119 replies to this topic

[jonnybravo0311]

What are you calibrating your flat frames with? When you're taking the long exposure flats, you need to be matching them with darks. So, 10" flats need to be calibrated with 10" darks at the same gain/bin/temp.

[MarMax]

What are you calibrating your flat frames with? When you're taking the long exposure flats, you need to be matching them with darks. So, 10" flats need to be calibrated with 10" darks at the same gain/bin/temp.

I'm using SharpCap and its built in tools for calibration frames. With flats you have three choices, flat, flat-dark and flat-bias. I'm using the flat-dark for the 294MM and 294MC sensors. I'm basically useless when it comes to the technical discussion about flats and calibration and assume Dr. Robin Glover has done all that for me so I can just click the GO button.

Edited by MarMax, 28 January 2023 - 08:06 PM.

[jonnybravo0311]

You're definitely on the right path. The 294 series is painfully finicky. I have spent many days and nights of my life doing nothing but taking flats with it. Where I've ended up is close to what you're doing. I use a high CRI video light projected against my garage wall (which is painted a flat white). The light is kept quite dim so I can take the long exposure flats. I gain/temp/bin/offset match everything. If my lights are taken at gain 200, -10C, bin2... then my flats, darks and flatDarks are as well.

 

The reason I asked you how you're calibrating your flats is because of the residual amp glow you highlighted. Amp glow is removed by subtracting the darks, which must be gain/offset/temp matched to the light it is calibrating. I thought maybe you were getting some over-/under-correcting of the flats because you might be using a bias to calibrate the flat. Since you confirmed you're using the flat-dark option, then the flat over-/under-correction doesn't seem to be the problem. You sure you used the proper dark (again, gain, offset, temp, bin matched) to calibrate the lights?

[MarMax]

Thanks for the follow up jonny and I'm thinking at this point it's best to ditch all my darks and start over. I'm going to revise my EAA imaging approach for all cameras. Mainly to simplify the number of gains and exposures. Here is what I'm planning to do. The asterisk's mean I need to confirm the setting.

 

Calibration frames for exposures of 8, 15, 30, and 60 seconds.

 

QHY183C, Bin1, WB R=127, G=127, B=127
Gain| Offset
  11 | 10
  18 | 10

 

ASI533MC, Bin1 and Bin2, WB R=50, B=50
Gain| Offset
200 |  *
350 | 20

 

ASI2600MC, Bin1, WB R=50, B=50 (need to test Bin2)
Gain| Offset
200 |  *
350 | 40

 

ASI294MC, Bin1, WB R=50, B=50
Gain| Offset
200 | 10
350 | 10

 

ASI294MM, Unlocked Bin1
Gain| Offset
200 | 15

 

ASI294MM, Bin2
Gain| Offset
200 | 10
350 | 10

 

ASI290MM, Bin1
Gain| Offset
200 | 30
300 | 50

 

Player One Uranus-C, Bin1, WB R=0, G=0, B=0
Gain| Offset
200 | 10
350 | 10

 

The 294MC may still be a bit of a wildcard since it's non-cooled and there is no way to match the darks with the flat-darks. But it's correcting reasonably well so far. Using four different exposures and two gains will really simplify my dark library. It's not really necessary for all the sensors but it can't hurt to standardize them all.

 

With clouds and rain back again it's a good time to run through all the cameras and put together a fresh dark library. Getting a dimmable LED light set up for the white marker board is also a priority to be able to adjust the ADU for flats at the set exposures.

[Clouzot]

 Bin 1 on the 294MM doesn't look like it has HCG mode, but bin 2 does.

For the sake of fastidiousness, I shall mention that HGC mode is always on when using the "unlocked"/47Mpix/bin 1 mode of the ASI294MM

(hence the relatively "low" level of noise across all possible gains. Contrary to what is commonly used in bin2/11Mpix, gain 0 seems the best choice here as the fullwell stays almost reasonable, around 14000ADU)

[steveincolo]

For the sake of fastidiousness, I shall mention that HGC mode is always on when using the "unlocked"/47Mpix/bin 1 mode of the ASI294MM

(hence the relatively "low" level of noise across all possible gains. Contrary to what is commonly used in bin2/11Mpix, gain 0 seems the best choice here as the fullwell stays almost reasonable, around 14000ADU)

Ah, that makes sense!

[dcweaver]

For the sake of fastidiousness, I shall mention that HGC mode is always on when using the "unlocked"/47Mpix/bin 1 mode of the ASI294MM

(hence the relatively "low" level of noise across all possible gains. Contrary to what is commonly used in bin2/11Mpix, gain 0 seems the best choice here as the fullwell stays almost reasonable, around 14000ADU)

Interesting. Does the 120 < gain < 200 guidance still apply since there is no switching between LGC and HGC in unlocked bin 1 mode.

[MarMax]

Interesting. Does the 120 < gain < 200 guidance still apply since there is no switching between LGC and HGC in unlocked bin 1 mode.

I've been researching and asking this question and about all I can make of the responses is to avoid 119 < gain < 200 regardless of the mode (Bin1 or Bin2). Most of the information out there regarding the 294MM is based on Bin2 (or Bin1 - 11 Mpix for SharpCap).

 

EDITED to say that 120 should be avoided and 200 and above is OK.

Edited by MarMax, 30 January 2023 - 10:10 PM.

[Clouzot]

Interesting. Does the 120 < gain < 200 guidance still apply since there is no switching between LGC and HGC in unlocked bin 1 mode.

I don't know, because Sony never published the detailed spec sheet of the IMX492 where they supposedly advise against some gains.

 

In the spec sheet from the IMX294 color, Sony clearly state that when HCG is switched on, analog gain should not be between 0 and 8dB.

 

However, at gain 120, probably to keep some form of continuity with gain 119 where HCG is off, ZWO switches back the 294MC to a 0dB analog gain. Hence the recommendation to avoid gains 120-170 with that 294MC camera. (which doesn't mean it can't be used at gain 120: everyone does it).

 

Altair has a better implementation on their 294c, where HCG is switched on only above gain 900, just outside of the forbidden zone.

 

The issue with the 294MM bin1 is we don't know how gain 0 translates into analog gain terms. Maybe ZWO applied a 8dB minimum analog gain...or maybe they didn't. One would have to check if all pixels can reach full saturation or not (the latter case being the sign of the non linearity issue)

[MarMax]

Even after Dr. Robin Glover simplified what the Sony datasheet said I still could not follow it. I made this graph just so I could visualize what was happening and I still don't get it.

 

 

I bought the 294MM to use it in Unlocked Bin 1 mode but after all the discussions here I don't think that will be the case. I will use Bin2 mode much more often and it has me wondering why I bought it in the first place. I should have picked up a 533MM or 2600MM.

[Bob Campbell]

Even after Dr. Robin Glover simplified what the Sony datasheet said I still could not follow it. I made this graph just so I could visualize what was happening and I still don't get it.

 

 

 

I bought the 294MM to use it in Unlocked Bin 1 mode but after all the discussions here I don't think that will be the case. I will use Bin2 mode much more often and it has me wondering why I bought it in the first place. I should have picked up a 533MM or 2600MM.

It is confusing, for sure.

 

The way I interpret it is that unlocked bin 1 (46MP) operates in low noise mode all the way to 0 gain, whereas the 11MP bin2 mode has the transition at 120 gain.

 

Not sure why that should be true by just decoupling the adjacent pixel from each other (ie 4 pixel unit versus 4 separate pixels), unless the combining circuitry is the source of the noise.

 

FWIW

 

 

Bob

 

 

 

Bob

[dcweaver]

I bought the 294MM to use it in Unlocked Bin 1 mode but after all the discussions here I don't think that will be the case. I will use Bin2 mode much more often and it has me wondering why I bought it in the first place. I should have picked up a 533MM or 2600MM.

The 294MM was still a good choice. The 4/3 sensor is a great size and the sensitivity in red wavelengths looks a little better than the 533MM with comparable read noise (although it's hard to decipher Zwo's plots). Just a matter of learning the 294MM's quirks and using it effectively. With a couple of tests in the "dead zone" using bin1, you should have what you need to forge ahead.

[Clouzot]

The 294MM was still a good choice. The 4/3 sensor is a great size and the sensitivity in red wavelengths looks a little better than the 533MM with comparable read noise (although it's hard to decipher Zwo's plots). Just a matter of learning the 294MM's quirks and using it effectively. With a couple of tests in the "dead zone" using bin1, you should have what you need to forge ahead.

I second that. Just try and find the sweet spot.

 

For what it's worth, I've been using the ASI294MM a couple times in bin1 (a perfect match to Hyperstar) with no particular issues. I just made sure to capture long (like, 3-5-second) flats as I do in bin2, and that was pretty much it. I used either gain 0 or gain 120 and flats calibrated out vignetting and dust donuts just fine.

[Clouzot]

@Marmax, as far as I understood from Dr Golver's explanations, the gain numbers you marked (numbers in blue, green, orange) are the overall, composite gain. It's technically composed of the analog gain and of that additional, fixed gain that HCG provides when it's switched on.

 

The real issue in bin2 is that, in order to compensate for the additional gain brought by HCG, ZWO backs off the analog gain when the gain setting reaches 120. That brings the analog gain back to the forbidden (0-8dB) zone that Sony warns against

[Bob Campbell]

Even after Dr. Robin Glover simplified what the Sony datasheet said I still could not follow it. I made this graph just so I could visualize what was happening and I still don't get it.

 

 

 

I bought the 294MM to use it in Unlocked Bin 1 mode but after all the discussions here I don't think that will be the case. I will use Bin2 mode much more often and it has me wondering why I bought it in the first place. I should have picked up a 533MM or 2600MM.

In a thread over on the sharpcap forum, Robin mentioned that he uses the 533-based cooled osc sVbony SV605CC. Seems like it works very well, and fully compatible with sharpcap.

 

Bob

[MarMax]

As an update to Post #104, I've started creating a new dark library for each of the 294 sensors. I'm also going to standardize on exposures of 8, 15, 30 and 60 seconds. With one or two gains per camera and four exposures I'll simplify things a lot.

 

Another workflow change for flats is "same-gain". With this change, all parameters will be the same for flats and lights except exposure. I'm also standardizing on 10 second exposures for flats (flat w/dark option). The last change is to use the 294MM in Bin2 (11 Mpix - 4144x2822) all the time except when (and if) I use it with the Hyperstar.

 

I used the projected light method to take flat frames for both the 294MC and 294MM. With a dimmer on the projected light I was able to get 10 second exposures for gains of 200 and 350, with the histogram at about the 50% level.

 

The first night using both cameras was on 2/5 with the Full Moon. My overall sense is that the flats provided the best correction I've seen so far with both cameras. For the 294MM it seems like the amp glow is gone. Last time out using a gain of 300 I had amp glow in all the images. This time at gain 350 there was no noticeable amp glow.

 

Here is a the Seagull Nebula with the ACL200, ASI294MM-Pro and Antlia ALP-T filter.

 

[30" / 22;], Link to Large

 

 

Post-Processed Version

 

[BrentKnight]

I think I'll be sticking with a gain around 200 regardless of the gain I set for lights.  I'm also going to stick with 5 second exposures. 

 

I think my issue is that I am having problems getting my projected light to be uniform.  The camera is very sensitive to differences in brightness (that's what it does, right?), so my thinking is to keep the exposure shorter so that it does not pick up on the uneven brightness I may have.  I will also be working on getting the light as even as I can.  I'm seeing linear gradients in my flats (no matter what the settings are).  Those gradients go away and I get a more normal looking bright round center when I'm careful about the illumination on the ceiling. 

 

Kinda getting tired of all the hassles, but I look forward to any improvements you might get with your setup (I'm always open to good suggestions).

[MarMax]

I think I'll be sticking with a gain around 200 regardless of the gain I set for lights.  I'm also going to stick with 5 second exposures. 

 

I think my issue is that I am having problems getting my projected light to be uniform.  The camera is very sensitive to differences in brightness (that's what it does, right?), so my thinking is to keep the exposure shorter so that it does not pick up on the uneven brightness I may have.  I will also be working on getting the light as even as I can.  I'm seeing linear gradients in my flats (no matter what the settings are).  Those gradients go away and I get a more normal looking bright round center when I'm careful about the illumination on the ceiling. 

 

Kinda getting tired of all the hassles, but I look forward to any improvements you might get with your setup (I'm always open to good suggestions).

I would be much easier if 200/5 will work for me as well. I'm going to stick with same-gain and 10 seconds until I see how some New Moon images look. I have a feeling the gradient in the Seagull above is Full Moon related.

 

It's too early to tell if I have any non-uniformity at the target. I can't see any but the sensors will if it's there. I'm using a 100W equivalent Reveal (daylight) LED bulb that projects against a white vinyl window shade that's across the garage (24 feet) from the white marker board panel. The path in-between is clear and I'm not seeing any reflections at the target.

 

Setting up and shooting flats at the marker board is actually easier than using the LED light panel. And I'm finally getting used to the whole flats process so it's less irritating. Heck, I never did throw the the 294MM across the yard, so this hobby must have a calming effect.

 

EDITED to add that the gradient can be removed in post but that's outside the purpose of this forum.

Edited by MarMax, 08 February 2023 - 10:47 AM.

[BrentKnight]

Do you mind posting a pic of your flat capture configuration?  I'm interested in where you place the light source in relation to the telescope and the screen.

[MarMax]

Here you go, the first is looking right and left from the center of the garage at the light source and target. The second is a wide shot of the garage which is 24 feet wide. The light source is just one of those work lights with the large aluminum reflector and a regular size Reveal (daylight) bulb and this dimmer. The white marker board is from Lowe's. The marker board has a shiny finish which I figured is needed since the light source is reflected (diffused). I would think if you use direct light you would want a flat white target.

 

Left side (white board) - Right side (light source)

 

 

Wide view

 

 

EDITED to say that there are shadows/reflections on the white board because I have the garage lights on and it's daytime. When I run it at night with all the other lights off the white board (target) appears evenly illuminated with no shadows or reflections.

Edited by MarMax, 08 February 2023 - 05:51 PM.