94 replies to this topic

[Wildetelescope]

GRRRR!  I have ONE leHance image with my 294 pro and never thought twice about it.  Now I have to go back and look my flats and the image, lol.   THANKS GUYS!:-).  

 

More seriously, this has been a great thread and it does explain some things.   The analysis by Steve's friend makes a GREAT deal of sense to me.  I am generally pleased with the 294MC for broadband when I use Dark Flats and I have the 571 chip if I want to do narrowband stuff.  

 

Cheers!

 

JMD

[Edarter]

A bit gutted as I've been given the green light to buy a dedicated astro camera and had long thought eh 294MC pro would be my choice. I'm now seriously doubting this!

What is the latest on this issue?

 

Thanks
Ed

[mrkhagol]

A bit gutted as I've been given the green light to buy a dedicated astro camera and had long thought eh 294MC pro would be my choice. I'm now seriously doubting this!

What is the latest on this issue?

 

Thanks
Ed

it still happens so you're lucky in that knowing it before you get it. There seems to be way of getting around this issue but it's a lot of work and pain. You'd be better off getting any other but this 294mc pro. I wish I knew before hand.

[StevenBellavia]

A bit gutted as I've been given the green light to buy a dedicated astro camera and had long thought eh 294MC pro would be my choice. I'm now seriously doubting this!

What is the latest on this issue?

 

Thanks
Ed

All cameras exhibit this effect to some degree or another (see post # 58 of this thread, where it is nearly identical on the 183MC).

This effect only happens with narrowband, and a one-shot color camera is inefficient in narrowband, even with a duo or multi-band filters.

So if you plan on doing a significant amount of narrowband work, why not get the MM version? And if you are not going to do narrowband, the 294MC should be fine.

I have the 294MM, and it is an excellent camera for the price. It has a high QE, low noise, and not much amp glow (easily calibrated out with dark frames).

And it seems others have worked around the issue with the 294MC, but I cannot comment as I do not own one (yet).

[bobzeq25]

A bit gutted as I've been given the green light to buy a dedicated astro camera and had long thought eh 294MC pro would be my choice. I'm now seriously doubting this!

What is the latest on this issue?

 

Thanks
Ed

A number of people have had a number of problems with the 294.  Two good options.

 

A 533 camera, smaller chip.  A camera with the 571 chip, more expensive.

 

Compared to the 294, both are more troublefree.

 

There are many happy 294 owners, also.  But it does appear to be something of a risk.

 

My 183s have never had an issue.  But they've been retired for the ZWO 2600s, with 571 chips.  And the small pixeled 183 are specialized for short fast scopes, I got mine for my C8 RASA.

Edited by bobzeq25, 04 May 2022 - 08:53 AM.

[StevenBellavia]

A number of people have had a number of problems with the 294.  Two good options.

 

A 533 camera, smaller chip.  A camera with the 571 chip, more expensive.

 

Compared to the 294, both are more troublefree.

 

There are many happy 294 owners, also.  But it does appear to be something of a risk.

 

My 183s have never had an issue.  But they've been retired for the ZWO 2600s, with 571 chips.  And the small pixeled 183 are specialized for short fast scopes, I got mine for my C8 RASA.

Good points!

 

Just to add another option:  

You can software bin the 183, which gets it close to the 294 in terms of resolution and SNR. (though also a smaller chip, with identical diagonal measurement as the 533. i.e., 16.0mm)  If you extract the R-G-B from the MC version, it is sort of "automatic binning". 

I only mention this as the 183 can also be an option and has a more traditional rectangular framing, versus the square 533 (I own the 183 and 533 and use them interchangeably, deepening on the framing I want.)

[Craigar]

I see the same thing in my imx294 flats using a Triad TriBand and I'm in an almost bortle 9 backyard so far they seem to calibrate out pretty well in DSS especially when including Dark Flats .. right now I  have been trying to "shoot what I can" and the best positioned of the very few nebula is M57. I've only had 3 sessions with this filter and very poor mostly cloudy conditions so far I think it's a pretty good way to get around serious LP (although no where near as good as a "gasoline filter"..)

 

Here's a master flat (1600x1600px center ROI) stretched max using GIMP's with levels>autostretch

 

 

 

and here's my first try using this master flat (unstretched)

 

M57 about 1.6hrs total integration, Bortle 8/9
ES 152mm DHL MakNewt > TriadTriBand > Ds10c(imx294)
DSS 2x drizzle average, calibrated with (t-shirt)Flats + DrkFlats + dark

 

 

Edited by Craigar, 02 June 2024 - 03:02 AM.

[fieldsweeper]

I've had a lot of problems with flats on my 294 along with a lot of other people on CN, and it sounds like this might explain what's happening.  It's encouraging that ZWO is at least getting close to acknowledging a problem with the sensor.  Staying tuned...

Well, 2024 and I see nothing from them in my searches lol.  Guess they don't really care THAT much lol.

[mrkhagol]

Well, 2024 and I see nothing from them in my searches lol.  Guess they don't really care THAT much lol.

I really had a back and forth with their support person..think for one here in US-and I just couldn't get my results to get calibrated at all. This was a year or two ago. I basically then gave up and got 2600mc pro. Now using 294 as more of a planetary cam instead which just works fine especially it's hi-res that helps a lot .

[fieldsweeper]

I really had a back and forth with their support person..think for one here in US-and I just couldn't get my results to get calibrated at all. This was a year or two ago. I basically then gave up and got 2600mc pro. Now using 294 as more of a planetary cam instead which just works fine especially it's hi-res that helps a lot .

Honestly, I would not have gone with another ASI in that case tbh, in fact I was looking at the same one, but rather switching to mono, but between things like that, the grease gate issue, asi air lock down I think I decided to go mini PC/with nina and other brand stuff if this is the norm for them lol. 

[StevenBellavia]

Well, 2024 and I see nothing from them in my searches lol.  Guess they don't really care THAT much lol.

It might not be that they don't care, but that there is not much they can do about it.

If you go back to post, #58, and #59, on page 3, you will see that other cameras do this as well, and perhaps they all do, to some extent, as silicon is somewhat transparent in certain wavelengths, so you get an etalon effect from what is behind it (in this case, the glue that holds the sensor to its support frame, and why the pattern looks essentially the same in all models).

 

Steve

[fieldsweeper]

what is the etalon effect? and you are saying you can see the actual glue behind it? lol how

Edited by fieldsweeper, 11 June 2024 - 08:45 PM.

[mrkhagol]

Honestly, I would not have gone with another ASI in that case tbh, in fact I was looking at the same one, but rather switching to mono, but between things like that, the grease gate issue, asi air lock down I think I decided to go mini PC/with nina and other brand stuff if this is the norm for them lol. 

yeah sure but so far 294 is the one with issues while other cams have been super good or just fine.

[mrkhagol]

what is the etalon effect? and you are saying you can see the **** glue behind it? lol how

if you search for etalon effecct in threads-there was one who posted good details on it. something good to know. wish i had bookmarked it.

[fieldsweeper]

well I can't seem to do it, sky flat or led panel, lol both look similar, and i tried a few different exp values while still being near the middle of histogram. The fact neither direction seems to make any change, makes it even harder to tell if maybe just a little more expl or a little less, I can't imagine that there isn't a small range where it will be perfect... then slightly out of that range not instantly like these screen shots.   Surely it has a even if small range where you will notice something. So having done several exp values, and skys and panels, I just don't know what else to try

Edited by fieldsweeper, 11 June 2024 - 10:02 PM.

[mrkhagol]

well I can't seem to do it, sky flat or led panel, lol both look similar, and i tried a few different exp values while still being near the middle of histogram. The fact neither direction seems to make any change, makes it even harder to tell if maybe just a little more expl or a little less, I can't imagine that there isn't a small range where it will be perfect... then slightly out of that range not instantly like these screen shots.   Surely it has a even if small range where you will notice something. So having done several exp values, and skys and panels, I just don't know what else to try

don't bother with sky flat. it just won't work. how is the result with another filter not triad or NB?

[fieldsweeper]

Here are one of each of the files... Organized if any of you wanna take a stab) https://drive.google...?usp=drive_link

 

 

this is also in another thread I opened and someone asked, but since others are commenting here on this one too I figured I would add this here as well:

 

There is a doc in there that states why i put more 10s ones in.   10s was the sky flat I did yesterday, but it was kind of late, so it was hard to get more than 2 seconds without jumping to a longer one, BUT since it was so long at 10s per, as it went on, it's histogram started dropping as the daylight slowly went away. (but fast enough apparently) So I included the first frame and the last and 2 in between, but only one bias.   I organized everything to be easy.

I also added a few different flats/bias.  each matching of course, also separated by sky and panel, there are at least 3 of each. 

If you are able to figure it out, that would be greatly appreciated.   Here is also a link to the settings i had in WBPP in pixinsight... (https://imgur.com/a/3xAMZxE) .  Ignore the master flats etc, as that was just there when I was done with one today (and it still didn't work) but for each of these runs, I deleted the masters and ran with the full count of the calibration frames. I also had drizzle on but that particular run I forgot to select it, but I did rerun it with it, and still no difference in the pattern.  (here is the pattern...   https://i.imgur.com/zmeGfiO.jpeg..., and it's almost always the same in all of those pics, with very little difference between all the different variations of flats etc.   I also tended to purge the cache before each run as well.

[mrkhagol]

Here are one of each of the files... Organized if any of you wanna take a stab) https://drive.google...?usp=drive_link

 

 

this is also in another thread I opened and someone asked, but since others are commenting here on this one too I figured I would add this here as well:

 

There is a doc in there that states why i put more 10s ones in.   10s was the sky flat I did yesterday, but it was kind of late, so it was hard to get more than 2 seconds without jumping to a longer one, BUT since it was so long at 10s per, as it went on, it's histogram started dropping as the daylight slowly went away. (but fast enough apparently) So I included the first frame and the last and 2 in between, but only one bias.   I organized everything to be easy.

I also added a few different flats/bias.  each matching of course, also separated by sky and panel, there are at least 3 of each. 

If you are able to figure it out, that would be greatly appreciated.   Here is also a link to the settings i had in WBPP in pixinsight... (https://imgur.com/a/3xAMZxE) .  Ignore the master flats etc, as that was just there when I was done with one today (and it still didn't work) but for each of these runs, I deleted the masters and ran with the full count of the calibration frames. I also had drizzle on but that particular run I forgot to select it, but I did rerun it with it, and still no difference in the pattern.  (here is the pattern...   https://i.imgur.com/zmeGfiO.jpeg..., and it's almost always the same in all of those pics, with very little difference between all the different variations of flats etc.   I also tended to purge the cache before each run as well.

i'll take a look when i get a chance but remembered something else from my previous usage of pixinsight (not using it currently though).

 

there's a way to do synthetic flats and 'remove' such nasty gradients. http://trappedphotons.com/blog/?p=756

 

that's one ref i tried and then came up with my own approach which kind of worked for me and was satisfied with final result.

 

you may want to try it if you haven't come across it. plus it's little fun if you enjoy doing math

 

these are the formulas I used to try do both flats and then synthetic flats.

 

rgb - rgb_background * 0.8

 

(rgb * mean(flat) / flat)
 

[sharkmelley]

Here are one of each of the files... Organized if any of you wanna take a stab) https://drive.google...?usp=drive_link

You need to be really methodical about taking calibration frames.  Use the same software (you switch between NINA and ASIAir) and the same parameters (your dark has a different offset to your lights, flats and bias).  Furthermore there is evidence of light leaks in some of your calibration frames.  Also your panel flats do not calibrate your sky flats so don't use the light panel any more.  There is no way you can easily use the set of files you already have.

 

Having said all that, I was able to perform a decent calibration of one of your lights by adjusting the offset of your dark to match the offset of the other files and by using the sky flat (not the panel flat).  Here is the resulting red channel:  

 

 

There's not the slightest hint of the "pretzel" pattern.  So I'm convinced you will have no further problems if you use sky flats and take a consistent, methodical approach to calibration frames.

[mrkhagol]

You need to be really methodical about taking calibration frames.  Use the same software (you switch between NINA and ASIAir) and the same parameters (your dark has a different offset to your lights, flats and bias).  Furthermore there is evidence of light leaks in some of your calibration frames.  Also your panel flats do not calibrate your sky flats so don't use the light panel any more.  There is no way you can easily use the set of files you already have.

 

Having said all that, I was able to perform a decent calibration of one of your lights by adjusting the offset of your dark to match the offset of the other files and by using the sky flat (not the panel flat).  Here is the resulting red channel:  

 

CN_fieldsweeper_ASI294_calibrated_red.jpg

 

There's not the slightest hint of the "pretzel" pattern.  So I'm convinced you will have no further problems if you use sky flats and take a consistent, methodical approach to calibration frames.

nice..sky flats working out.