I would appreciate a critique of my M33 image, which was awarded as IOTD a few days ago. To my eyes, the image is far from excellent, and yet it was my first IOTD. In particular, I noticed several artifacts caused by BlurXT + NoiseXT ("mottling" as Adam Block calls them in his video on the subject), especially in the periphery of the galactic disk. So much so that to get a decent result I had to lower the BXT sharpening to 0.25. Subsequently, I had some difficulty blending L with RGB, in fact I even considered not using luminance at all. In the end, I produced the LRGB with ImageBlend and a blending percentage of 0.6. I treated the HA similarly, blending only a minimal percentage. All this, with an additional difficulty: it was impossible to work on a starless image, since SXT extracted large portions of the galaxy. Therefore, the entire workflow took place with the stars present (and in fact many are bloated). Ultimately, I don't think this is my best image at all, but I'd appreciate your opinion on it.š· M33
[RCC] My very first IOTD
I donāt mind your stars at all. I often will not remove them from Galaxy images just because removing them tends to be destructive. On an object like M33 that has so many resolved stars, the last thing I want to do is minimize them.
On the rest, all I can say is Shhhhhh!
I donāt tend to do star removal for galaxies either. It tends to remove the knots and other features. Also, I donāt mind doing a greater stretch on the stars than usual since you can really bring out the super faint fuzzies in the background.
You image had a great composition rather than the normal diagonal M33, and nice framing as well. I think that was what they were looking at rather than jumping to a 100% view right away.
Iām kinda glad you came forward, because I agree with you. I donāt think it belonged there, but Iām known for holding high standards that seem silly to others. Iām also partially salty because my own didnāt make it there, but I digress.
The main thing that stands out to me from afar is the color. Itās quite blue and the red of the Ha structures do not stand out, if anything the Ha appears washed out, bordering colorless. The background is not neutral and also has a nice blue/green cast to it. The stars lack meaningful saturation as well.
You should not attempt starless processing with M33 at all. This is honestly a fault of many processed images of galaxies. M33 is one of our closest galaxies, and weāre able to get up close and personal with itās structures. All that blue is a sea of stars, very little is nebulosity. And yet, a large percent of M33 images you will find, the structure appears as nebulosity instead. Your image is not entirely there, but you were getting close in some areas.
See here:
Iām fairly confident your scope has the ability to resolve most all the stars in this field of viewā¦ yet for some reason, your star field is a few pin point stars, and alot of soft, denoised mottle and hallucinated structure. Alot of this comes from the combination and subsequent overuse of denoise with blurX. Just my 2 cents, I donāt use hardly ANY denoise on galaxy images until Iām nearing the end of the process. It ensures all those little stars do not become a blur when denoise mistakes them for noisy pixels.
Here is a clip from my dataset (processed by Charles Hagen) to demonstrate what I mean by the rich star field that exists. Denoise will easily nuke all those details. Our scopes are fairly similar in size.
The outer background structures of your image you are well aware of. I think you will figure out how to fix that on your own. The answer is simple overuse of the BX/NX suite. Try processing an image without them. Also, the āpower settingā of blurX is not the same for every person, PSF fitting could be off as well. In your image, I definitely feel both these tools were over used.
Every image is a learning opportunity. Even for those of us more experienced, we are always finding ways to improve. For those who struggle I always like the saying āLess is moreā. Respect the data. Tab between the raw image vs your final image, at the pixel level. Do you think it looks better? Or worse? These are the things I do for every step.
To my eye I think this is a great rendition of M33. I wouldnāt have pushed the Ha regions any more than you didā¦ The ol ājust cause you can doesnāt mean you shouldā applies to Ha in galaxies I think.
The color all looks nice and natural to meā¦but Iām old school with processing and will always gravitate toward less HDR and more natural looking galaxies.
Brad
Surprisingly refreshing and straightforward discussion, especially since AB is usually very sanitized. Some threads read like ChatGPT praising Claude and vice versa, to be honest.
You, sir, deserve real credit for starting this topic. Very few people I know would question their own IoTD and openly invite public discussion about it.
Now, candidly ā the image looks great, but only up to about 50% of the original resolution. Brian is right: M33 is a unique galaxy for us amateurs. The grain is real structure, but itās so fine that any denoising tool will smear it, and aggressive deconvolution ā whether the oldāschool approach or BX ā will introduce strange artifacts (i call them āwormsā).
Iām not sure whether you used BX after NX; if you did, you should deconvolve a raw, cropped image instead.
I usually do the first pass of noise reduction just before stretching, and I work with extremely noisy data (faint PNe). Even then, I often run NX with masks to moderate its influence on the faint signal or very fine-contrast details (it smears those too).
If the data is good (LUM or bright Ha), I tend to skip noise reduction entirely during the linear phase.
In general, noise reduction should be treated like a series of small, precise specialāops missions ā not a oneāpass carpet bombing.
I am certainly not qualified enough to judge if your image is IOTD worthy. I think that it is quite beautiful and very enjoyable to gaze at and into. Bottom line is that it is an IOTD and was judged to be.
I admire that you are sharing your feelings here.
I recently imaged M33 myself because I have never been happy with any previous rendition and wanted to try again with a mono camera. I really struggled with it and like many others have mentioned in this thread.
I have learned a lot from the comments here about the unique aspects of M33 and how over denoising can destroy the delicate data. I am guilty of that in my process. Along with overprocessing.
To me, M33 is just one of those targets that can be processed to create strikingly different results. When I look at your image and then my last attempt, they look like completely different galaxies to me.
I am very impressed with your image and your very sincere openness. I have learned a lot from this thread and might just be doing a re-process soon!
Jim
I agree, truly admirable to open up this type of discussion. In my observation with this site, there are more elements involved when giving recognition to images. Remember, this site drives commerce (subscriptions/equipment/advertising) so the controlling levers are always going to lean towards mechanisms that drive that growth. Sometimes you see truly quality work, that simply canāt be ignored. They are more rare than not. More often than not, I see those who have great social characteristics get rewarded for mediocre work. What drives commerce better than the social charismatic personalities? Then you have artistic expression which has gained a lot of momentum over authentic / quality work. Iām ok with all of it, however, if there were an area here, where some standards were established, and a pursuit of authentic/quality was the goal, Iād be the first to sign up for it. Sadly, that would limit the scope of a commerce based goal so I donāt think it would ever happen. None the less, as to your M33. I think itās fantastic and as many point out here, itās actually a very difficult image to process to completion. I also think you did a good with a authentic / quality process. Does it stand out to me in a IMOD way? If I were honest, no, as Iāve seen others who didnāt get IOTD that I feel is of better quality. But again, there are other factors that determine these results so as it stands, itās just everyoneās opinion based on what they rank as priority in an image. Congratulations to you, regardless of my, or anyone elseās, ultimate opinion, your out there doing the hard work and are very deserving under these conditions (Just like anyone who participates).
Brian Puhl Ā· Mar 5, 2026, 01:41 AM
You should not attempt starless processing with M33 at all. This is honestly a fault of many processed images of galaxies. M33 is one of our closest galaxies, and weāre able to get up close and personal with itās structures. All that blue is a sea of stars, very little is nebulosity. And yet, a large percent of M33 images you will find, the structure appears as nebulosity instead. Your image is not entirely there, but you were getting close in some areas.
See here:
Here is a clip from my dataset (processed by Charles Hagen) to demonstrate what I mean by the rich star field that exists. Denoise will easily nuke all those details. Our scopes are fairly similar in size.
Every image is a learning opportunity. Even for those of us more experienced, we are always finding ways to improve. For those who struggle I always like the saying āLess is moreā. Respect the data. Tab between the raw image vs your final image, at the pixel level. Do you think it looks better? Or worse? These are the things I do for every step.
Thank you Brian, for your excellent post and critique and for explaining so well what makes a great M33 image. I learned from it. āSimpleā objects like M42 and M33 are often the hardest to process well. Your image and Charles Haagenās process of it certainly deserve an IOTD in my eyes.
I would also like to say, that this topic is refreshing and I really appreciate it being opened by the OP and the CC in the answers.
First - let me congratulate you to your first IOTD. I am contributing for several years now and I am still waiting for this special notification - your image was selected IOTD. Well - I have read all previous comments and think they are correct - I need to rethink my own processing avenue, I guess.
Looking at the IOTD images daily, I see, that many, too many are awarded to images, which a ānormalā astrophotographer is never ever able to get - either 70-150 hours of integration (which in my area means about two yearsā¦) or acquired with a 100000-500000$ equipment somewhere in Bortle 1. OK, maybe fine, but I sit back and think - nice image, yes, but out of range forever.
You and your image have achieved something, which demonstrates, that one can get this IOTD with an equipment and within a time and darkness range, which encourages most others to keep doing and aiming for that applause!
Thank you for your image, I like it!
Very timely topic. I have been reprocessing M33 for the last week or so and wasnāt happy with the results. This thread provided some great insights for me to explore. It was eye opening to totally skip BXT and NXT (I was already skipping SXT).
Comparison:
At the moment, Iām leaning toward the āliteā version but Iāll sleep on it.
Thanks for the constructive thread!
ETA: congrats on the IOTD!
This is a refreshing topic, years ago I had 1 IOTD and promptly got butchered by just a few elitistist (i had Salvatorie demote to remove the arguments and abuse i got), I almost gave up this hobby but some persuaded me not to let others sway my interest. Iām pleased to say iām still going 6 years down the line and much better at processing.
Galaxies are an interesting mix, take the stars out and lose a bucket load from the image. Leave them in and your stars go wonky. Your question is a good one and everyone will work on a solution to an age old problem. Your image BTW is amazing and deserved the accolade.
I think the advice given so far is great and positive, which sadly I didnāt get. But itās super refreshing to see, good luck for future IOTD. š
Greg has hit an interesting point that alot of images of recent are images taken with equipment that is out of reach of alot of folk. While its nice to see new and exciting images, just occasionally itās nice to see images taken and processed at the lower end. It can be disheartening if your chasing this award, personally I donāt bother myself but simply enjoy the challenges of this amazing hobby. In fairness the community will give advice even to a question that you may think is silly to ask. Thereās never a dumb question to ask in this technical hobby.