Best Tracker For Night Sky Photography Using Medium Format Equipment

[happy-kat]

19 hours ago, JoeKitchen said:

Also, for noise reduction from heat on a long exposure, my camera will automatically take a second exposure of equal length as the sensor is cooling down to map out the heat distribution and use that mapping to reduce noise from heat for that specific capture.  I can not turn this function off.  So effectively if an exposure is 5 minutes, it will take 10 minutes to actually process.  (I am an extremely patient man by the way.)

it might be an idea to test your camera now on a static tripod. Face east or west keeping altitude reasonably low take a single shot with your widest lens on a dark clear night.

Taking your widest lenses 600/28 gives roughly 21 seconds before startrails.

Take an image

Turn off in camera noise suppression if you can

Take another shot

Compare images to see what impact in camera noise suppression has on the stars, you may find some have disappeared, this may be acceptable for you.

 

[JoeKitchen]

2 hours ago, happy-kat said:

it might be an idea to test your camera now on a static tripod. Face east or west keeping altitude reasonably low take a single shot with your widest lens on a dark clear night.

Taking your widest lenses 600/28 gives roughly 21 seconds before startrails.

Take an image

Turn off in camera noise suppression if you can

Take another shot

Compare images to see what impact in camera noise suppression has on the stars, you may find some have disappeared, this may be acceptable for you.

 

Thanks for the advice, but I can not turn it off.  Also, I am not sure it would really matter. 

I don't know all of the engineering, but from what I have been told the noise from heat develops in the top portion of the diode within each pixel.  Some how, Phase One, on long exposures, has their backs ignore the top portion of the diode and only records information from the bottom 2/3s.  This, combined with the dark frame, allows the digital back to remove the noise from heat on a per pixel basis, but also has the effect of increasing the base ISO by one and a third stops.   It is more of a physical property of the digital back. 

Beyond that, you would have to talk to the engineers at P1 on how this works, but I would doubt they would be willing to give any more information then that. 

Of course, this only works for so long.  Eventually heat will travel into the remaining portions of the diodes, negating what the dark frame can do.  At 68F the limit is one hour.  At, say, -20F, I have been told the limit is 10 hours. 

Now the increase in ISO for the long exposure mode does increase noise, but noise from heat working at base ISO on long exposures is much worse.   To get around this added noise from increased ISO, working on the principle that all sensors are ISOless, I over expose by one stop on long exposure mode and then pull the file one stop back in raw processing, which almost completely eliminates noise down to base ISO levels in normal working exposures.  Once again, this is not a noise reduction algorithm, just me working with the inherent properties of sensors.  All noise reduction algorithms are applied by Capture One (P1's raw processor) and can be adjusted by the user. 

Edit

Just to clarify, by ISOless I mean all sensors have a base ISO of 65, assuming there is a bayer grid on it.  Sensors that can only take black and white have no bayer grid and therefore have a base one stop higher.  If you are shooting at a higher ISO, assuming the entire resolution of the sensor is being recorded, the sensor itself is not operating at a higher level of sensitivity.  Instead the file is being pushed.  Assuming your camera records the entire raw file (Canon clips theirs when it is over exposed at higher then base for some reason), pulling the exposure down in post will give you the same quality as shooting the number of stops your pulled the exposure down in ISO. 

The exceptions to this are with cine cameras, which usually have a base of 800 ISO.  Although their sensors still have a working base of 65, cine cameras reduce the file size dramatically from the native resolution.  Reduction in file size inherently reduce noise and partially "recover" the color data (from over sampling) that was lost due to a higher working ISO. 

Edited December 13, 2023 by JoeKitchen

[JoeKitchen]

4 hours ago, Elp said:

The issues you'll have with your existing camera is whilst you're imaging even doing 30s or so exposures, if you're capturing multiple panels of the sky in order to stitch them together (you will need to do this if you want high resolution paired with a long focal length lens) plus taking multiple exposures per panel, you'll find framing the panoramic stitch will not be perfectly landscape, it'll be stepped in a zigzag arrangement and you'll have to crop. You'll be surprised how quickly the earth rotates.

Regarding manual function of the shutter, have you tested this on star fields? A lot of DSLRs have a shutter lock function to stop the mechanicals from moving, minute vibrations are disastrous for star imaging, even walking around your setup whilst imaging is enough to cause this.

 

I did think about this, dealing with the zig zag.  The panorama software I use compensates for field of view, giving me files that kind of the look like the profile of an extreme concave lens to help maintain parallax.  Now whether or not I want to use this setting for stars and/or how much the zig zag would interfere with normal cropping I wont know until I actually shoot something.

Insofar as the mechanical vibrations, my camera is a digital version of a 4x5 camera.  There is no mirror, there is no auto focus drive, the apertures are set manually, the digital back has no moving parts.  The only thing that moves is the Copal shutter.  For reference, below is a 100% crop captured from a concrete pier in Key West using my 55mm at 12 seconds, ISO 140. 

Edited December 13, 2023 by JoeKitchen

[vlaiv]

7 hours ago, JoeKitchen said:

Speaking of nodal points, is it important to have the nodal point of the lens over the axis of rotation of the mount?  Or are the stars so far away it does not really matter?

Stars are effectively at infinity so it really does not matter at all.

You can have your lens mounted at any sort of angle with respect to rotation of the mount and at any sort of distance - as long as lens follows mount's rotation - you'll be fine.

From what I gathered in this thread - you really have two options:

1. guiding

2. encoders

First is more involved but less expensive, second is less involved but probably way over your budget.