T3i 600D Mirror lock-up not working with BYE 3.1.13

[rtemen]

I need to take many very fast exposures.

It is suggested that the problem I am having with my flats is that there might be mirror mechanism shaking or timing.

It suggested that I set a mirror lock-up for a second or two before the shot to see if that settles the problem down.

Here is what is happening with my process:

1. I enable mirror lock-up on my camera
2. I set the mirror lock-up to one second
3. I hit Start Capture
4. In 30 to 45 seconds the mirror goes up
5. In another 30 to 45 seconds, it sounds like the camera takes the shot
6. In 59 seconds, I get an error message that the download has timed out
7. At this point the progress clock at the top right keeps spinning and BYE is locked up and I have to kill it because after many minutes it will not respond
8. I even try Aborting, and nothing will wake it up.

Any suggestions?

Rich

 

[admin]

2. Where do you set this 1 second?  Camera?  BYE?

 

[rtemen]

BYE in the Mirrorlock under the Histogram.

[admin]

Don't setup mirror lock in-camera in step #1, BYE should turn it on and off automatically since you're setting the number of second in BYE.

[rtemen]

It does the same whether the camera has lock enabled or not.

 

[s3igell]

Mirror Lockup (MLU) is NOT worth the hype, unless you are taking Short Exposures on an Unsteady Photographer's Tripod.  And, if you are, MLU is amongst the least of your Challenges...

MLU is not worth it - Why:

1. AP Image Exposures are Long Duration - any Mirror-Slap induced Vibration will die-out in the first 1-2 seconds (during which very few Photons have been captured)
2. AP Mounts are Massive (compared to the DSLR Mirror-Box - any Mirror-Slap induced Impulse will be overwhelmingly dampened by the Mass of the Mount
3. AP Imaging involves Purposefully Induced Motion - Dithering will overwhelm any slight perceived Motion attributed to Mirror-Slap
4. AP Imaging involves Stacking with Statistical Noise-Reduction Functions which will reduce/remove any recorded Vibration from Mirror-Slap

[rtemen]

Well, the idea here is not to stop Mirror-Slap Vibration.

Several forum discussions seem to point out that when taking many flats, and the exposure time is in the 1/1500 second range that maybe the mirror mechanism is still partly in the way causing a false shadow on the sensor.

So I was trying to put a 1 or two second wait time to make sure the mirror is out of the way and settled in.

Does this make any sense?

 

[astroman133]

That seems like a malfunctioning camera, especially considering that some cameras allow much shorter exposures.

Do you actually see a false shadow in your flats?

Have you thought about dimming your flats light source so that you can use a slight longer exposure?

[astroman133]

I had thought from your original post that you want to raise the mirror and keep it up while shooting an exposure plan. But instead it seems as though you would be OK with raising the mirror waiting a second taking the shot and lowering the mirror, and repeating.

With my T5i I had no problem doing this with BYE version 3.2.1 - RC3. However I did not make any changes in the camera's menus for this. the Mirror Lockup setting under Custom Functions is Disabled when I look while the camera is disconnected. All I did in BYE was to unhide the Mirror Lock control and set it to 1 second.

It works as expected. Perhaps it is because the T3i is an older model than my T5i.

 

[rtemen]

Hi, Rick.

Yes, I have done all of those things and at this point in my investigations on what is wrong, I have decided to stop and do a major reset and start all over fresh. That being said, I am just wondering why my Mirror Lock functionality will not work. Just so I know what is wrong for future endeavors.

 

[astroman133]

I don't know what you mean by "a major reset and start all over fresh". I would not recommend uninstalling and re-installing BYE. Because everything else works, it is very unlikely that this would do anything to give the program different functionality with your camera, unless you install a newer version than what was previously installed. Then you would likely only get the incremental changes from the newer version.

Well, the Supported cameras table says that MirrorLock is supported over USB for the T3i AND that it was tested, I assume by the BYE developers. Perhaps another T3i owner can test MirrorLock and report their results.

If you have a friend with a different, supported camera model then perhaps they could help you test to see if there camera operates as expected with Mirror Lock.

You could also try a factory reset on the camera.

Reviewing the log file may also provide a clue about what is happening.

[rtemen]

By reset I just meant that I would take a complete new set of Bias, Dark, Flat and Lights while being more careful on my setup, etc.

I have added a Starizona Hyperstar unit and I am learning how to use it and set things up.

 

[astroman133]

If no light is reaching the sensor, whether the shutter is completely open or not should not matter. Flats should be taken for each imaging session or whenever the camera is touched. Based on that I am not sure what you expect to gain by taking new darks and biases. Shooting flats is sometimes more art than science.

If you have shadows in your flats, I would look at how you are routing the USB cable to the camera. Also see if the shadows change after moving the cable path in front of the corrector plate.

What is your aperture? I hope that it is 11" or 14". With a DSLR, the larger aperture is a must due to the amount of obstruction of the camera body.

How are you taking flats? This could be tricky for HyperStar due to the wide FOV. Have you asked Starizona for their recommended technique?

[rtemen]

I have a Celestron C11. My camera is a Canon T3i (600D)
I have tried changing the orientation of the camera and cable, no difference.

I take my flats right after taking my Lights by placing my white panel over the end of the scope. Using a dew shield to extend the tube above the camera. My light panel has brightness control and I have tried several levels.

I use Pixinsight to do my processing. I have been using the Weighted Batch Preprocessing script.

The attached file is the result and we are trying to find out why this dark band is there at the bottom.

 

 

[astroman133]

How do your flats look, as far as the dark band and the histogram? I would not use the BYE histogram as this is based on the JPG that embedded in the RAW and as such has an auto-stretch applied. Look at the histogram of an individual flat frame in PI

Does the processed image change if you don't apply the master flat? Like does the dark band disappear?

I don't use PI, so I can't comment on changing that process.

I use an electroluminescent panel to shoot flats. The panel operates with AC power so its brightness changes at a 60 Hz frequency. I need to be sure that my exposures are several cycles in length so I don't see refresh artifacts in my flats. My panel is not dimmable, so I put a few sheets of white printer paper in front of the panel to darken it. My shortest exposure for flats is 0.2 sec with a QSI583wsg monochrome CCD camera. I don't typically shoot prime-focus long exposures with BYE. I would suggest setting your flat exposure to say 1/8 second and adjust the brightness of the panel to give a well centered histogram from the RAW unstretched image. Then look critically at an individual flat frame, looking for the unexpected gradients.

I hope this helps.

[rtemen]

Yes, if I don't use the flats, the shadow is gone.

[rtemen]

Here is a screenshot in my PI.

I think this shows what you are asking earlier.
In the statistics, I was taught that the mean and median values should be in the 7 to 8 thousand range.
I also show the histogram.

What do you think?

[astroman133]

I am not sure where you heard that the pixel values for flat frames should be 7-8K. Here is what I have heard:

An ideal flat frame histogram would have a peak at about the 50-60% level and have all the histogram between the 20% and 80% levels.

With that in mind I would suggest increasing your exposure by a factor of 3-4.

Here is what one of my master flats looks like. It is from a mono camera and is a full-sized luminance image:

The histogram is logarithmic which has the effect of stretching out the X-axis. A linear histogram would be narrower and right at the x-axis midpoint. The line profile shows that the values range from 37045 at the center (mid-point of the diagonal line) and drops off to 32808 at one end. The pixel values are in a sixteen-bit unsigned integers with a range from 0 to 65535.

I hope this helps.

[rtemen]

By the way, I had been tutored by a few different people on taking good flats.
Here is one of them.

https://www.myastroscience.com/proper-flats-with-dslr

Perhaps I am not understanding something correctly???

[astroman133]

The link in your post does not agree with this post --> https://www.cloudynights.com/topic/616102-fullwell-capacity-vs-max-adu-misconceptions-about-taking-flats/

Not sure how to reconcile the differences.

The myastroscience article says to take flats at 50% of the full well depth and then uses the ADU value which is scaled in the camera by the gain. The CN article says that this is wrong and to take flats at 50% of the max ADU, which is what I do.

This quote is from the CN link:

Quote

 

The problem is many people confuse their cameras FW with ADU values. When using our cameras on a daily basis the FW number doesn't mean anything. The only number we care about is the ADU. So when someone says take your flats at 50% of your cameras FW capacity, what they are talking about is 50% of 65535 ADU value the software shows not the published FW of your camera. 

 

 

[s3igell]

Both of you are arguing from an essentially outdated effort to ensure that the Flats were within the most Linear (center) portion of the CCD Chip's Output.  The days of CCD are Long Past, and with modern CMOS Sensors and ADC Units, Linearity exists almost to the far ends of the Histogram.  Assuming your Flats are far enough separated from the Null and Fully Saturated, and within the Linear Response area of the Sensor, everything else is straight forward Math performed within the Master Flat Calibration routine.

(I'm having no luck finding my preferred reference - site may be down for repairs.)

[astroman133]

Not that long past. I still image mostly with a CCC camera.

Any idea what is wrong with rtemen's flats?

[dkerber]

9 hours ago, astroman133 said:

Not that long past. I still image mostly with a CCC camera.

Any idea what is wrong with rtemen's flats?

Looking at the flat itself, I didn't see anything noticeable, but my first guess would be that the flat panel isn't as flat as it should be. I would also suggest lengthening the exposure, as was suggested above. I have had the best results putting the histogram peak near the middle of the range, and I use the same imaging camera as the OP does (T3i/600D).

[rtemen]

Hi, everyone.

I have a number of questions and I will start by apologizing for my lack of detailed knowledge on these issues.
One of my problems is that I do not understand how all of these numbers and mnemonics apply to my equipment. So in many of the articles these various numbers are explained, but I cannot seem to find any 'tutorials' showing me what to do with my equipment and software.
For example, I have my T3i (600D) and I use BYE to capture my shots. It has been explained that I should not use BYE's histogram and instead, look at the numbers of my flat in my processing software. I use Pixinsight for my processing. So, when I open up a flat and look at the statistics, I see the mean and median numbers. I had been coached in the past that these numbers should be in the 7 to 8 thousand range. 
Now, here is one of the questions that I have not found an answer to: in PI, should I change the display to show 14 bit like my camera is, or do I leave it at 16 bit? 

Then the next question is that when taking the shots with BYE, I need to make my exposures longer and longer to get this ADU level up so that PI shows mean and median to be in the 20 to 30 thousand range. As I make my exposures longer to get towards that median number, I seem to run into the following issues:

1. it looks like the BYE will let me lengthen my exposures, but it seems to clamp off at a certain level and exposing the flat longer makes no difference. Therefore, I have not found a way to get the mean and median anywhere close to the 20 to 30 thousand level.
2. When trying to take longer exposures to attain these levels, the shot is way over exposed and is basically all white.

A couple of years ago when I built my lightbox, someone (I cannot remember who any more) helped look at the panel and we determined that it was very even across the whole panel. Perhaps this is wrong as well?

And as a last item for this discussion, I have not has this issue with the dark shadow band on any of my old shots, only now that I added my HyperStar unit.

I know this is a lot, but thanks for any help or guidance.

 

[astroman133]

Can you make a raw flat image available so that I can download it and look at it with the software that I use.

I am concerned that I may have given you bad advice because I am not familiar with PixInsight.

[rtemen]

Well, I am open as to how to send it to you. It is too big for this forum.
I can dropbox it but I would need your email. Have you another suggestion?

 

[rtemen]

Try this.
One is Straight, the other is rotated 180 degrees.

https://drive.google.com/drive/folders/1OFapkq8zy8PDozuJwQar1rMdCleEFti9?usp=sharing

Rich

 

[astroman133]

I'll send you a link via private message.

 

[rtemen]

Thanks for your time and analysis on this.

I did another test and I put the flat Test_Light in the google drive. Can you do that analysis on that one too and see if it is better?

Thanks.

[astroman133]

One other thing occurred to me. Do you cover the camera's viewfinder when shooting flats?

[rtemen]

Yes