Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

The order in which I should take Light/dark/bias frames in? Also, camera warmup shots?


Recommended Posts

Hello, I won't be taking many shots until astronomical twilight returns but this has been bugging me.

Because I'm using a DSLR which, being a DSLR, does not have cooling of any kind, I thought it might be a good idea to get my camera to take a 10 minute exposure just to... Warm up before I start taking light frames. I thought of this because otherwise the darks I take will correct for the heat noise incorrectly. Although, I am kinda worried that might not bein the best interests of my camera's health. DSLRs not exactly being built for 10 mins exposures 'n' all.

Also, I have not taken any bias frames before but I intend to start. I don't know whether to take them before I start the imaging session when the heat noise hasn't built up yet or after when I get home. I don't know how read noice changes in a camera over time... Or even if it does at all.

So, to summarise:

    Will taking really long exposures just before taking my light and then dark frames cause damage to the camera in the long term?

    When is the best time to take my bias frames? Before, or after my imnaging session. Or maybe I only need one set ever?

   (p.s. will a peice of A4 paper work for flats and does the image have to have white balance set correctly?)

        ~pip

Link to comment
Share on other sites

Don't worry about "warming the camera up" - I have never noticed any obvious difference between my first and last "light" of a session (maybe 4 hours or more of 5 min subs).  I usually start a long imaging run before it is truly dark and therefore often throw away my first two or three frames anyway.  Better to do that than miss out on valuable imaging time by starting late!

Darks can be taken and kept for future use - they should be taken at the same temperature as your lights.  I took a whole series of darks (at different temperatures - about every 5°C) in my garage which was close to outside temperature, over a few months of cloudy nights.  I must admit that i now find that darks add to the noise in my images and have therefore stopped using them.

Bias are quick and easy - again you can make a library.  Temperature is not important - just fire off a load of frames at the shortest shutter speed you have.  I use 100 frames at 1/4000 sec.

You mention 10 min exposures - i would suggest 2 mins to start with.  This will allow you to sort out your technique and equipment etc.  You don't say if you will be guiding?  If not then you may only get away with 1 min exposures without trailing.  If you are guiding then you may prefer 5 min subs as a maximum.  This way you get a fair number of subs in a reasonable time and the loss of a sub is not so critical.  You only get 12 5min subs in an hour!

For flats (which must be taken at the same settings as your lights I use one of these:  http://www.amazon.co.uk/gp/product/B00J3NRAV2/ref=ox_sc_act_title_1?ie=UTF8&psc=1&smid=A23HHFRIW02ZZW   And put it over the end of the scope at the end of the session, put the camera on "Av" and take around 30 flats.  If you set the light panel to low you will have an exposure of about 1/3 - 1/4 second which seems to work well for me.

Link to comment
Share on other sites

Don't worry about "warming the camera up" - I have never noticed any obvious difference between my first and last "light" of a session (maybe 4 hours or more of 5 min subs).  I usually start a long imaging run before it is truly dark and therefore often throw away my first two or three frames anyway.  Better to do that than miss out on valuable imaging time by starting late!

Darks can be taken and kept for future use - they should be taken at the same temperature as your lights.  I took a whole series of darks (at different temperatures - about every 5°C) in my garage which was close to outside temperature, over a few months of cloudy nights.  I must admit that i now find that darks add to the noise in my images and have therefore stopped using them.

Bias are quick and easy - again you can make a library.  Temperature is not important - just fire off a load of frames at the shortest shutter speed you have.  I use 100 frames at 1/4000 sec.

You mention 10 min exposures - i would suggest 2 mins to start with.  This will allow you to sort out your technique and equipment etc.  You don't say if you will be guiding?  If not then you may only get away with 1 min exposures without trailing.  If you are guiding then you may prefer 5 min subs as a maximum.  This way you get a fair number of subs in a reasonable time and the loss of a sub is not so critical.  You only get 12 5min subs in an hour!

For flats (which must be taken at the same settings as your lights I use one of these:  http://www.amazon.co.uk/gp/product/B00J3NRAV2/ref=ox_sc_act_title_1?ie=UTF8&psc=1&smid=A23HHFRIW02ZZW   And put it over the end of the scope at the end of the session, put the camera on "Av" and take around 30 flats.  If you set the light panel to low you will have an exposure of about 1/3 - 1/4 second which seems to work well for me.

You've found darks are adding to your noise? Hmm. Maybe I'll have to check that when I get back to taking images ~1.5 months from now.

So... I can just make a huge library of biases and they last forever? Sounds great! I understand they're best taken with lens cap on and at the loswest exposure time. I presume I use the same ISO as I do when taking lights?

Maybe when I get some money I'll consider that light panel. But I might as well be broke right now so I'll have to pass. Maybe I should experiment with the paper anyway...

I'm not guided (unfortunately, same issue as stated above) but it's on my list. If there's little to gain from a warmup shot I'll just ignore it completely. I've ended up using 20~ second exposures for my 35mm focal length. it seems to work.

Link to comment
Share on other sites

oops!  I forgot to mention ISO - you should have the same ISO for all images.  You don't say which camera you have so I cannot say specifically, however with both my modded Canon 1000D and unmodded 1100D I use an ISO of 800.  

If you are unguided then 30 sec or maybe 1min may well be your maximum exposure time.  However you can take 100-200 subs and throw away any with trailing.  When I started out I was unguided and reckoned on throwing away 30 out of every 100 subs !!  So taking 3 hours worth (about 180) of 1min subs is not too difficult and will result in a decent image.

You will also have to choose your targets - clusters and galaxies maybe rather than faint nebulae to start with.

This is one of my very earliest images from Sept 2009 - about 100 x 1 min subs of the double cluster, no flats, bias or darks:

post-4502-0-62550200-1434303464.jpg

Link to comment
Share on other sites

Over the course of an imaging session, the sensor temperature will rise.  Here is a graph of dark current for successive 5 min exposures.  I haven't yet added the Sony A7S to the chart but it's pretty much identical to the Nikon 7000D - both are Sony Exmor sensors. 

post-19658-0-06639400-1434303804.jpg

So darks taken at the end of the imaging session will, overall, be a better match but they won't be a good match for the lights taken at the start of the imaging session.  So better results will be obtained by switching on the option in your processing software that automatically matches the master dark to each individual light.  The software calculates a scaling factor to apply to the dark to best match the thermal noise in each light.

In fact, you can make it it much easier on yourself - I actually use the same master dark I created a few years ago - the software happily scales this as appropriate to every individual light in every imaging session I do at a large range of ambient temperatures.

Mark

P.S. I should have added that if you are imaging with one of the Nikons that does black level clipping then darks will not work very well at all unless the Nikon Hacker hack for black level is installed.

Link to comment
Share on other sites

Hello, I won't be taking many shots until astronomical twilight returns but this has been bugging me.

Because I'm using a DSLR which, being a DSLR, does not have cooling of any kind, I thought it might be a good idea to get my camera to take a 10 minute exposure just to... Warm up before I start taking light frames. I thought of this because otherwise the darks I take will correct for the heat noise incorrectly. Although, I am kinda worried that might not bein the best interests of my camera's health. DSLRs not exactly being built for 10 mins exposures 'n' all.

Also, I have not taken any bias frames before but I intend to start. I don't know whether to take them before I start the imaging session when the heat noise hasn't built up yet or after when I get home. I don't know how read noice changes in a camera over time... Or even if it does at all.

So, to summarise:

    Will taking really long exposures just before taking my light and then dark frames cause damage to the camera in the long term?

    When is the best time to take my bias frames? Before, or after my imnaging session. Or maybe I only need one set ever?

   (p.s. will a peice of A4 paper work for flats and does the image have to have white balance set correctly?)

        ~pip

The two most important Calibration frames you should take are the Bias and the Flats. Darks will do more damage than good if not taken correctly and this means set temp cooling that a DSLR does not have so they are best left alone. Also the newer breed of cameras from Canon and perhaps Nikon do on chip dark current manipulation which complicates the matter further ( the so called raw data is not really raw) . But for now concentrate on the Bias ( the easiest to take ) and the Flats, perhaps the most difficult as how well they work depends on 1- the imaging train must not be disturbed after finishing the Lights 2- the illumination must be absolutely even 3- The level of illumination should be between 30%~ 50% of the full well depth of the sensor.

Condition 1 means that the camera and the lens should preferably not be moved before Flats are taken. For 2 a decent EL panel through a neutral diffuser is the best way of going, just set the camera to Av mode and take about 30~50 flats if you find that exposure is not correct ( check the live histogram ) you can use the exposure compensation in 1/2 stop increments to take another series and see which one works best. A Flat frame as applied to the light frames is only luminance so the colour does not come into the equation with the caveat that none of R,G,B channels should be allowed to be clipped.

Bias is the easiest one to take as you just set the camera to the fastest speed possible ( theoretically the exposure should be Zero to establish the pedestal value but fastest speed is fine ) and it is best taken with lens removed and the camera off the scope but with both the viewfinder shield and the camera cap on so no stray light gets into the camera and lift the Bias value.

As to how many to take , the answer is the more the better. The reason is that every calibration frame brings it's own noise into the equation by taking a lot of these you are allowing the calibration software to optimise the data before applying them to the Lights. I have taken as many as 200 Bias and over 100 Flat frames and it does make a difference. I don't bother with Darks for neither my DSLR nor my CCDs.

Regards,

A.G

Link to comment
Share on other sites

Coincidently spent some time today researching DSLR Flats exposure time.


Many sources suggest the flats should be exposed at the camera's Median level, which on my 12bit DSLR is 2048 out of 4096, but I failed to find a way to display the bit level on a RAW image.


The sources stressed that 2048 would not be halfway along the camera's histogram, due to the gamma correction applied to the RAW in order to display it, the Median level would be way to the right on the histogram.

Exposing in AV mode did give a histogram about 3/4 along, so this may be not be too far off "right".


Also suggested taking Flats at a low ISO, say 200 instead of 800 or 1600 used for the Lights, to avoid adding more noise to your images.


Michael

Link to comment
Share on other sites

Coincidently spent some time today researching DSLR Flats exposure time.
Many sources suggest the flats should be exposed at the camera's Median level, which on my 12bit DSLR is 2048 out of 4096, but I failed to find a way to display the bit level on a RAW image.
The sources stressed that 2048 would not be halfway along the camera's histogram, due to the gamma correction applied to the RAW in order to display it, the Median level would be way to the right on the histogram.
Exposing in AV mode did give a histogram about 3/4 along, so this may be not be too far off "right".
Also suggested taking Flats at a low ISO, say 200 instead of 800 or 1600 used for the Lights, to avoid adding more noise to your images.
Michael

This is true, I usually set the Exposure compensation to + 1/2 and sometimes to +2/3.

A.G

Link to comment
Share on other sites

Hmmm, if darks are only going to hurt my images noise wise. What am I going to do about hot pixels? Are those removed with bias frames because darks were the only thing stopping DSS from interpreting 3000 hot pixels as stars and accidentally causing me to make a star trail image :rolleyes:

Link to comment
Share on other sites

Hot pixels can be dealt with by a combination of applying darks, and using sigma clipping when stacking (requires a min of 16 subs to be effective in DSS). Sigma clipping also work well in getting rid of satellite trails etc.

Darks wont hurt DSLR images if properly applied (quite the opposite, they can do a great job!), you need a matching set of bias frames to calibrate the darks (they remove camera read noise) before they are applied to your lights. DSS sorts all this out for you, but its best to have a full set to be on the safe side (dark/flat/bias).

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.