Bin 1 vs Bin 2 for ASI294MM pro

Topic starter

Jul 29, 2021

Hi All
I just wanted to see what everyone's experience has been with different binning options available on ASI294mm. I'm new to astrophotography and read few discussions online but it's hard to know the benefits specifically for ASI294mm which was primarily designed to use 2x2 mode. I've been primarily using Bin 1 but given the file sizes integrations have been painfully slow. I recently tried Bin 2 and was impressed with the processing speed and trying to understand what I am missing out by not using Bin 1 mode. I understand the resolution part and thinking of using Bin 1 for 100-200mm wide angle shots and Bin 2 for longer focal lengths (800mm). Would love to hear what others have to share on this.

Respectful Engaging

LV426

Jul 29, 2021

I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

Helpful

Olaf Fritsche

Jul 29, 2021

I must admit that by now I don't worry about oversampling (too many pixels for the object) and undersampling (too few pixels for the object).

I always record with bin 1. When processing images with StarTools, I usually reduce to bin 2 by software. From the result, it makes no difference, because even in the ASI 294, the pixels are probably only calculated together.

The disadvantage of Bin 1: You get very large files.
The advantage of Bin 1: For really detailed objects the camera captures the detail information.
From Bin 1 you can always calculate to Bin 2. The other way around is not possible.

Helpful Concise

Dan Bryan

Jul 29, 2021 · Edited: Jul 29, 2021

I've really only used my 294mm pro twice now due to the weather, moon, smoke in the sky. Coming from a Sony mirrorless camera,
the 294 is a whole new beast for me. I don't think my old laptop would handle the files very well at Bin1. Till I know more and get some time on the 294 I will use Bin2 @ 4.63u. I'm using the EvoStar 80 with flattener which comes out to about 510mm.

Jonny Bravo

Jul 29, 2021

I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

I've got the same resolution: 1.25"/px in bin1 and 2.5"/px in bin2. I've used both modes, and to be honest, I'm still experimenting. My current project is M16, and for that I'm imaging in bin1 mode. However, I've also done M101 in bin2 and then a 2x drizzle integration. My bottleneck is definitely my processing computer. Dealing with those 100M subs in bin1 mode with my 15" MacBook Pro (8 core i9, 16G RAM) is unpleasant. I'm seriously considering building a processing box.

Well Written Concise

Amrinder Singh

Topic starter

Jul 29, 2021 · Edited: Jul 29, 2021

I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

Thanks. How do you calculate the resolution? Also curious to understand your gain selection for either options. I used gain of 200 for my bin 1 shots and 120 for bin 2. I did see more noise stretching with gain 200 so planning to stick with 120 from now on.

Amrinder Singh

Topic starter

Jul 29, 2021

Olaf Fritsche:
I must admit that by now I don't worry about oversampling (too many pixels for the object) and undersampling (too few pixels for the object).

I always record with bin 1. When processing images with StarTools, I usually reduce to bin 2 by software. From the result, it makes no difference, because even in the ASI 294, the pixels are probably only calculated together.

The disadvantage of Bin 1: You get very large files.
The advantage of Bin 1: For really detailed objects the camera captures the detail information.
From Bin 1 you can always calculate to Bin 2. The other way around is not possible.

I tried to convert bin 1 to bin 2 in Astro Pixel Processor and it produced blotchy stars. I guess I need to shoot same target with both options to really see the difference for my kit (Vixen AX103S and Canon 70-200mm).

Jonny Bravo

Jul 29, 2021 · Edited: Jul 29, 2021

Amrinder Singh:
I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

Thanks. How do you calculate the resolution? Also curious to understand your gain selection for either options. I used gain of 200 for my bin 1 shots and 120 for bin 2. I did see more noise stretching with gain 200 so planning to stick with 120 from now on.

Resolution is calculated as follows:

(s/f) * 206.265

s = the pixel size of your camera's sensor
f = the focal length in mm of your scope (including any modifications by a reducer or Barlow).

Example using my kit of a 294MM Pro in bin1 mode and 0.8x reduced GT81:

s = 2.315
f = ( 478 * 0.8 ) = 382.4

(2.315 / 382.4) * 206.265 = 1.2487

Helpful

Monty Giavelli

Jul 30, 2021

I've read that some guys use Bin 1x1 for their luminance filter and Bin 2x2 for RGB filters. There are some pretty good videos on youtube that get into that.
I just bought a wider angle telescope that should deal with the small pixel size of the 1x1. Lots of experimentation to come.
Monty

Brian Meyerberg

Jul 30, 2021

I use bin 1 with the RedCat 51. I can’t believe the detail considering the 250mm F/L.

North American Nebula with a slice of Pelican

Jonny Bravo

Jul 30, 2021

Brian Meyerberg:
I use bin 1 with the RedCat 51. I can’t believe the detail considering the 250mm F/L.

North American Nebula with a slice of Pelican

That combo gets you a 1.91"/px resolution, which depending on your seeing conditions puts you either in the sweet spot or just slightly under-sampled. I suppose you could always do a 2x drizzle integration. Pretty sure my computer would curl up in the fetal position and just whimper if I asked that of it

Well Written Insightful Respectful Concise Engaging

Brian Meyerberg

Jul 30, 2021

I never drizzled before, might be a good idea with this setup. 2x or maybe even 4x.
thanks for the idea.

LV426

Jul 30, 2021

Higher gains usually reduce noise on CMOS cameras, at the expense of dynamic range (more color in stars/shades of grey in mono). I am getting pretty nice HA shots at gain 0 and bin1. I have not posted them here, so I can't link yet.

Amrinder Singh

Topic starter

Jul 30, 2021

Brian Meyerberg:
I use bin 1 with the RedCat 51. I can’t believe the detail considering the 250mm F/L.

North American Nebula with a slice of Pelican

Nice shot. I did the same region with bin 2. It was actually my first shot with bin 2 with canon 70-200mm lens at 200mm. My untrained eye could not tell the difference. Maybe shoot some Ha frames with both bin 1 and 2 for comparison. I don’t want to sacrifice the details considering the time investment into getting these shots but the integration is a pain. It can take a whole day to process 15 hrs of data on my MacBook Pro. Here’s my bin 2 shot.

Helpful Engaging Supportive

Amrinder Singh

Topic starter

Jul 30, 2021

Amrinder Singh:
I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

Thanks. How do you calculate the resolution? Also curious to understand your gain selection for either options. I used gain of 200 for my bin 1 shots and 120 for bin 2. I did see more noise stretching with gain 200 so planning to stick with 120 from now on.

Resolution is calculated as follows:

(s/f) * 206.265

s = the pixel size of your camera's sensor
f = the focal length in mm of your scope (including any modifications by a reducer or Barlow).

Example using my kit of a 294MM Pro in bin1 mode and 0.8x reduced GT81:

s = 2.315
f = ( 478 * 0.8 ) = 382.4

(2.315 / 382.4) * 206.265 = 1.2487

Thanks. So for my Canon 200mm it comes out to be 2.38 and for my Vixen Ax103s (825mm) it comes out to be 0.58. How do you know when it’s undersampled vs oversampled and which bin to select for each option? Sorry if I am asking too basic questions.

Well Written Respectful

Brian Meyerberg

Jul 30, 2021

Amrinder Singh:
Brian Meyerberg:
I use bin 1 with the RedCat 51. I can’t believe the detail considering the 250mm F/L.

North American Nebula with a slice of Pelican

Nice shot. I did the same region with bin 2. It was actually my first shot with bin 2 with canon 70-200mm lens at 200mm. My untrained eye could not tell the difference. Maybe shoot some Ha frames with both bin 1 and 2 for comparison. I don’t want to sacrifice the details considering the time investment into getting these shots but the integration is a pain. It can take a whole day to process 15 hrs of data on my MacBook Pro. Here’s my bin 2 shot.

Beautiful shot! I agree with everything you stated. I might try a small region of interest to keep the file size and processing time down for a 2x drizzled image. Perhaps the Cygnus Wall. Will it be worth it maybe, maybe not. I’m really satisfied with the details of my bin 1 non drizzled image with only 5.5 hours of total integration. Seeing was average that night which helped.

Jonny Bravo

Jul 30, 2021

Amrinder Singh:
Amrinder Singh:
I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

Thanks. How do you calculate the resolution? Also curious to understand your gain selection for either options. I used gain of 200 for my bin 1 shots and 120 for bin 2. I did see more noise stretching with gain 200 so planning to stick with 120 from now on.

Resolution is calculated as follows:

(s/f) * 206.265

s = the pixel size of your camera's sensor
f = the focal length in mm of your scope (including any modifications by a reducer or Barlow).

Example using my kit of a 294MM Pro in bin1 mode and 0.8x reduced GT81:

s = 2.315
f = ( 478 * 0.8 ) = 382.4

(2.315 / 382.4) * 206.265 = 1.2487

Thanks. So for my Canon 200mm it comes out to be 2.38 and for my Vixen Ax103s (825mm) it comes out to be 0.58. How do you know when it’s undersampled vs oversampled and which bin to select for each option? Sorry if I am asking too basic questions.

The seeing plays an important factor in whether you're under-sample, over-sampled, or in the sweet spot. It's like Goldilocks LOL. For example, the exact same resolution can go from under-sampled to over-sampled depending on your sky conditions. In general, with "OK" seeing (stars are between 2" and 4" FWHM), a resolution between 0.67"/px and 2"/px is considered properly sampled. The better your seeing (i.e. the smaller the FWHM of the stars), the smaller the resolution you want to maintain "good" resolution.

The vast majority of us have "OK" seeing. Some are lucky and have "Good". It's the rare case indeed where you've got "Exceptional" (think mountaintops in Chile). Live in an area with high humidity and a turbulent jet stream, you might even have "poor" or "very poor" seeing.

So, how do you determine your seeing? Well, the easiest way is with your eyes. Look at the stars. If they're twinkling like the lights on your Christmas Tree, your seeing is not good. On the other hand, if they're steady and you can't really discern any twinkle, your seeing is pretty good. A more accurate way is to take some images and run a script like FWHMEccentricity in PixInsight to get an idea of what your stars look like.

Well Written Helpful Engaging

Amrinder Singh

Topic starter

Jul 30, 2021

Amrinder Singh:
Amrinder Singh:
I am using bin1 with my 80mm refractor at 384mm FL so I can have camera resolution of 1.25"/px. Otherwise I wind up with 2.5"/pix which is WAY undersampled. It has worked out pretty well I think. On the other hand, I am using bin2 with my 860mm FL refractor to get 1.11"/px. I really like this combination and is why I got this camera. It just works out that the Jekyll/Hyde nature of the camera is like having two for the price of one. I do have to change the gain to 0 when using bin1 and 120 when using bin2 to optimize noise. File sizes are obnoxious. I have 32GB ram and am thinking about moving up to 64 to handle it better.

Thanks. How do you calculate the resolution? Also curious to understand your gain selection for either options. I used gain of 200 for my bin 1 shots and 120 for bin 2. I did see more noise stretching with gain 200 so planning to stick with 120 from now on.

Resolution is calculated as follows:

(s/f) * 206.265

s = the pixel size of your camera's sensor
f = the focal length in mm of your scope (including any modifications by a reducer or Barlow).

Example using my kit of a 294MM Pro in bin1 mode and 0.8x reduced GT81:

s = 2.315
f = ( 478 * 0.8 ) = 382.4

(2.315 / 382.4) * 206.265 = 1.2487

Thanks. So for my Canon 200mm it comes out to be 2.38 and for my Vixen Ax103s (825mm) it comes out to be 0.58. How do you know when it’s undersampled vs oversampled and which bin to select for each option? Sorry if I am asking too basic questions.

The seeing plays an important factor in whether you're under-sample, over-sampled, or in the sweet spot. It's like Goldilocks LOL. For example, the exact same resolution can go from under-sampled to over-sampled depending on your sky conditions. In general, with "OK" seeing (stars are between 2" and 4" FWHM), a resolution between 0.67"/px and 2"/px is considered properly sampled. The better your seeing (i.e. the smaller the FWHM of the stars), the smaller the resolution you want to maintain "good" resolution.

The vast majority of us have "OK" seeing. Some are lucky and have "Good". It's the rare case indeed where you've got "Exceptional" (think mountaintops in Chile). Live in an area with high humidity and a turbulent jet stream, you might even have "poor" or "very poor" seeing.

So, how do you determine your seeing? Well, the easiest way is with your eyes. Look at the stars. If they're twinkling like the lights on your Christmas Tree, your seeing is not good. On the other hand, if they're steady and you can't really discern any twinkle, your seeing is pretty good. A more accurate way is to take some images and run a script like FWHMEccentricity in PixInsight to get an idea of what your stars look like.

Thanks. I mostly get average seeing based on Astrospheric app data. I’m working on a Cygnus area wide angle shot and will have more than 30hrs of Bin 1 data. Trying to stack individual data as it will bring my system down 😁

LV426

Jul 31, 2021

I use Astronomy Tools website for some of this. For me, the big factors to decide sampling is my mount accuracy and the steadiness of my skies.

starfield

Jul 31, 2021

I'm still experimenting, but do find for some objects the extra resolution pays off. Took one of M13 and was really happy with the extra detail in the stars. I have a 550mm length scope and with good seeing should be able to take advantage of the extra resolution. You can see the pic here: https://cdn.astrobin.com/solutions/images/54791/2021/f4216bb7-98d6-415e-a0d2-7242f7d90299-1624732400.png

Another tool I've found helpful is the CCD calculator found on the astronomy tools website. Helps show impact of seeing with given bin setting. Here's a link: astronomy.tools

Helpful Engaging Supportive

Amrinder Singh

Topic starter

Aug 2, 2021

So I did a quick test this weekend but forgot to take some dark and bias shots. Basically I took 3 sets of shots (6 x 300sec subs Ha subs)
1. Bin 1 Gain 120
2. Bin 2 Gain 120
3. Bin 2 Gain 0
Also In APP I added these 2 processed shots
4. Bin 2 Gain 120 with 0.5 Drizzle setting
5. Bin 2 Gain 120 with 0.1 Drizzle setting

I've never done drizzle before so not sure if I did it correctly. I simply changed the droplet size from default 1 to the above 2 numbers.

Some of my observations:
- The drizzle seems to have no impact
- Bin 2 provides much better exposure
- Bin 2 loss on resolution was not significant (atleast for this wide angle, it may be more perceivable in higher focal length?)

What do you guys think? Adding zoomed in and full version pictures of each scenario. I can share tiff files as well if needed.

Brian Meyerberg

Aug 2, 2021

Excellent detective work and research! From my findings bin 1 mode (pixel size 2.315)for the Asi294mm with my RedCat made a huge difference over using the asi1600 with pixel size of 3.8 microns. My stars are pinpoint and round. With the larger pixel size my stars were terrible. Maybe the nature of the 1600 with the micro-lensing? Either way I’ll sacrifice sensitivity for detail and I guess I won’t drizzle unless I want to zoom in on a region of interest. Below image Bin 1 with the Asi294mm.

North American Nebula with a slice of Pelican

Helpful Insightful Respectful Engaging

Amrinder Singh

Topic starter

Aug 2, 2021

Brian Meyerberg:
Excellent detective work and research! From my findings bin 1 mode (pixel size 2.315)for the Asi294mm with my RedCat made a huge difference over using the asi1600 with pixel size of 3.8 microns. My stars are pinpoint and round. With the larger pixel size my stars were terrible. Maybe the nature of the 1600 with the micro-lensing? Either way I’ll sacrifice sensitivity for detail and I guess I won’t drizzle unless I want to zoom in on a region of interest. Below image Bin 1 with the Asi294mm.

North American Nebula with a slice of Pelican

I feel I’ve to repeat the test with another target maybe North America nebula and maybe with my Vixen Ax103s scope to see the actual detail loss. Great shot by the way. How were to able to get so much blue in your shot? I was only able to get a small area blue. Did you expose more for OIII?

Brian Meyerberg

Aug 2, 2021

Thank you.
All 3 filters pretty much the same exposure time and number. I like to use color masks in my workflow to bring out curtain colors.
I also used a super luminance in that one. Maybe that helped with color saturation.

Jonny Bravo

Aug 2, 2021

A test I keep meaning to do is to collect data at bin1, then collect data at bin2, but do a 2x drizzle integration. In both cases, I would end up with a final image at 8288x5644. I've done both independently in the past, but I haven't shot the same target with both methods.

Here's a side-by-side of M101. The one of the left is bin2 not drizzled. The one on the right is the same data, but 2x drizzle integrated.

Obviously drizzling bin2 data makes a difference. Yes, I realize you need to zoom in quite a bit to see that difference. The question is: how does that drizzled data compare to data taken at bin1?

Helpful Insightful Respectful Engaging

What binning you use most with your ASI294MM pro