Which is better for the bubble?

[MARS1960]

Tonight i am hoping to image 4-5hrs of NGC 7635.

I will be using my 100ED native and my 60D.

I am thinking either around 40 x 6min subs at ISO 800 or 80 x 3min subs at ISO 1600.

Which would suit this object best?

[Yamez]

4 minutes ago, MARS1960 said:

40 x 6min subs at ISO 800

Not an expert but this way i'm sure will have a lot less noise

[MARS1960]

4 minutes ago, Yamez said:

Not an expert but this way i'm sure will have a lot less noise

I would tend to agree but this Dynamic range, ADC and SNR is getting me really confused, would twice as many subs actually be better than half as much sensor amplification?

[Yamez]

1 minute ago, MARS1960 said:

I would tend to agree but this Dynamic range, ADC and SNR is getting me really confused, would twice as many subs actually be better than half as much sensor amplification?

I'm terribly sorry, i'm not the best person to ask i am very much still a beginner to astrophotography. More subs means you have more data to stack and with a lower ISO you are avoiding high noise levels. Just remember to take your dark frames, bias frames and flats. I'm again really sorry if this doesn't answer your question properly 

[dph1nm]

The answer is, it depends! In particular on how much sky noise you pick up in 3 mins. If the sky noise drowns out the read noise in 3 mins then there will be no difference in the final signal-to noise between the two scenarios. If it doesn't then you probably want the 6min subs (assuming you can track OK that long).

NigelM

[MARS1960]

1 minute ago, dph1nm said:

The answer is, it depends! In particular on how much sky noise you pick up in 3 mins. If the sky noise drowns out the read noise in 3 mins then there will be no difference in the final signal-to noise between the two scenarios. If it doesn't then you probably want the 6min subs (assuming you can track OK that long).

NigelM

tracking/guiding isn't a problem and i use a CLS LP clip filter so 6mins would possibly be the better alternative.

[dph1nm]

5 minutes ago, MARS1960 said:

tracking/guiding isn't a problem and i use a CLS LP clip filter so 6mins would possibly be the better alternative.

I guess so - in fact 40x6min at ISO1600 should have slightly less read noise overall - so better signal-to-noise, but the stars would suffer more saturation.

NigelM

[MARS1960]

1 minute ago, dph1nm said:

I guess so - in fact 40x6min at ISO1600 should have slightly less read noise overall - so better signal-to-noise, but the stars would suffer more saturation.

NigelM

Thanks Nigel.

Now i just have to hope there isn't bucket loads of dew like last night or it will be 120 x 2mins at ISO 3200 unguided lol.

I must get that guidescope dew heater.

[Adam J]

40 x 6 mins for sure....but I never use ISO 1600 on most cameras its a poor trade off as its more than twice the noise of ISO800.....however I have never used a 60D so who knows. The bubble is a difficult target for an non modded dslr. Discarding your options I might go for 24 x 10min ISO800...if it was me.  Exposure length will be important for you to get the signal on a non modified camera. 

[MARS1960]

13 minutes ago, Adam J said:

40 x 6 mins for sure....but I never use ISO 1600 on most cameras its a poor trade off as its more than twice the noise of ISO800.....however I have never used a 60D so who knows. The bubble is a difficult target for an non modded dslr. Discarding your options I might go for 24 x 10min ISO800...if it was me.  Exposure length will be important for you to get the signal on a non modified camera. 

I totally agree Adam it's mag 10  too, so after giving it more thought i am all set up and guiding on the wizard at a more comfortable mag 7.20, hopefully it will be more forgiving on an umodded DSLR. going to try 6mins at ISO 800 and see where the histogram sits.

[Adam J]

never tried the wizard myself so i can't comment, good luck

[dph1nm]

17 hours ago, Adam J said:

but I never use ISO 1600 on most cameras its a poor trade off as its more than twice the noise of ISO800

Not so - the noise in electrons (which is what matters) is either roughly the same, or in most (older) Canon cameras better, as the read noise is lower.

NigelM

[Knight of Clear Skies]

17 hours ago, MARS1960 said:

I totally agree Adam it's mag 10  too, so after giving it more thought i am all set up and guiding on the wizard at a more comfortable mag 7.20, hopefully it will be more forgiving on an umodded DSLR. going to try 6mins at ISO 800 and see where the histogram sits.

Probably a good idea, I seem to remember the Bubble is quite a difficult target.

One thing to bear in mind - magnitude is total brightness so can be a poor indication of how easy a target is to image, surface brightness (magnitude per unit area) is a far better indication. Although the Wizard is mag 7.2 it's fairly large its light is spread out over a wide area, it could be quite challenging with an unmodded DSLR. 

[symesie04]

Not sure i get the problem. If your saying you can track nicely for 6 mins and LP isnt an issue with the clip (in my limited experience LP can still be an issue with a clip) then why wouldnt you go for iso 1600 for 6 mins? especially for such a faint object.  Within a few subs your going to know if the stars are being bloated and if so just shoot some subs at 3 mins or is0 800 and stack them for a star mask?

[MARS1960]

I don't know, thats why i'm asking. 

There is so much contradiction on the internet when considering ISO and sub exposure length with Dynamic range, ADC and SNR.

[symesie04]

Fair point. Im sure there is a post somewhere on here where a member does a lengthy test of SNR V ISO ratios and the conclusion was that iso 400 was more ideal if you can spare the extra time. Between iso 800 and 1600 there was actually very little to choose. Sure some one can come up with the link.

[symesie04]

heres one discussion but if you do  a search there are loads and some very interesting ones as well.

 

 

[Adam J]

3 hours ago, dph1nm said:

Not so - the noise in electrons (which is what matters) is either roughly the same, or in most (older) Canon cameras better, as the read noise is lower.

NigelM

Having done the testing myself I am going to firmly disagree with you.

The only situation in which higher ISO will result in superior signal to noise ratio is when the signal is week and the exposure is short. In this instance a higher ISO will help raise the signal above the fixed read noise. Its true to say that the lower the ISO the longer it will take to raise the signal above the read noise (requiring longer exposures), however we are talking about 6 min guided exposures here not 60 second unguided exposures.  The histogram should have fully separated from the left hand side of the chart even at ISO 400 for the majority of targets. For very dim targets you will hit a wall of thermal noise problems from lack of cooling irrespective of ISO, as you would need to increase the exposure length significantly.

Once the histogram has fully separated from the read noise / left side of the chart all that you accomplish by using ISO 1600 as opposed to ISO 800 is to increase the amplifier's contribution to the total noise reducing overall signal to noise ratio, while also reducing dynamic range meaning that your stars will saturate sooner. 

Most DSLR hit unitary gain at between ISO 400 and ISO 800 (with some notable modern exceptions like the A7), although I admit that I don't know what unitary gain is for a 60D. So for most cameras by using ISO 1600 all you are doing is saturating the image faster at the expense of more amp noise while achieving nothing that cant already be achieved (with less noise) by stretching the image digitally. ISO 1600 wont gain you any additional information. 

Found the below from you on another much older thread:

"Another point worth making is that I see various comments on various astro forums about high ISO images looking too noisy, so people use lower iso. But this is not what matters - it is the signal-to-noise that concerns us, not just the noise. So a high iso image will have higher noise in terms of the numbers coming out of the camera, but the signal-to noise will just the same (or slightly better if you are read noise limited) as any other iso."

The problem is that you are not just amplifying the signal / total pre-amp noise, you are also amplifying the noise internal to the amp itself. So unless (as you note) you are read noise limited you are just increasing the component of noise derived from the amp itself and as this is not linear with ISO you are reducing signal to noise ratio. 

I would also say that if a picture only looks noisy, its got a lower signal to noise ratio. If it did not have a lower signal to noise ratio it would not look noisy, you would just be able to set the black point to remove the noise without removing the data at the same time.  

ISO selection is about balancing various noise sources to increase signal to noise, the optimum will depend on lots of different factors such as: Brightness of your target, sensor temperature, maximum guiding period, light pollution and total integration time / number of frames and others. While under some conditions higher than unitary gain ISO is justified, the fact is that under the vast majority of conditions its not and on average ISO 800 / 400 performs better than ISO 1600 in terms of signal to noise ratio.  In most conditions when higher than unitary gain is justified then you are probably setting yourself up for a poor image irrespective as you have come up against some other limiting factor.   

 

[Adam J]

1 hour ago, symesie04 said:

Fair point. Im sure there is a post somewhere on here where a member does a lengthy test of SNR V ISO ratios and the conclusion was that iso 400 was more ideal if you can spare the extra time. Between iso 800 and 1600 there was actually very little to choose. Sure some one can come up with the link.

In the end it depends on your camera and the ratio of its various noise sources, ambient temperature, total exposure length and a bunch of other things. But I will stick to my guns with this and say that ISO 1600 is less optimal for most cameras under most conditions in comparison to ISO 400 or ISO 800. 

[symesie04]

Indeed the thing is camera technology moves on rapidly and DSLR cameras over the last year or two have huge improvements in high ISO noise and SNR, not that 1600 is high iso anymore. So perhaps we should be talking about iso 1600 in terms of yesterdays iso 400-800 and yesterdays iso1600 is really more equivalent to todays iso3200-6400. maybe?

[Adam J]

55 minutes ago, symesie04 said:

Indeed the thing is camera technology moves on rapidly and DSLR cameras over the last year or two have huge improvements in high ISO noise and SNR, not that 1600 is high iso anymore. So perhaps we should be talking about iso 1600 in terms of yesterdays iso 400-800 and yesterdays iso1600 is really more equivalent to todays iso3200-6400. maybe?

Perhaps, but i would still argue that its more to do with lower noise electronics tolerating higher ISO better, which is not necessarily the same thing as higher iSO being optimal.

[MARS1960]

1 hour ago, Adam J said:

Perhaps, but i would still argue that its more to do with lower noise electronics tolerating higher ISO better, which is not necessarily the same thing as it being optimal.

Sure, that goes without saying.

I think after reading the comments the best approach may just be deciding which image, after processing, will give the best results, but, by that very nature it will depend on what is being imaged.

So i will have to ask myself, will this object respond better with an increase in PP noise reduction or PP stretching!

[ollypenrice]

The only person who can answer the OP's question decisively is the OP. There are many variables defining the 'initial conditions' which affect his captures. This game is about experimentation. If I posted a similar question, stating my kit and my sky conditions, I would get answers varying between, 'Take 300x1 minute subs,' to 'Take 10 x 30 minute subs.' But I don't ask the question because I know the answer, and I know the answer because I have done the experiments. Honestly, I think that is the only way.

(Part of the answer lies in processing. Saturated stellar cores, for instance, can be fixed in processing. The colour information lost in the core itself is still available just outside the core. This is a perfectly reliable source of colour information, possibly more reliable than colour information captured close to saturation.)

Astrophotography is a world populated by comminicative autodidacts. Embrace it!

Olly

[MARS1960]

2 hours ago, Adam J said:

Perhaps, but i would still argue that its more to do with lower noise electronics tolerating higher ISO better, which is not necessarily the same thing as it being optimal.

Sure, that goes without saying.

I think after reading the comments the best approach may just be deciding which image, after processing, will give the best results, but, by that very nature it will depend on what is being imaged.

So i will have to ask myself, will this object respond better with an increase in PP noise reduction or PP stretching!

[MARS1960]

9 minutes ago, ollypenrice said:

But I don't ask the question because I know the answer, and I know the answer because I have done the experiments. Honestly, I think that is the only way.

 

I know the answer too, i realized the answer about 10 minutes ago when i posted and it's the same answer you got .