typical range), and I have consistently seen my detail get better down to about ~0.3"/px with no improvement at finer image scales.

Edited by SilverLitz, 23 June 2025 - 11:13 AM.

[astrokeith]

SolverLitz has put it well. Also note that we routinely exceed these two limits when we observe the Cassini division.