47 replies to this topic

[mmalik]

Given that EOS 6D's manual is quite ambiguous in explaining whether or not 'Long Exposure Nose Reduction' (LENR) [often used in astrophotography] is possible with its 'Multiple Exposures' feature, let's have a dedicated discussion on it.


This discussion will focus primarily on LENR, and whether or not LENR can be coupled with 'Multiple Exposures' where one dark gets applied to multiple lights [in RAW mode].


Note: Secondary discussion can be had about merging 'Multiple Exposures' without LENR [without taking a dark] which is NOT something typically used in astrophotography.


Relevant background discussions here..., here... and here.... Regards

[mmalik]

@Thirteen: What additional testing are you guys proposing on the 6d?

What I would propose for the start is that we visually inspect what's happening with the shutter as multiple LENR images are coupled with one dark. If a dark of equal length gets taken at the end of say three lights of e.g., 300 sec, let's examine if shutter remains closed for 300sec at the end of taking 3 lights (while busy is displayed in LCD). A control test would be repeating the test without LENR and seeing that camera doesn't take a dark at the end of taking three lights. Regards

[jgraham]

Wow, it sounds like someone at Canon may have suffered a brain cloud. To apply one dark to a series of images is insane. The temperature changes alone would wipe out any benefit. Imprinting the same 'noise' over and over again wouldn't help either.

I use the long exposure noise reduction all the time, but this sounds crazy. I'm surpised that a company that has shown some knowledge of the unique challenges of astrophotography would implement somehing like this. Very odd.

[mmalik]

Wow, it sounds like someone at Canon may have suffered a brain cloud. To apply one dark to a series of images is insane. The temperature changes alone would wipe out any benefit. Imprinting the same 'noise' over and over again wouldn't help either.

I use the long exposure noise reduction all the time, but this sounds crazy. I'm surpised that a company that has shown some knowledge of the unique challenges of astrophotography would implement somehing like this. Very odd.

John, this is NOT confirmed yet; that's the reason for this thread to find out exactly what they mean since 6D manual is not clear on the subject. Regards

[Thirteen]

@Thirteen: What additional testing are you guys proposing on the 6d?

What I would propose for the start is that we visually inspect what's happening with the shutter as multiple LENR images are coupled with one dark. If a dark of equal length gets taken at the end of say three lights of e.g., 300 sec, let's examine if shutter remains closed for 300sec at the end of taking 3 lights (while busy is displayed in LCD). A control test would be repeating the test without LENR and seeing that camera doesn't take a dark at the end of taking three lights. Regards

This is what I did last night. 120s exposures separated by 5s interval. If you program the camera to go take shots with LENR on it will pause and do a dark after every trio of lights. I inspected the shots and I can't tell much of a difference in the appearance. When LENR is off, the camera just happily goes on taking exposures.

This are 100% crops of the upper left corner (please excuse the vignetting and fixed tripod).


EDIT: I don't really think this is a terrible way to implement it, if it is doing this as we speculate. You get back some clear sky time and the temperature really wouldn't drift too much. It still takes care of any pattern noise. You go from 50% dark frame down time to 25%, so that's worth something. I just don't see enough of a difference here to justify it yet, though.

Image is 120s, ISO800, f/5 @24mm through a 24-105 f4L

Attached Thumbnails

• 

[G. Hatfield]

I have been using a Hutech modified 6D for about six months now and have not used the in-camera noise reduction. And I don't use darks nor calibration with this camera. But I do use the Camera Raw to reduce the appearance of noise. It really smooths out the noise and I have control over what it does. For an excellent review of the technical aspects of the 6D see Roger Clark's review.

http://www.clarkvisi...n-6d/index.html

He suggests not doing dark subtraction.... in-camera or otherwise.

George

[Thirteen]

I have been using a Hutech modified 6D for about six months now and have not used the in-camera noise reduction. And I don't use darks nor calibration with this camera. But I do use the Camera Raw to reduce the appearance of noise. It really smooths out the noise and I have control over what it does. For an excellent review of the technical aspects of the 6D see Roger Clark's review.

http://www.clarkvisi...n-6d/index.html

He suggests not doing dark subtraction.... in-camera or otherwise.

George

I think I agree with this so far...

[Moromete]

If possible please try 6D's LENR with MagicLantern DualISO function on some DSOs to see if you can get a better dynamic range too.
You need to put the 6D in RAW mode for this.

I tried it once by mistake with my T2i and it works but I didn't know this function was activated.

[Thirteen]

I can't install magic lantern on the 6d I used, sorry.

[ignacio_db]

DualISO makes sense in single-shot photography (on a changing/moving subject). In astroimaging, one can combine different ISOs and exposures sequentially, so no need to interlace.

Also, I would never waste a minute of good sky to take darks. I always do those at home.

Ignacio

[G. Hatfield]

I posted this comparison on another thread, but it seems applicable here too. Here is a comparison of the noise in the unprocessed image from a 6D with the noise after Camera Raw noise reduction and after a median combine of 10 images.

http://www.geoandpat..._comparison.jpg

The ambient temp when these images were taken was 62 degrees F. The image exposure is 2 min at ISO3200.

George

[mmalik]

What I would propose for the start is that we visually inspect what's happening with the shutter as multiple LENR images are coupled with one dark. If a dark of equal length gets taken at the end of say three lights of e.g., 300 sec, let's examine if shutter remains closed for 300sec at the end of taking 3 lights (while busy is displayed in LCD). A control test would be repeating the test without LENR and seeing that camera doesn't take a dark at the end of taking three lights. Regards

If someone could do these visual tests with 6D, that would be great? Note: This test can be done indoors; all one needs to do is just look at the sensor/shutter sequence, i.e., Light(Shutter Open)+Light(Shutter Open)+Light(Shutter Open)+Dark(Shutter Close). Thanks in advance. Regards

[Ralph Paonessa]

I was the one who posted about this feature in the previous EOS noise thread.

I just did a test triggering the EOS 6D with the Canon TC-80N3e intervalometer set to:

- 4 minute exposures
- 5 second delay between exposures
- numbers of frames from 1 to 20+

The 6D was set to Bulb and ISO 3200. (I put on a 20mm lens with the lens cap on and set the aperture to f/5.6.)

For convenience of seeing the results, I tethered the camera to Lightroom 5 on my PC, which imported the images as they were taken.

I enabled LENR in the 6D menu by setting it to On (not Auto).

Results:

1. If I set the intervalometer for 1 frame, the camera shot a 4 minute exposure; then remained busy (red LED illuminated) while it recorded the dark. Thus after 8 minutes the camera was ready and the image appeared in LR. (The actual EXIF time was 240.4 sec.)

2. If I set the intervalometer to multiple frames (>20), the camera would take frame 1, then 2, then 3 (with a few seconds delay between as set in the itervalometer), then spend another 4 minutes recording its dark. When that was done, the intervalometer timing was such (trying to take 4 minute exposures separated by 5 seconds) that the camera would be triggered to take another brief ("throwaway") of 3.2 seconds (the tail end of the 240 secs where the timer was triggering the shutter).

At this point, the four images would download to Lightroom while the next set 3 lights (+1 throwaway) were being taken.

3. I also shot some exposures with LENR turned Off, to see what they'd look like.

Conclusion:

Comparing all the images, it was clear that with Long Exposure Noise Reduction turned on, the EOS 6D subtracts a single dark from a prior sequence of three lights before going onto the next frames and repeating the process.

I first discovered this option in Alan Dyer's updated article DSLR Cameras for Astrophotography. He described this as a nice feature of the 5D, but has now updated it to include the 6D (although he claims 4 lights/dark, but I've only gotten 3.)

[Ralph Paonessa]

I have been using a Hutech modified 6D for about six months now and have not used the in-camera noise reduction. And I don't use darks nor calibration with this camera. But I do use the Camera Raw to reduce the appearance of noise. It really smooths out the noise and I have control over what it does. For an excellent review of the technical aspects of the 6D see Roger Clark's review.

http://www.clarkvisi...n-6d/index.html

He suggests not doing dark subtraction.... in-camera or otherwise.

George,

I looked at Clark's article. In his conclusions:

No long exposure dark frames are needed when making long exposures if recording raw.

Is that what you're referring to? Also, I'm a little puzzled by the "if recording raw"; why would raw vs. JPEG make a difference? I didn't follow all the technical nuances of the article.

I have found that Adobe Camera Raw has excellent tools for noise reduction.

[G. Hatfield]

Generally speaking, you should only process RAW images for astrophotography. They are very close to being unprocessed by in-camera software. This allows you to control the white balance, color balance, brightness, color saturation, contrast, noise reduction, sharpening and all the other stuff that we do in image processing. With JPEG the camera does a lot of this automatically and sometimes the results are less than we would like and irreversible. Also JPEG images are always 8 bit while RAW yields 16 bit TIFF files.

I actually collect both JPEG and RAW, but the former is only there to give a rough idea of what the image looks like and if the histogram is appropriate (i.e., exposure and ISO).

So one should ALWAYS start with RAW images. I don't think Roger is suggesting that you need darks with JPEGs. JPEG images are basically "throw away" images since all your processing should be done with the RAW files. I think what he is saying is to get optimal noise reduction, you need to use RAW files.

One side note about the 6D is that CR2 RAW files cannot be directly imported into Photoshop. They must first be converted by digital negative files (DNG) using a program like Adobe's DNG Converter which is a free download.

In general terms this is my workflow for 6D images.
1. Convert CR2 RAW files from the camera to DNG files
2. Open all the DNG files (usually 10-20) in Adobe's Bridge
3. Then open all of them at once into Camera RAW from Bridge
4. Do noise reduction, etc in Camera RAW
5. Save as 16 bit TIFF files
6. Register and combine (median) the TIFF files using Registar
7. Open the combined image in Photoshop or CCDStack depending on what process I'm using (RGB vs LRGB).

Note that I do not use darks nor do I use calibration. It is my impression that the color is preserved much better without calibration, but I have no data to back this up.

George

[Ralph Paonessa]

Generally speaking, you should only process RAW images for astrophotography ...

One side note about the 6D is that CR2 RAW files cannot be directly imported into Photoshop. They must first be converted by digital negative files (DNG) using a program like Adobe's DNG Converter which is a free download.

I agree with you completely on the value of raw. I've never shot a JPEG since my first EOS 10D.

What version of Photoshop are you using? I've got Photoshop CC and Lightroom 5, and they import 6D images. (Of course, that compatibility is how Adobe encourages you to upgrade. )

[G. Hatfield]

Ralph....I use CS5 and Camera Raw is a bit different in CS6. Might upgrade, but the monthly fee now required has made me think twice about upgrading.

[G. Hatfield]

Ralph...P.S. Should have checked your website before replying on "RAW"....obviously you know everything I posted! Beautiful shots!

[keithlt]

George , there is a special for cc and light room for 9.99 month for a year but the buy ends this may.

[mmalik]

I just did a test triggering the EOS 6D with the Canon TC-80N3e intervalometer set to:

- 4 minute exposures
- 5 second delay between exposures
- numbers of frames from 1 to 20+

The 6D was set to Bulb and ISO 3200. (I put on a 20mm lens with the lens cap on and set the aperture to f/5.6.)

For convenience of seeing the results, I tethered the camera to Lightroom 5 on my PC, which imported the images as they were taken.

I enabled LENR in the 6D menu by setting it to On (not Auto).

Results:

1. If I set the intervalometer for 1 frame, the camera shot a 4 minute exposure; then remained busy (red LED illuminated) while it recorded the dark. Thus after 8 minutes the camera was ready and the image appeared in LR. (The actual EXIF time was 240.4 sec.)

2. If I set the intervalometer to multiple frames (>20), the camera would take frame 1, then 2, then 3 (with a few seconds delay between as set in the itervalometer), then spend another 4 minutes recording its dark. When that was done, the intervalometer timing was such (trying to take 4 minute exposures separated by 5 seconds) that the camera would be triggered to take another brief ("throwaway") of 3.2 seconds (the tail end of the 240 secs where the timer was triggering the shutter).

At this point, the four images would download to Lightroom while the next set 3 lights (+1 throwaway) were being taken.

3. I also shot some exposures with LENR turned Off, to see what they'd look like.

Conclusion:

Comparing all the images, it was clear that with Long Exposure Noise Reduction turned on, the EOS 6D subtracts a single dark from a prior sequence of three lights before going onto the next frames and repeating the process.

I first discovered this option in Alan Dyer's updated article DSLR Cameras for Astrophotography. He described this as a nice feature of the 5D, but has now updated it to include the 6D (although he claims 4 lights/dark, but I've only gotten 3.)

Ralph, thanks for your tests, that settles it then!

Here is the quote from the reference... you cite, "What about the new 6D? It can take four exposures before forcing a dark frame. That speeds up exposure acquisition a lot..."


Is everyone OK with this conclusion? Speak now or hold your peace until next big (AP) thing from Canon

[mmalik]

Here is sum total of Multiple LENR analysis thus far:

6D..............3-4 long exposures + 1 internal dark
5D MkIII......3-4 long exposures + 1 internal dark
5D MkII.......5 long exposures + 1 internal dark

[mmalik]

Relevant discussion regarding 6D's LENR tests here.... Regards

[mmalik]

I'll be testing 6D's LENR and Multiple Exposures features soon; in the meantime a look at Canon 6D Settings for Astrophotography below. Suggestions welcome to optimize 6D's LENR and Multiple Exposures Settings, focus of this discussion!

Note: Astrophotography settings for Canon 60Da can be found here...

[mmalik]

1.

Attached Thumbnails

• 

[mmalik]

2.

Attached Thumbnails

•