I just bought a Seestar S50 Pro. I am almost embarrassed to say it out loud.
And somehow, buying another ZWO product has made me more frustrated with ZWO.
That probably requires some explanation.
I’m not new to astrophotography, and I wasn’t looking for an easy way to get started. My current imaging gear includes an AT115EDT, AT60EDP, Celestron RASA 8, ASI585MC Air, ASI2600MC Duo, and a Rokinon 135mm f/2 with autofocus. Between those systems, I can cover everything from enormous emission-nebula complexes to extremely fast f/2 imaging and smaller targets at longer focal lengths.
I guide. I autofocus. I dither. I shoot calibration frames. I build mosaics. I process in PixInsight.
I’ve also owned the original Seestar S50. And I sold it.
The original S50 was clever, convenient, and genuinely impressive as a concept. But ultimately, I found the imaging mediocre compared with what I could produce with conventional astrophotography equipment. And honestly I slammed it a lot, making a lot of Seestar users upset.
So why did I just spend $900 on the new S50 Pro?
Not because I needed another telescope. My wife will attest to that….
I bought it because ZWO seems to be putting some of its most interesting computational-photography innovation into Seestar, while ASIAIR increasingly feels like it’s being left behind in that particular area.
And as someone invested in the ASIAIR ecosystem, that bothers me.
I Already Tried the Seestar Experiment and there was a lot I liked about my original S50.
Put it outside. Turn it on. Choose an object. A few minutes later, an image starts developing on your phone.
That’s an extraordinary accomplishment when you consider what astrophotography used to require.
For outreach, travel, casual observing, and simply exploring the night sky, the S50 was fantastic.
But I eventually sold mine for a simple reason: I wasn’t impressed enough with the images.
Once you’re accustomed to cooled cameras, good optics, higher resolution, proper guiding, dithering, autofocus, calibration, and serious post-processing, the compromises become obvious.
When I wanted to make an image I was genuinely excited about processing, I reached for my conventional equipment.
So when ZWO announced the S50 Pro, my initial reaction was obvious: I don’t need this.
And technically, I don’t. I Already Have Better Imaging Hardware.
My AT60EDP and ASI585MC Air give me a serious 300mm wide-field imaging system with a cooled camera, guiding, dithering, and autofocus.
My Rokinon 135mm f/2 with the ASI585MC Air and autofocus gives me an extremely fast system for huge emission nebulae and large areas of the Milky Way.
Then there’s my RASA 8.
That’s 203mm of aperture operating at f/2.
If what I needed was a faster way to collect deep-sky photons, buying another 50mm smart telescope would be a strange solution.
And my AT115EDT gives me the longer focal length I want for smaller targets.
So the S50 Pro doesn’t really fill a meaningful optical gap in my equipment.
But it does fill a software gap.
That’s why I bought it.
ASIAIR Was One of ZWO’s Best Innovations
Before I criticize ZWO, it’s worth explaining why this frustrates me in the first place.
I really like ASIAIR.
There’s a reason I’ve invested in the ZWO ecosystem.
ASIAIR was transformative.
Plate solving became easy. Polar alignment became easy. Guiding, autofocus, dithering, meridian flips, imaging plans, filter changes, and sequencing could all happen inside one coherent interface.
Instead of assembling a miniature IT department every time I wanted to photograph a nebula, I could turn on the ASIAIR and start imaging.
That was real innovation.
And ASIAIR remains extremely good at what it does.
The problem is that I feel like the definition of what it does hasn’t expanded enough.
ASIAIR has become an excellent astrophotography equipment controller and acquisition platform.
Seestar, meanwhile, is increasingly becoming a computational astrophotography platform.
Those aren’t quite the same thing. I will explain and Compare Them Fairly
I don’t want to pretend Seestar is somehow more capable than ASIAIR overall. It isn’t.
ASIAIR controls sophisticated astrophotography equipment that a Seestar can’t. And I don’t think a feature should be considered superior merely because Seestar hides it from the user or performs it automatically.
Automation and abstraction are not automatically innovation.
So here’s the comparison I think actually matters:
CapabilityASIAIRSeestar / S50 ProAssessmentPlate solve & centerYesYesEquivalentAutomated target acquisitionYesYesEquivalentAutofocusYesYesEquivalent core function; ASIAIR more configurableAutomated tracking workflowYesYesEquivalent core automationBasic live stackingYesYesEquivalentMosaic planningYesYesEquivalent basic capabilityAutomated mosaic acquisitionYesYesEquivalent basic capabilityAdvanced sequencingYesLimitedASIAIR advantageConfigurable external guidingYesNoASIAIR advantageCooled-camera controlYesNoASIAIR advantageFilter-wheel automationYesNo external EFWASIAIR advantageExternal rotator controlYesNoASIAIR advantageModular hardware ecosystemYesNoASIAIR advantageLive mosaic constructionNoYesSeestar advantageLive/software drizzleNoYesSeestar advantageAI denoiseNoYesSeestar advantageAI image enhancementNoYesSeestar advantageAutomated star reductionNoYesSeestar advantageAutomated Milky Way workflowNoYes — S50 ProSeestar advantageAutomated Milky Way panoramaNoYes — S50 ProSeestar advantageAutomated wide-field stitchingNoYes — S50 ProSeestar advantageAutomated star-trail workflowNoYesSeestar advantageOriginal FITS for advanced processingYesAvailable in supported workflowsASIAIR more flexibleSession-aware automated calibration workflowNoN/AASIAIR opportunity, not a Seestar advantage
That table tells a much more interesting story than simply saying one platform is better.
ASIAIR clearly wins at controlling serious astrophotography equipment.
But once the photons start arriving, Seestar increasingly does more with them.
And that’s the part I find frustrating. ASIAIR Stops Too Soon
Here’s where my criticism of ZWO gets stronger.
ASIAIR has become extremely good at:
Planning → Pointing → Focusing → Guiding → Dithering → Acquiring → Saving
And then it basically stops. There’s a common defense of this approach that goes something like:
“Serious astrophotographers want to process their own data.”
And you know what? That’s probably true in many cases.
I generally want to process my own data.
I use PixInsight. I want my original FITS files. I want control over calibration, integration, stretching, noise reduction, color, sharpening, star processing, and every other decision that ultimately determines what my finished image looks like.
But that argument creates a false choice.
Why does giving experienced astrophotographers control mean ASIAIR can’t also provide sophisticated automated processing?
Those things aren’t mutually exclusive.
Give me a switch. Examples…
AI Denoise: ON / OFF
Live Mosaic: ON / OFF
Drizzle: ON / OFF
Gradient Correction: ON / OFF
Star Reduction: ON / OFF
AI Enhance: ON / OFF
Computational Live Stack: ON / OFF
Keep every one of them disabled by default if that’s what ZWO believes experienced astrophotographers prefer.
I don’t care. Just give me the choice.
And this is where the argument that ASIAIR is intentionally keeping things manual for “real astrophotographers” starts to feel less like a product philosophy and more like an excuse for ZWO not doing the work.
From the outside, it starts to look like laziness.
Not because implementing good computational astrophotography is trivial. It isn’t.
But because ZWO has already done a substantial amount of that work.
It’s in Seestar.
Seestar Already Proves These Features Can Be Optional
This is what makes the situation particularly frustrating.
ZWO’s own Seestar software undermines the argument that advanced automation somehow has to replace user control.
Seestar already allows computational features to be used selectively depending on the workflow.
ZWO has already embraced the concept of letting the user decide how much automation they want.
So why couldn’t ASIAIR work the same way?
One night I could run ASIAIR exactly as I do today:
Capture my data. Save my FITS. Leave everything else alone.
Another night I could enable a sophisticated live stack because I’m showing friends what the telescope is doing.
Another night I could enable live mosaic construction because I’m shooting a six-panel Cygnus project.
Maybe I enable AI denoise because I want a cleaner preview on my tablet.
Maybe I don’t.
Maybe I enable software drizzle.
Maybe I don’t.
Maybe I want automatic gradient correction applied only to the live preview while preserving completely untouched subs underneath it.
Again:
Give me a switch.
That isn’t taking control away from experienced astrophotographers.
That’s giving us more control.
“Real Astrophotographers Do It Themselves” Isn’t a Good Enough Answer
There’s also something slightly absurd about drawing the line at processing.
Astrophotographers used to manually find targets. Now ASIAIR plate solves.
We used to manually focus. Now ASIAIR autofocuses.
We used to manually guide. Now ASIAIR guides.
We used to manually calculate framing. Now ASIAIR does it.
We used to manually perform meridian flips. Now ASIAIR can handle them.
We used to manually calculate mosaic panels. Now ASIAIR does that too.
Nobody argues that using autofocus means you’re not a real astrophotographer.
Nobody says: “A serious astrophotographer should manually guide for six hours.”
That would be ridiculous.
Automation didn’t destroy astrophotography.
It removed tedious work and allowed us to spend more time making better images.
Computational photography is simply the next logical layer of that automation.
And importantly, none of it has to touch the raw data.
Mosaic Is the Most Obvious Example
This is probably my biggest frustration.
ASIAIR can plan a mosaic.
It knows my telescope. It knows my camera. It calculates my field of view. I can select the number of panels and overlap. ASIAIR can slew to those panels, plate solve, center them, and collect the data.
That’s useful automation. But the workflow essentially stops at acquisition.
Now compare that with the direction ZWO has taken Seestar.
Seestar can construct mosaics as part of the imaging experience.
Imagine pointing my Rokinon 135mm f/2 and ASI585MC Air at Cygnus and creating a large mosaic.
ASIAIR could continue doing everything it currently does.
Save every FITS file. Guide. Dither. Autofocus. Maintain the imaging plan.
But simultaneously, why couldn’t it build a computational preview?
As each exposure arrives, register it. Add it to the appropriate panel. Reject bad frames. Stack the good ones. Align the panels. Update the mosaic.
By midnight I could look at my tablet and see the North America Nebula, Pelican, Sadr, Crescent, and surrounding structures gradually appearing as one enormous image.
By sunrise I’d have a beautiful preview of the entire project.
And I’d still have every untouched FITS file waiting for me to process properly in PixInsight.
Why can’t I have both?
ZWO has already demonstrated that it understands the value of live mosaic construction.
It just hasn’t brought that experience to ASIAIR.
Live Stacking Feels Increasingly Dated
The same thing is happening with live stacking.
ASIAIR has live stacking. So does Seestar. At the most basic level, that’s an equivalent capability.
The difference is what ZWO has started building around Seestar’s stack.
Seestar has been getting computational features such as software drizzle, AI denoise, AI enhancement, and automated star reduction.
ASIAIR’s live stack remains comparatively basic.
That’s difficult for me to understand.
Where is optional modern noise reduction?
Where is gradient correction?
Where is drizzle?
Where is a sophisticated adaptive stretch?
Where is intelligent frame rejection?
Where is a live mosaic?
Again, I’m not asking ASIAIR to replace PixInsight.
I don’t want it touching my original FITS files.
But this is 2026.
Computational photography has completely changed what cameras can do. A smartphone can combine a series of mediocre exposures and use sophisticated processing to create an image that would have seemed impossible from that tiny sensor not very long ago.
Astrophotography is arguably an even better candidate for computational imaging because we already capture dozens or hundreds of exposures.
The data is sitting right there.
And ZWO obviously understands this.
I know they understand it because they’re already doing it in Seestar.
Flats Are Another Missed Opportunity
I also think ASIAIR could be considerably smarter about calibration.
To be clear, this isn’t an ASIAIR-versus-Seestar comparison.
A fixed smart telescope and a modular astrophotography system have completely different calibration requirements. Hiding calibration from the Seestar user doesn’t somehow make Seestar superior.
ASIAIR already gives us tools for acquiring flats.
My complaint is about how much further the workflow could go.
ASIAIR knows what I just did.
It knows which camera I used.
It knows my gain and binning.
It knows which filters I used.
It knows the imaging plan.
It knows the focuser configuration.
And if I’m using a rotator, it knows the camera orientation.
So imagine finishing an imaging session and pressing:
“Create the calibration frames required for this session.”
ASIAIR examines the completed plan.
It determines which flats are required.
It restores the relevant configurations.
It calculates the appropriate exposures.
It acquires the necessary calibration frames.
The individual pieces largely exist.
What’s missing is the intelligence that connects them.
That’s the kind of automation I expected ASIAIR to be moving toward.
This Feels Like a ZWO Problem, Not a Technical Philosophy
If Seestar didn’t exist, I’d be more forgiving.
Maybe sophisticated live mosaics are too difficult.
Maybe computational drizzle is too demanding.
Maybe AI denoise isn’t practical.
Maybe automated image enhancement simply isn’t something ZWO wants to develop.
Except Seestar exists.
And ZWO is doing those things there.
That’s why I’m frustrated.
I don’t think the problem is that ZWO hasn’t thought about computational astrophotography.
Clearly they have.
I don’t think the problem is that ZWO believes users shouldn’t have selectable automation.
Seestar itself demonstrates otherwise.
And I don’t think preserving a traditional astrophotography workflow prevents these features from existing.
Every computational feature could operate on a temporary preview stack while the original FITS files remain completely untouched.
Maybe there are hardware limitations.
Maybe there are development-resource limitations.
Maybe ZWO simply believes the return on investment is higher with Seestar.
I don’t know what the internal reason is, and I’m not going to pretend that I do.
But from the customer side, after watching ASIAIR’s computational capabilities move relatively slowly while Seestar accumulates increasingly sophisticated imaging features, it starts to feel like ZWO has become complacent with ASIAIR.
And yes, sometimes it feels like laziness.
Not because ZWO’s engineers aren’t working.
But because ASIAIR has become successful enough that ZWO doesn’t appear sufficiently motivated to push this part of the product forward.
That’s what I’m upset about.
ASIAIR became successful because it eliminated unnecessary work from astrophotography.
I’d like ZWO to remember that philosophy.
Don’t tell experienced astrophotographers what they supposedly want.
Give us the tools.
Give us the switches.
And let us decide for ourselves.
Then ZWO Announced the S50 Pro
And this is where ZWO got my money again.
The S50 Pro isn’t simply an updated version of the original S50.
It adds an entirely separate wide-field imaging system.
And ZWO has built software around that camera specifically for something I don’t currently have:
automated Milky Way photography.
That caught my attention immediately.
Not because I can’t photograph the Milky Way.
I absolutely can.
In fact, I already have equipment capable of producing substantially better raw data.
A good full-frame camera with a quality wide-angle lens will provide much better individual frames.
My Rokinon 135mm f/2 can produce enormously more detailed images of Milky Way structures.
Put a full-frame cooled astronomy camera behind that 135mm and I could create some ridiculous high-resolution panoramas.
But there’s one important difference:
I’d have to create them.
The S50 Pro is offering something closer to:
Put it outside → select Milky Way → capture → stack → stitch → enhance → finished panorama.
That is interesting.
And more importantly, I don't own anything else that automates that entire process.
I Bought the S50 Pro for Software, and yes I paid $900 for my S50 Pro.
I’m not expecting its 50mm telescope to outperform even my my AT60EDP. It won’t.
I’m not expecting it to compete with my RASA 8 for collecting photons. That’s not even remotely fair.
I’m not expecting the S50 Pro’s wide camera to outperform a good full-frame camera. It won’t.
And I don’t expect it to replace my Rokinon 135mm system.
What I’m buying is automation.
My RASA 8 can spend an entire night collecting serious data on a nebula.
My AT115EDT can spend ten hours on a smaller target.
My AT60EDP can work on a wide-field object.
My Rokinon 135mm can collect a carefully planned, high-quality wide-field project.
And while all of that is happening, I can put the S50 Pro outside and say:
Go make me a Milky Way panorama.
That’s useful and fun to me. Not because the Seestar has better hardware.
Because it has software that does something none of my better hardware currently automates.
And That’s Why I’m Still Upset With ZWO
This is the irony.
The S50 Pro demonstrates that ZWO understands exactly where astrophotography software could go next.
They understand live mosaics.
They understand computational stacking.
They understand drizzle.
They understand automated denoising and enhancement.
They understand automated panoramas.
They understand that software can take responsibility for more than simply operating the equipment.
It can help construct the image.
Yet somehow, those ideas are showing up most aggressively in ZWO’s self-contained smart telescopes rather than the ASIAIR platform controlling far more capable astrophotography systems.
That feels backwards.
Imagine what these algorithms could do with data from a cooled full-frame ZWO camera.
Imagine live mosaics coming from a RASA.
Imagine computational stacking from ten hours of properly guided and dithered data.
Imagine a sophisticated live preview built from narrowband filters.
Imagine ASIAIR automatically constructing a giant Rokinon mosaic while simultaneously preserving every individual FITS exposure for PixInsight.
The better the hardware gets, the more useful these computational tools become. Not less.
I’m Not Asking ASIAIR to Become Seestar
This is important.
I don’t want ASIAIR dumbed down.
I don’t want it hiding advanced controls.
I don’t want ZWO deciding how my final image should look.
And I absolutely don’t want AI permanently altering my raw data.
Give me the FITS.
Every exposure. Untouched.
But give me an optional computational layer on top.
Call it Advanced Live Stack.
Call it Computational Preview.
Call it whatever you want. And you are greedy holding out for more cash, charge me a 50$ license fee for it,
Let me turn on live mosaic construction.
Let me turn on drizzle.
Let me turn on gradient correction.
Let me turn on intelligent frame rejection.
Let me turn on modern denoise.
Let me turn on star reduction.
Let me turn on adaptive stretching.
Maybe eventually let me live-combine SHO data while I’m acquiring it. Dang… That would be awesome.
And if I don’t want any of that?
I turn it off.
That's the part of the argument I find so difficult to dismiss.
Optional automation doesn't reduce control.
Optional automation increases control.
Tomorrow morning I’ll still open PixInsight.
I’ll still process the original data myself.
But tonight, while my telescope is actually collecting those photons, I’d like my advanced astrophotography controller to do something more interesting with them.
I hope ZWO Hasn’t Abandoned ASIAIR
I don’t think ZWO has stopped developing ASIAIR. That wouldn’t be a fair accusation.
ASIAIR continues to receive improvements in remote operation, hardware integration, device management, networking, and other areas. But lets be honest its really slow and less innovative.
Those things still matter. But much of that development feels like infrastructure.
Seestar development feels more experimental and forward thinking.
More visible. More focused on the actual imaging experience.
It feels like ZWO is using Seestar to explore what happens when acquisition, stacking, processing, and computational photography become one integrated system. And I think that’s exciting.
I just want some of that excitement brought back to ASIAIR.
Because right now there’s an increasingly strange divide inside ZWO’s own ecosystem:
Seestar has less capable hardware but increasingly sophisticated computational photography.
ASIAIR controls dramatically better hardware but largely expects the photographer to take over once the photons have been collected.
I’d like to see those worlds meet.
So Yes, I Bought One I’m genuinely looking forward to the S50 Pro, not so much for the hardware though.
The original S50 didn’t earn a permanent place in my equipment collection. Maybe this one won’t either. I’ll find out.
But this time I’m not really buying it because I want another 50mm telescope.
I’m buying it because I want to experience where ZWO seems to think automated astrophotography is going.
And there’s some wonderful irony in that.
I own an AT115EDT.
I own an AT60EDP.
I own a Celestron RASA 8 at f/2.
I own a Rokinon 135mm f/2 with autofocus.
I own a cooled ASI585MC Air and ASI2600mc duo.
I have equipment capable of producing considerably better astrophotography than a Seestar.
And yet I just gave ZWO another $900 because its smart telescope can automate an imaging experience that my much more sophisticated ASIAIR-controlled equipment still can’t.
So yes, ZWO:
You got another $900 out of me.
I’m excited about the S50 Pro.
I think some of what you’re doing with Seestar is genuinely innovative.
And I’m still upset with you. Because the S50 Pro makes something very clear:
You already have the technology.
You already understand the value of optional automation.
Now please bring more of it to ASIAIR. Id love to hear others thoughts.
