First OSC Workflow tutorial -UGC300 vs 10 Part WBPP

Roger Smith

Hi Adam,

I watched your 10 Part WBPP videos on Youtube before deciding to purchase Fundamentals. Very clear, and with reasons behind the usage. Part 9 was very helpful.  Previously I used a CCD camera had been using more the CCD workflow for my new CMOS cooled camera. Happily I discarded all my biases and made some flat darks! No more optimizing with super long exposure darks! Everything matching is best.

I know many things in image processing are not clearly right or wrong, and I really appreciate what I learned especially in the latter parts of OGC300 Workflow example. In this example the data provided to you had bias frames and no flat darks. You had darks that did not match the lights. This really surprised me after watching WBPP Part 9 for CMOS cameras. 

Using bias frames as flat darks (and forcing the times to match), and forcing the darks times to match the lights seemed wrong. Thinking back the provided flats for OGC300 were several different exposures, I think difficult to provide matching flat darks. 

The OGC300 example bends the logic of WBPP where Calibrate with Flat Darks checkbox means they time-match, and then you show how the bias can be a Flat Dark. With the data given, you certainly processed the calibration as best possible with the data given, and in the end show show an image that is very nice. This should  give imagers without the greatest calibration data confidence how to process it. 

For me, I will be making matching Flat Darks with light panel Flats, and matching Dark exposure times with Lights! Now I really need to work on post calibration!

Thanks,

     Roger

BlockHead

So, using a bias to calibrate a flat is a perfectly fine thing to do. It DEPENDS on the sensor. Unfortunately not all sensors behave well...and require Flat Darks. I think in the future CMOS sensors will improve and this will not be necessary. Bias frames to calibrate all short exposures will be fine.

But in the meantime it is best to calibrate Flats with a Flat Dark if you do not know the behavior of your chip or what to look out for that can be subtlety wrong. 

In the NGC 300 the darks DID match the lights. So... we disagree on this..unless I am really mistaken about something. It was on the Flats... If you watch FastTrack you will see the same thing of using Biases to calibrate flats. This is actually the NORMAL thing to do..except some CMOS sensors screw it up!

-the Blockhead 

Roger Smith

Thank you for your further insights.  I really like your style of explaining what is behind what we are doing, and not blindly telling us Step 1 to Step 10. For example Part 8 of WBPP you created the Pictorial showing the components that make up the Lights, Darks, and Flats., and how the electronic signature is removed fro the final lights. 

So a sincere thank-you to you! 

Regarding using bias with flats, I do not know my CMOS camera characteristic, and I like the idea of matching flat darks to the flats. 

I did double the NGC300 Part 1 video, and (at 9:30) you examine the darks as 100 sec and lights at 120sec, and logically move forward since the dark current is very low. (I did not download the data since my main PI computer is in for repair.)

As I mention before, I learned from this example more about the strictness needed, or not needed to have a great looking end result.

Thanks,

     Roger

BlockHead

Hi Roger,

You are right! I did not remember that detail.

When I receive data from other people... it is remarkable what I get... can't keep track.

However, this is exactly what people want me to do. "Adam, don't use your data... use data just like mine!"

Well..this is a consequence of that. lol

I guess I will need to look back at when I do the calibration...I sure hope I did it correctly. 

I probably ended up loading them as custom and forced the time to match. 

-the Blockhead (living up to it!!!!!!!!!)

Roger Smith

Hi Adam,

You are definitely a funny guy! That is what makes the world, and galaxies, fun.

Now that I have read, as you say, the last chapter in the book, I am going to watch Fundamentals for additional details and understanding of what I have missed during my last 2 years with PI.

Roger