DSLR Noise reduction

[Chefgage]

Hello all. I have just realised why my dslr takes so long to write data to the card after a long exposure. Its the noise reduction setting on the dslr. So obviously if i take an 8 second exposure it will then take another 8 seconds as it takes another 8 second 'dark' picture.

My question is does this need to be done if i am taking my own dark frames? If i am taking lots of say 30 second exposures then its going to take twice as long to take the light frames. 

[PeterCPC]

Definitely not - turn it off. Your own dark/bias frames will do the job.

Peter

[Alien 13]

As above, the only time the "long exposure noise reduction" is useful is night time landscapes or single exposure milky way shots that have foreground detail.

Alan

[smr]

as above, imagine doing 5 minute exposures and then waiting for the camera to do another 5 minute dark exposure straight after. 

Also, depending on your DSLR it may not be necessary to do darks at all, if it is one of the more recent sensors. 

[Chefgage]

I thought as much. It did seem a bit daft waiting twice as long ?

[alacant]

10 hours ago, Chefgage said:

taking my own dark frames

Hi. You'll probably find that dark frames introduce more noise, so best without them. Take bias frames instead and dither between exposures. Stacking with a clipping algorithm along with flat frames could be a better way against noise. That's certainly true with our bashed around Canon DSLRs. HTH.

[Chefgage]

12 hours ago, alacant said:

Hi. You'll probably find that dark frames introduce more noise, so best without them. Take bias frames instead and dither between exposures. Stacking with a clipping algorithm along with flat frames could be a better way against noise. That's certainly true with our bashed around Canon DSLRs. HTH.

When you say they will probably introduce more noise is that becsuse its a dslr?

[carastro]

Quote

When you say they will probably introduce more noise is that becsuse its a dslr?

Yes unfortunately, because they are not cooled.  

[alacant]

1 hour ago, Chefgage said:

When you say they will probably introduce more noise is that becsuse its a dslr?

Not necessarily dslrs. The problem would be the same with any camera where you cannot match the temperature of the light frames with that of the corresponding dark frames.

HTH

[PeterCPC]

Noise can be significantly reduced by dithering between exposures.

Peter

[Chefgage]

1 hour ago, PeterCPC said:

Noise can be significantly reduced by dithering between exposures.

Peter

I will look into it. Just need some clear skies now, looks like cloud/rain for the next week for me. Typical ?

[vernmid]

Hi All,

I seem to recall mention of using Bias frames as darks in DSS. Is that correct and if so then why would it make a difference over using Bias as bias and not using any darks? Is there a flow chart for showing how DSS applies calibration frames?

Many thanks

Vern

[happy-kat]

See the link below to answer.

DSS Link to processing of calibration files

[vernmid]

26 minutes ago, happy-kat said:

See the link below to answer.

DSS Link to processing of calibration files

Thanks happy-kat, Very interesting. The document seems to suggest that not using Darks leads to "Improper calibration". So is the suggestion that you use the same bias files as both bias and darks?

Thanks again

Vern

 

[happy-kat]

Register your lights, bias, flats, dark flats. Then copy and rename the master bias and use as dark. 

[alacant]

15 minutes ago, vernmid said:

same bias files as both bias and darks?

What works here for canon dslr where we have spiral dithered between light frames. We are using Siril for stacking.

Produce a master bias frame y stacking the bias frames with median and no normalisation.

Stack the flat frames with the master bias to get a master flat frame using median with multiplicative normalisation.

Now use the master bias and master flat frames to calibrate the light frames. Combine them first then debayer and stack using linear fit clipping with additive normalisation.

That's it:)

HTH

 

 

[dph1nm]

3 hours ago, vernmid said:

So is the suggestion that you use the same bias files as both bias and darks?

Pointless doing this in DSS. You just end up subtracting the bias from the bias! Works perfectly fine if you just ignore darks completely. I just use lights, flats and bias.

NigelM