can't seem to take proper flats with asi 294mm - Forums

hello,
so recently I made a big upgrade and bought a asi 294mm pro with chroma 1,25" filters (LRGBSiiHaOiii), and also I control everything from the eagle 5 s. this is my first time imaging with a mono camera, and also been using Nina for a few days only, I watched the video from Dark Rangers Inc about how to connect everything and how to image, and also how to take calibration frames. now ive made first light on the Wizard nebula, planning to do an SHO image, and everything went well during the aquisition of the lights. the day after I wanted to take the caibration frames, right now im still using the tablet with t-shirt method, but I have a deepskydad flat panel for my pleiades 68 refractor on the way. now the problem is that the flats seem a bit off. when I took them I thought that maybe that's how mono narrowband flats look like, but when I went and stacked everything I knew something was up. im gonna post some pictures of the stacked frames, and the flats that I took using flat wizzard.
first the stacked Ha and the master flat:

The Oiii:

and the Sii:

these are the unstretched master flats.
if y'all need something else to figure out the problem I can retake them, with no problem. first I thought the filter wasn't all the way turned so I unscrewed the camera with the filter wheel from the telescope, and it was dead center so I don't know what the problem could be.

I also had issues with flats and calibration when I first started using my 294MMPro. Here were my initial issues:

the flat frames and the master flats look "weird". They are not uniform and show "blotchy" patterns
I was using the 294 in bin 1 at f/5. With the narrowband filters the signal level was quite low and my flat frames and dark frames ended up swamping out the actual signal and I got patterns on my calibrated light frames similar to what you are seeing
During pre-processing (stacking) with Astro Pixel Processor I would get error messages that my light frames were under-exposed

Here is what ended up working for me:

When using narrowband filters I boost the gain to 240; this provides enough signal to provide good calibrated light frames
I use bin 1 mode sparingly - only when seeing and my desired resolution justify it. Obviously bin 2 will capture more light

The flat frames still look "weird" but they calibrate out in the process and I get a very nice signal to noise ratio

Mark Savan:
I also had issues with flats and calibration when I first started using my 294MMPro. Here were my initial issues:
the flat frames and the master flats look "weird". They are not uniform and show "blotchy" patterns
I was using the 294 in bin 1 at f/5. With the narrowband filters the signal level was quite low and my flat frames and dark frames ended up swamping out the actual signal and I got patterns on my calibrated light frames similar to what you are seeing
During pre-processing (stacking) with Astro Pixel Processor I would get error messages that my light frames were under-exposed

Here is what ended up working for me:
When using narrowband filters I boost the gain to 240; this provides enough signal to provide good calibrated light frames
I use bin 1 mode sparingly - only when seeing and my desired resolution justify it. Obviously bin 2 will capture more light

The flat frames still look "weird" but they calibrate out in the process and I get a very nice signal to noise ratio

i saw a video a while back of someone who made a in depth review of the 294mm and they suggested that for narrowband you should use ha bin 1 gain 100 offset 10 and for oiii and sii gain 120 offset 10. So i just naturally went with that because i didnt know what else to use

Hi Claudio,

Everyone always recommends to not evaluate stretched flats, but I agree those look weird. I am not use to seeing that sharp of light fall-off.

I have a 294MM, and have found the most success with 5s-10s flats taken with NINA Flat wizard, matched with equivalent darks

How long were your exposures?
I also switched from iPad/T-shirt to an actual flat panel (Pegasus) and that seemed to help

I am switching to Chroma, and although Chroma says orientation doesn't matter, there is a big thread on Cloudy Nights that seems to say it does

But, I don't think you can switch your filter orientation since they are 1.25" mounted?

Here is one of my stretched Sii flats (ZWO 7nm/107PHQ) as a reference, although I'm not sure how different yours should look given the faster scope and Chroma filters:

-Chris

I use ASIAir so I don't have control over offset … 240 gain worked for me … and I use bin 1 only when it makes sense

There is no need to use very high gains with the 294MM. Doing so unnecessarily sacrifices dynamic range.

I use Gain 120 and 2x2 binning. The key to good flats with the 294MM is to:

Meet or exceed 4s exposure for your flats achieving ADU of 20,000 approximately.
Calibrate the flats with flat darks (not biases!) that match the exposure time, gain, and temperature of your flats exactly

You can multiple images I have taken with the 294MM in narrow band using this method and it works fine.

Hi Claudio,

Everyone always recommends to not evaluate stretched flats, but I agree those look weird. I am not use to seeing that sharp of light fall-off.
I have a 294MM, and have found the most success with 5s-10s flats taken with NINA Flat wizard, matched with equivalent darks
How long were your exposures?
I also switched from iPad/T-shirt to an actual flat panel (Pegasus) and that seemed to help

I am switching to Chroma, and although Chroma says orientation doesn't matter, there is a big thread on Cloudy Nights that seems to say it does
But, I don't think you can switch your filter orientation since they are 1.25" mounted?

Here is one of my stretched Sii flats (ZWO 7nm/107PHQ) as a reference, although I'm not sure how different yours should look given the faster scope and Chroma filters:

-Chris

thank you for your reply, my exposure time on the oiii was indeed short, around 2 seconds, but the sii filter was 6.25s and still was not correcting right. ive taken new flats for the oiii with corresponding dark flats, they are around 8 seconds but I can already tell that they are looking the same as the other ones. @Arun H said that I should aim for a 20k ADU which I also changed and still looks the same. now looks can be deceiving so ill run the stacking again for the oiii, but I don't have high hopes. this is what my settings are in flatwizzard:

the image shown is the last flat taken before I told it to run dark flats.
also this is the first time I see such "cut off" on the bottom of flats, but there is no obstruction in the imaging train, the filters are dead center aligned with the sensor, so I couldn't explain it why there is such a cut off, also in the Sii it looks like this:

still "cut off".

Arun H:
There is no need to use very high gains with the 294MM. Doing so unnecessarily sacrifices dynamic range.

I use Gain 120 and 2x2 binning. The key to good flats with the 294MM is to:
Meet or exceed 4s exposure for your flats achieving ADU of 20,000 approximately.
Calibrate the flats with flat darks (not biases!) that match the exposure time, gain, and temperature of your flats exactly

You can multiple images I have taken with the 294MM in narrow band using this method and it works fine.

all of the above that you mentioned with the exceeding 4 s or more, and calibrating the flats with dark flats that match exposure time, gain, and temperature, I did all of that and it didn't work.

Looking more closely at your flats - you may have significant non uniformities in the flat panel you are using which is exacerbating the problem. Your SII flat with the strong light coming from the bottom left is not normal. The band at the bottom of the OIII flat is not normal either and you can see that this is impacting your light correction. The deep sky dad panel has been well characterized and, so long as you are following good practice like assuring the panel is fully covering the lens, no light leaks, good baffling, etc., should work well.

It is quite normal for the 294MM flats to have patterns such as what @apennine104 has shown. This is the flat recording the non uniform pixel response of the 294MM sensor. This pattern is present in your lights too, but not visually obvious due to noise and overlapping signal, but still important to correct to get a good image.

Arun H:
Looking more closely at your flats - you may have significant non uniformities in the flat panel you are using which is exacerbating the problem. Your SII flat with the strong light coming from the bottom left is not normal. The band at the bottom of the OIII flat is not normal either and you can see that this is impacting your light correction. The deep sky dad panel has been well characterized and, so long as you are following good practice like assuring the panel is fully covering the lens, no light leaks, good baffling, etc., should work well.

It is quite normal for the 294MM flats to have patterns such as what @apennine104 has shown. This is the flat recording the non uniform pixel response of the 294MM sensor. This pattern is present in your lights too, but not visually obvious due to noise and overlapping signal, but still important to correct to get a good image.

I guess I'll have to wait for the flat panel to arrive then.

Beside the fact that your flats look like there’s either a light leak or a bad flat panel…

do you use an oag? If so it could be protruding into your light path.

Sean Mc:
Beside the fact that your flats look like there’s either a light leak or a bad flat panel…

do you use an oag? If so it could be protruding into your light path.

there is no light leak, nor a flat panel, they are taken with iPad and t-shirt, and no im not using a OAG, there is no obstruction from the sensor till the last lens. I also took flats with the L filter and there is no "cut-off" parts that im seeing with the narrowband filters. I will have to see how it is with the flat panel from deepskydad, it should be arriving in a week or two.
and to be sure I removed the filterwheel with camera and hooked them up to my pc to check if they are positioning themselves right infront of the sensor, and they do that.

Well that’s strange.

The straight “shadow” at the bottom of the flats goes away when you take lum flats?!?

maybe play around stretching the lum flat to see if it’s there but more subtle. It doesn’t make sense that it would go away like that.

I have had the same weird band along one side as well. I thought it was my OAG, but I moved it entirely out of the light path, and it still occurred. This sensor is very challenging to get proper flats; I've seen different patterns in the same run of recording flats. For example, if I take 15 flats, some will have quite different patterns, which they shouldn't. I've used the T-shirt, iPad, and a proper light panel; however, even at its lowest setting, it was too bright to give me 4+ seconds of exposure, so I used multiple layers of T-shirts. I even resorted to using neutral-density film on the light panel, but I had no success getting consistent results.
Finally, what worked for me was the old-fashioned method of using a T-shirt, aiming the scope at a uniform part of the sky after sunset(solid cloud cover works, too), and shooting over 4 seconds.

Sean Mc:
Well that’s strange.

The straight “shadow” at the bottom of the flats goes away when you take lum flats?!?

maybe play around stretching the lum flat to see if it’s there but more subtle. It doesn’t make sense that it would go away like that.

this is the stretched lum flat.

I have had the same weird band along one side as well. I thought it was my OAG, but I moved it entirely out of the light path, and it still occurred. This sensor is very challenging to get proper flats; I've seen different patterns in the same run of recording flats. For example, if I take 15 flats, some will have quite different patterns, which they shouldn't. I've used the T-shirt, iPad, and a proper light panel; however, even at its lowest setting, it was too bright to give me 4+ seconds of exposure, so I used multiple layers of T-shirts. I even resorted to using neutral-density film on the light panel, but I had no success getting consistent results.
Finally, what worked for me was the old-fashioned method of using a T-shirt, aiming the scope at a uniform part of the sky after sunset(solid cloud cover works, too), and shooting over 4 seconds.

seems like the flat panel you used couldn't go as far down (dim) as you needed to. ill give the t-shirt sky method a try.

I bought a cheap dimmable tracing panel, and use neutral density film to give me 5-10 second flats. Seems to work ok for me now but I did struggle initially.

well, the “shadow” thing definitely isn’t there in the lum shot. Not sure how it would be there with some filters and not others. It’s the same orientation for the narrowband filters so it’s not the filter that’s the problem.

If you have lots of time to burn, you could try again with the same setup, but just use one filter and rotate the ipad between a few shots to see if the pattern changes. If it doesn’t change, it’s not the ipad. Then try rotating the camera. If the pattern doesn’t change, it’s the camera.

¯\_(ツ)_/¯

Sean Mc:
I bought a cheap dimmable tracing panel, and use neutral density film to give me 5-10 second flats. Seems to work ok for me now but I did struggle initially.

well, the “shadow” thing definitely isn’t there in the lum shot. Not sure how it would be there with some filters and not others. It’s the same orientation for the narrowband filters so it’s not the filter that’s the problem.

If you have lots of time to burn, you could try again with the same setup, but just use one filter and rotate the ipad between a few shots to see if the pattern changes. If it doesn’t change, it’s not the ipad. Then try rotating the camera. If the pattern doesn’t change, it’s the camera.

¯\_(ツ)_/¯

by rotating the camera you mean rotating it without moving the filterwheel right?

Interesting….all of you mention such low gain numbers….I was told by the company to use gain 1600 for 11MPx and 2600 for 47Mpx… can you comment on this please?

Thanks and CS

Georg

Georg N. Nyman:
Interesting....all of you mention such low gain numbers....I was told by the company to use gain 1600 for 11MPx and 2600 for 47Mpx... can you comment on this please?

Thanks and CS

Georg

on the website it recommends to use a gain of 120 because that's where the read noise is lowest. now I haven't had the chance to up the gain much because I only had one clear night with this camera.

Claudio Boicu:
Georg N. Nyman:
Interesting....all of you mention such low gain numbers....I was told by the company to use gain 1600 for 11MPx and 2600 for 47Mpx... can you comment on this please?

Thanks and CS

Georg

on the website it recommends to use a gain of 120 because that's where the read noise is lowest. now I haven't had the chance to up the gain much because I only had one clear night with this camera.

I can only show you the official diagram from QHY for their 294M and C..... 1600 is the right gain...or am I wrong?

Georg N. Nyman:
Claudio Boicu:
Georg N. Nyman:
Interesting....all of you mention such low gain numbers....I was told by the company to use gain 1600 for 11MPx and 2600 for 47Mpx... can you comment on this please?

Thanks and CS

Georg

on the website it recommends to use a gain of 120 because that's where the read noise is lowest. now I haven't had the chance to up the gain much because I only had one clear night with this camera.

I can only show you the official diagram from QHY for their 294M and C..... 1600 is the right gain...or am I wrong?

ah sorry, I wasn't clear up in my latest messages, I was talking about the zo asi 294 mm pro

Claudio Boicu:
Georg N. Nyman:
Claudio Boicu:
Georg N. Nyman:
Interesting....all of you mention such low gain numbers....I was told by the company to use gain 1600 for 11MPx and 2600 for 47Mpx... can you comment on this please?

Thanks and CS

Georg

on the website it recommends to use a gain of 120 because that's where the read noise is lowest. now I haven't had the chance to up the gain much because I only had one clear night with this camera.

I can only show you the official diagram from QHY for their 294M and C..... 1600 is the right gain...or am I wrong?

ah sorry, I wasn't clear up in my latest messages, I was talking about the zo asi 294 mm pro

Sorry, my mistake... I thought you were talking about the same camera, but from QHY...... !! I do apologize!

I think, the look of your flats is causes by something in your setup. Mine look very similar to those of @apennine104. It depends on the filter. With broadband filters they are not as bad as with nb filters. I never experienced such harsh "edges" in the frame myself.

In case of gain, I do my broadband stuff with 121 (just to be safe). With NB I usually go a bit higher to 180. As long as you do not have any really bright stars in the frame, you should not reach the full well depth. For me, it was much easier to process my images with a bit more signal in it. As many may argue, there are some downsides for going with higher gains. But in my experience, in a real world scenario I do not care much about it as long as I am not going crazy with it. The gain setting is a tool and I use it. To me it sometimes feel like buying a f/1.2 camera lens for a lot of money only to use it at f/4 all the time, because it is sharpest there. But that's just my opinion and I totally get why other people have different ones.

So I really hope you can fix your issues.

CS

Christian

Christian Großmann:
. For me, it was much easier to process my images with a bit more signal in it.

Hi Christian:

Just a couple points of clarification:

Signal has nothing to do with gain. Signal entirely depends of the photons impinging on the sensor and their conversion to recorded electrons through QE. All gain is is an amplification factor.
Between gain 120 and Gain 180, you are losing nearly an entire stop of dynamic range due to the finite bit depth of the A/D conversion while gaining nearly no benefit in noise. You can see this from the sensor graphs that George Nyman shared.

Obviously, if Gain 180 works for you, feel free to use it, but I did want to point out that the sensor data does not support a meaningful benefit versus Gain 120.