Focus stacking from a video

Here you can discuss topics such as focus stacking, stitching and other techniques that relate to the processing of micrographs.
Message
Author
User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Focus stacking from a video

#1 Post by fdupre » Mon Sep 04, 2023 4:49 pm

Hello everyone,

I am sharing with you a technical test of focus stacking of a butterfly (a copper moth) based on a video taken with a Lumix G5 in HD at 50 frames/s. I converted the film from the native Lumix .MTS format (HD at 50 frames/s) to .mp4 format (HD at 24 frames/s) PI: the video lasts 57 seconds and weighs 108 MB in .mp4 and 204 MB in .MTS native Lumix video.
I then opened the .mp4 movie in Helicon and got 1381 frames !!! which I then stacked with method C and smoothing 1.

The shooting :
pdv800px.jpg
pdv800px.jpg (143.28 KiB) Viewed 19393 times
The result :
papillon_lumix_film-MH.jpg
papillon_lumix_film-MH.jpg (263.26 KiB) Viewed 19393 times
The same butterfly taken the "old fashioned way" with a 5DmkII and zedified from 399 photos taken one by one (the shooting session was long long long ;) )
papillon_face_70-200_lomo3.7MH.jpg
papillon_face_70-200_lomo3.7MH.jpg (253.71 KiB) Viewed 19393 times
I lack resolution in HD but the method is very interesting and the result is honorable in my opinion (I still had to add more sharpness than usual in Photoshop*); not to mention that I definitely need to familiarize myself with this method. I'm looking at the price of a camera that does 4k or even 8K...but it seems very expensive!

I also plan to use a small motor (astro map motor for telescope) to do the different focusing because it still shook quite a bit when I turned the micrometric wheel but I found that Helicon realigned the photos really well but a Cropping of around 15% of the total area was still necessary because of realignment artifacts all around the photo.

* Regarding the loss of sharpness, I think it may be due to the compression I used when transforming the native Lumix video to mp4 format.
I use OpenShot Video Editor (https://www.openshot.org/) and here are the settings I used for mp4 conversion:
Parametres01.JPG
Parametres01.JPG (56.06 KiB) Viewed 19393 times
I kept the native .MTS file and would do further tests with other compression codecs.
If you have any information on the codec to use, I would be happy to take your valuable advice :)

Kind regards,
FRanck
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

User avatar
woyjwjl
Posts: 325
Joined: Tue Jan 04, 2022 1:55 pm
Location: Wuhan, China

Re: Focus stacking from a video

#2 Post by woyjwjl » Tue Sep 05, 2023 1:10 am

I use Panasonic GH4, you know, cameras from almost 10 years ago

4K, 30 frames, 10bit, you deserve it ;)
Micrographers from China, thanks to the forum for providing a platform for exchange

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#3 Post by fdupre » Tue Sep 05, 2023 1:34 pm

Hello everyone,

Thanks woyjwjl for the info.

I think I have solved the vibration problem when turning the micrometric dial with the addition of a motor, whether by taking photo by photo (moving the subject then shot, moving the subject then shot, etc.) or by taking a video. Here is a short video on YouTube (switch it to 2x speed otherwise it will be very monotonous :)): https://youtu.be/4lNegPG7SHU

It will also make it easier for me to take photo by photo sessions because I will no longer need to “move me” to turn the dial. I would use the remote control that you can see at the top of the video (in which there is the battery). A quick press is equivalent to a dimming of the dial (1/50th of a millimeter).

Well, now I still have to refine the tensioning of the belt and above all test the thing on a subject. . .

On the other hand, I started looking at 4k cameras; It's a bit pricey. For 8k, I'll wait a few decades for prices to drop ;)

Kind regards,
FRanck
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#4 Post by fdupre » Tue Sep 05, 2023 5:20 pm

I just did a video test with the engine and it's pretty conclusive but there are still small vibrations :(
Here is the result on YouTube without and with the engine:
(1st part of the video (28 seconds) without the engine and 2nd part (36 seconds) with the engine.)
https://youtu.be/od6jywsMOLs
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

MichaelG.
Posts: 4026
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Focus stacking from a video

#5 Post by MichaelG. » Tue Sep 05, 2023 5:30 pm

Excellent comparison, Franck
… you are doing interesting and informative work here

Thanks for sharing it.
MichaelG.
Too many 'projects'

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#6 Post by fdupre » Wed Sep 06, 2023 2:48 pm

Hello everyone,

Thanks for your feedback Michael.
Here is the result of the zedification from the video taken with the engine and with the same corrections as that from the video without the engine:

For this video with the engine, I converted the film to the native Lumix format in .mp4 with the same parameters.
(weight of video file in native Lumix format = 526 MB / weight of video file in .mp4 format = 259 MB)
3379 photos !!! extraites du film dans Helicon - Stacked with method C and smoothing to 1
3379 photos !!! extraites du film dans Helicon - Stacked with method C and smoothing to 1
papillon_lumix_filmMP4-MH.jpg (228.73 KiB) Viewed 19241 times
I don't see too much difference between the result without an engine and the one with an engine, apart from the fact that I had to crop less the original output from Helicon with the result with the engine (fewer artifacts around of the photo).



Here are the cropped and non-resampled results (one captured pixel equals one screen pixel) :

--- result of the zedification from the video without engine:
papillon_lumix_film_sm_non-reechan.jpg
papillon_lumix_film_sm_non-reechan.jpg (199.17 KiB) Viewed 19241 times
--- result of the zedification from the video with engine:
papillon_lumix_film_am_non-reechan.jpg
papillon_lumix_film_am_non-reechan.jpg (216.43 KiB) Viewed 19241 times
I still find that there is no real difference. The Helicon software compensated well for the vibrations of the video without a motor. On the other hand, we can clearly see the lack of result due to the HD resolution (1920x1080).
In my opinion, the only difference comes from the fact that it was necessary to crop more the photo from the video without engine because of the artifacts:
--- artifacts all around the raw photo without engine:
artefacts_sans_moteur.jpg
artefacts_sans_moteur.jpg (146.2 KiB) Viewed 19241 times
--- artifacts all around the raw photo with engine:
artefacts_avec_moteur.jpg
artefacts_avec_moteur.jpg (136.53 KiB) Viewed 19241 times
The really positive point of its tests is that the engine will serve me well for photo-by-photo shooting sessions.
I will no longer need to move around to turn the micrometric dial on the macro bench :) :) :) I will be able to stay in front of my screen and use the motor remote control : https://www.microbehunter.com/microscop ... 14&t=18887

Kind regards,
FRanck
Last edited by fdupre on Sun Oct 01, 2023 9:00 pm, edited 2 times in total.
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#7 Post by fdupre » Sat Sep 30, 2023 1:14 pm

Hello everyone,

@woyjwjl - I looked around the net and as you say, the Lumix GH4 is the best value for money for filming in 4k.
However, I have a question to ask you regarding the Lumix GH4:
Can we connect a Lumix GH4 to a PC with the possibility of seeing what the GH4 sees (live-view) as with the Canon and the EOS Utility II PC software ?

Sincerely,
FRanck
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

macnmotion
Posts: 548
Joined: Mon Sep 26, 2022 3:13 am

Re: Focus stacking from a video

#8 Post by macnmotion » Sat Sep 30, 2023 2:23 pm

The Lumix GH4 has HDMI out so you can connect it to an external monitor for live view without the need for a computer or software. You can then record internally. I can't comment on connectivity with a computer and software for viewing and recording, but even if you can with this camera, you would want to consider latency if you planned on using live view for focusing and tracking live organisms

Sure Squintsalot
Posts: 399
Joined: Mon May 16, 2022 3:44 pm

Re: Focus stacking from a video

#9 Post by Sure Squintsalot » Sat Sep 30, 2023 4:19 pm

FRank... a few questions and some observations:

1) Is focus stacking from video any easier, faster (or harder) than from stills?

2) Is it any better in terms of quality?

3) I see no real advantages, from here. Plus dealing with half a gig file size for a macro-shot seems absurd especially if the payoff is not inarguably better.

3) What will you do with these images? Print them wall-sized? Wallet sized? View them on a 4K monitor? Or look at them on a smart phone? I find that processing to a very high standard is sometimes a waste of time if I view everything on a mobile phone. It's hard to know where to draw the line.

4) Motor vibration is a real problem for macroscopy....imaging how much worse it is at 200x or even 400x! I am on the hunt for a low-RPM low vibration motor or system that will dampen vibration to near-zero that doesn't cost me an eye or a kidney. I understand worm-drives may be an acceptable option. Do you have any ideas?

Thanks for these tests. Head to head comparisons are tedious and a lot of work. I'm glad you're doing them!

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#10 Post by fdupre » Sun Oct 01, 2023 10:19 am

Hello everyone,

I hope that you all are ok.

@macnmotion - Thanks for the info. Indeed, you are right. Latency is an issue for videos of living organisms. I'm seriously considering buying a GH4 but it's 300/400 dollars ! . . .

@Sure Squintsalot - Thanks for your feedback - Your questions and observations are very interesting and very relevant!
I will allow myself to answer as best as I can while being aware that photography/imaging is a wonderfully interesting world (we learn every day) but also very complex because it is very subjective and very diverse.

1) Focus stacking from video is much faster. Making a video will take 1 to 2 minutes while taking 200 or 400 photos will take at least half an hour.
The "easy" thing about video is that you can't go wrong with the sharp points of the subject closest and furthest away without having to restart the shooting session.
All you need to do is start the video well before the first sharp point and stop it well after the last sharp point.
You work on the video afterwards in video processing software, removing the completely blurry shots from the beginning and end of the video.

On the other hand, the "problem" is that without a motor, you will have to turn the micrometric dial (or other manual system) to bring the subject closer to the camera or the camera to the subject without any hassle and with as little vibration as possible. With a motor, it's much simpler but there will always be more or less small vibrations but fortunately the stacking software will catch up without any problem.


2) Concerning only the video stacking part (I am not talking about the optical quality of the acquisition system which is also important of course), it is this question which is important in my opinion:

The resolution: In HD (1980x1020) or 2K (2048x1080), the resolution is not high enough in my opinion.
Not that I plan to make a 4x3 meter poster or a 120x160 cm poster of the final photo but with a resolution of "only" 2048x1080, we have much less possibility of managing the sharpness and fineness of the details of the image when digital development.
With my sharpness/fineness processing methodology, sub-resampling is one of the most effective ways to have good fineness of image details while having optimal sharpness and smoothness.

Example of a basic workflow for dealing with "artificial" sharpening on the final image (the result of Focus Stacking software) in PhotoShop (gain values can be increased, but be careful not to "sharpen" the image too much)
(this is only my way of working on sharpness/finesse/softness; there are certainly other equally effective methods)
:
- at the native resolution of the capture system: 1 emphasis filter such as:
accentuation50-2-0.jpg
accentuation50-2-0.jpg (48.34 KiB) Viewed 5910 times
- then a sub-resampling to 2000 pixels (sub-resampling of approximately 60%):
sous-reechanti60%.JPG
sous-reechanti60%.JPG (35.72 KiB) Viewed 5910 times
- and finally, 1 advanced sharpness filter:
net-optim-30-0.4.jpg
net-optim-30-0.4.jpg (99.87 KiB) Viewed 5910 times
As you may have understood, with this method of sharpness management, you need a resolution of at least 4K (3840x2160).
More resolution will give even more possibility of managing sharpness/fineness/softness of the image in the case, for example, of complex subjects or difficult to stack.

In short, the quality of the final rendering will mainly depend on the know-how, the lighting, the capture equipment and of course the quality of the optics used.
The manual focus stacking system, motorized or not, the method used, photo by photo or from a video, will remain important but will not be the only decider of the quality of the final image. As long as a reasonable distance between each photo is respected in relation to the macro enlargement and that the sharp points of the beginning and end subject are respected, all without too much vibration, then it will already be very good.

3) I prefer to have images with a minimum resolution of 1920 x 1080 because who can do more, can do less. This is my resolution limit as of today.
On the other hand, you are right, it is true that large storage disks are needed, especially since I keep all the stages of digital development of the photo once stacked.

4) You are absolutely right, vibrations are a big problem.
Regarding the cost of a turnkey or homemade manual motorized stacking system, I think you still have to pay the price. I mean it won't just cost you 20 dollars as you know.
I did a quick evaluation of my homemade macro bench (not counting the LED lighting, optics and cameras) and I arrived at a cost of around 400 dollars.

Sincerely,
FRanck

(Sorry for the Photoshop menus which are in French)
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#11 Post by bkt » Mon Mar 04, 2024 2:21 pm

Sure Squintsalot wrote:
Sat Sep 30, 2023 4:19 pm
1) Is focus stacking from video any easier, faster (or harder) than from stills?

2) Is it any better in terms of quality?

3) I see no real advantages...

4) Motor vibration is a real problem for macroscopy....imaging how much worse it is at 200x or even 400x!
I do focus stack also live beasts, using videos... using either 2 or 4k shots, between 60 and 30fps. The more frames you get, better can be the result. I often stack using several hundreds of frames, taken in 8 to 16 seconds... just taking care about the microscope stability whilst I turn the fine focus knob.

This is one of several videos I took from this beast, to dump frames, and to build a stack. As you see, it was often moving:



From one of those videos, a stack, using a 40x objective if I remember well. Picking a video from the set during which most movements were out of focus:

Image
fdupre wrote:
Sun Oct 01, 2023 10:19 am
1) Focus stacking from video is much faster. Making a video will take 1 to 2 minutes while taking 200 or 400 photos will take at least half an hour.
Take some videos, chose an acceptable one from the set, split and stack the frames...

That is a stack of 540 frames, due to the very large depth of field, a stack of a dead fly, using a 4x. I couldn't have done that in minutes using a DSLR or still frames:

Image

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#12 Post by fdupre » Thu Mar 07, 2024 5:16 pm

Hello bkt,
Really very interesting!
The result of the fly is really of very good quality.
Sincerely,
FRanck
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#13 Post by bkt » Thu Mar 07, 2024 8:34 pm

Hi fdupre,

thanks,

In the past, I had a look at what you are doing. Some were using a good 180 or 200mm focal, adding different microscope objectives in front of it, some russian oldtimers being of the preferred ones... By the way, those old optics went expensive.

I'm not sure you need a 4 or 8k cam. Using a stepper motor and a stage: keep the mirror up (live view mode), move one step or micro step forward, wait for hardware stabilization, take a shoot, repeat the process... a lot of shoots takes some time. At least, with the mirror up, that method saves the mirror mechanic. Several cameras now having no more mirrors. Taking so many shoots needing a port on the camera, for an external command synched with the stage movements.

I've not so good optics, so I don't care that much, and doing this manually, turning fine focus knob whilst concentrating.... This might be improved by a motor/stepper for shots at better intervals and positioning.

But hires videos seems to be a credible option.

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#14 Post by bkt » Thu Mar 07, 2024 9:03 pm

This being maybe an example of what people achieve with a vintage Lomo objective (not all are the same objectives), and maybe after hours of frames shoots: https://www.flickr.com/photos/radekvaltera/32474105550

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#15 Post by bkt » Thu Mar 07, 2024 9:39 pm

fdupre wrote:
Mon Sep 04, 2023 4:49 pm
I converted the film from the native Lumix .MTS format (HD at 50 frames/s) to .mp4 format (HD at 24 frames/s)
...
I then opened the .mp4 movie in Helicon and got 1381 frames !!! which I then stacked
Did you check what ffmpeg could dump from your original .MTS format?

With something like: ffmpeg -i file.mts -vcodec png -an output%05d.png

You could get images without recompression/degradation to .mp4 format...

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#16 Post by bkt » Fri Mar 08, 2024 1:47 pm

fdupre wrote:
Mon Sep 04, 2023 4:49 pm
taken with a Lumix G5 in HD at 50 frames/s.
...
I lack resolution in HD but the method is very interesting and the result is honorable in my opinion
...
The same butterfly taken the "old fashioned way"
...
I also plan to use a small motor (astro map motor for telescope) to do the different focusing
...
Regarding the loss of sharpness, I think it may be due to the compression I used when transforming the native Lumix video to mp4 format.
In what you show, I do prefer the old fashion way... Maybe use ffmpeg to split the MTS instead of converting it to mp4. You may get less alterations due to conversions?
fdupre wrote:
Tue Sep 05, 2023 1:34 pm
I think I have solved the vibration problem when turning the micrometric dial with the addition of a motor, whether by taking photo by photo (moving the subject then shot, moving the subject then shot, etc.) or by taking a video.
...
I started looking at 4k cameras; It's a bit pricey. For 8k, I'll wait a few decades for prices to drop
I had a quick look at your G5 specs. Its a 16mp, should provide nice detailed shoots, depending only on ISO, illumination and the optics you are using. It is mirror less, so you get no vibrations from there.

The G5 still has a mechanical shutter... But also a "silent mode", for an electronic shutter only, so no mechanical shutter vibrations? Plus a continuous shooting mode at 6FPS? Did you try this mode with your motorized stage, "silent mode" (should limit exposure times range...), continuously shooting, maybe slowing down the motor speed for more shoots?

If you can shoot permanently (RAW seems limited to 9 shoots, not jpeg?), you could already get your 4608 x 3456 or 4k frames, instead of full HD, 1,920 x 1,080.


User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#17 Post by fdupre » Fri Mar 08, 2024 3:45 pm

Hello bkt,
bkt wrote:
Thu Mar 07, 2024 8:34 pm
some russian oldtimers being of the preferred ones... By the way, those old optics went expensive.
It's true ! I have a Lomo microscope objective, the famous x3.7 / 0.11 and it is really very good indeed and really inexpensive compared to the quality of the renderings it gives.
And the Nikon BE Plan 10x 0.25, recommended by many and among other by ADi, is just fantastic especially if it is coupled with a 100 mm lens:
viewtopic.php?f=5&t=18880
viewtopic.php?p=138230#p138230


bkt wrote:
Thu Mar 07, 2024 9:39 pm
Did you check what ffmpeg could dump from your original .MTS format?
With something like: ffmpeg -i file.mts -vcodec png -an output%05d.png
You could get images without recompression/degradation to .mp4 format...
I'm going to check but it seems to me that I had done some tests not published on the forum because the difference between the native .mts format or uncompressed video formats and the .mp4 format was minimal if not absent.

Update: I found the results of my tests:
No compression (WEBM (vp9) lossless) - Video file size: 526 MB :
papillon_lumix_filmWEBM-MH.jpg
papillon_lumix_filmWEBM-MH.jpg (235.46 KiB) Viewed 2283 times
MP4 compression (h.264) - Video file size: 259 MB:
papillon_lumix_filmMP4-MH.jpg
papillon_lumix_filmMP4-MH.jpg (228.71 KiB) Viewed 2283 times

Apart from a slight difference in contrast, the fineness of the details seems identical to me.



bkt wrote:
Fri Mar 08, 2024 1:47 pm
In what you show, I do prefer the old fashion way...
Me too and notament in terms of the fineness of the details.


bkt wrote:
Fri Mar 08, 2024 1:47 pm
I had a quick look at your G5 specs. Its a 16mp, should provide nice detailed shoots, depending only on ISO, illumination and the optics you are using. It is mirror less, so you get no vibrations from there.

The G5 still has a mechanical shutter... But also a "silent mode", for an electronic shutter only, so no mechanical shutter vibrations? Plus a continuous shooting mode at 6FPS? Did you try this mode with your motorized stage, "silent mode" (should limit exposure times range...), continuously shooting, maybe slowing down the motor speed for more shoots?

If you can shoot permanently (RAW seems limited to 9 shoots, not jpeg?), you could already get your 4608 x 3456 or 4k frames, instead of full HD, 1,920 x 1,080.

Excellent !!! the continuous shooting mode at 6FPS on the G5.
I'm going to test this!
Thank you very much for your valuable feedback bkt.

Kind regards,
FRanck
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#18 Post by fdupre » Sat Mar 09, 2024 1:48 pm

Hello all,

Thanks to bkt's advice, i did some testing with the G5 in burst mode with various settings.
The two best results I obtained at 800 iso and 1/20 seconds speed in 16 MB jpeg format maximum quality are in slow speed burst mode (2 images seconds) and in medium speed burst mode (3.7 images seconds). Tests with other speeds at 6 frames per second (fast speed) and 20 frames per second (super fast speed) were not conclusive because the frame rate is well maintained for the first 3 or 4 seconds then the frame rate collapses and then is not constant ! :(

The SD memory card used at a transfer rate of 120 Mb/s which i thing explains the collapse of the frame rate from 6 frames per second and above but in any case, the shutter speed did not allow for go faster by 5 frames per second.

To conclude, I obtain in slow speed burst mode (2 images seconds) 233 images in 2 minutes and in medium speed burst mode (3.7 images seconds) 343 images in 2 minutes and above all all at constant images per second speed ( which is essential).

I am therefore satisfied because 343 images taken in 2 minutes suits me perfectly because as said bkt, I would only have to adjust the speed of the macro bench motor accordingly.

All I have to do is find an interesting subject and do some real-world testing on the macro bench.

For those who might be interested, a tutorial on the burst mode of the G5:
https://www.youtube.com/watch?v=9Ebg44cygME

Sincerely,
FRanck
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#19 Post by bkt » Sat Mar 09, 2024 7:40 pm

fdupre wrote:
Sat Mar 09, 2024 1:48 pm
at 800 iso and 1/20 seconds speed
You are welcome.

1/20 is really slow. This is related to aperture and light. That may not be enough to freeze stage movements and vibrations. To be checked.

I see you use video projectors with multiple LEDs to illuminate the scene, of a little object. You may have to replace them by some 3 or 5W single leds + diffusers, close to the subject. Shall allow to lower both ISO and exposure time.

If you are familiar with soldering (some LEDs + resistors), it wouldn't cost that much, maybe some $10 to 20 max.

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#20 Post by bkt » Sun Mar 10, 2024 12:28 pm

fdupre wrote:
Sat Mar 09, 2024 1:48 pm
800 iso and 1/20 seconds speed
At some point, you will make me buy a Lomo to test this also :D

For optimal quality, you will need ISO 200 or less, plus a speed of 1/500 or better, to freeze movements and vibrations.

I do not think this can be met with video projectors, their light seems to much spread to me. Neither with 1 to 5W power white LEDs only, as they seem all designed for a large 120° beam. I noticed several offers of cheap collimators, for resulting beams down to 30, 15, 10 or 5° (at such power or high concentrated luminosity, this could be dangerous for eyes, it would need extra attention...).

Building cheap spot lights using high power single LEDs and hot glue shouldn't be very complex. Could need a little heat sink behind the LEDs. I don't know if any exist on the market.

For eyes security, it would need two levels of illumination (can be done with resistors and a switch):
- a low eyes safe illumination level for spots positioning
- a high power illumination for shots, being away from the bench, newer looking at it, or better, being in next room during shots

Don't order such parts if you feel unconfident with security.

ColimatingLense.JPG
ColimatingLense.JPG (28.48 KiB) Viewed 2130 times

JWW
Posts: 152
Joined: Tue May 16, 2017 9:24 pm
Contact:

Re: Focus stacking from a video

#21 Post by JWW » Sun Mar 10, 2024 5:45 pm

FRanck,

Very nice. Is there something similar to the EOS Utility but for the Panasonic Lumix G5

-JW:

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#22 Post by bkt » Sun Mar 10, 2024 9:25 pm

JWW wrote:
Sun Mar 10, 2024 5:45 pm
FRanck,

Very nice. Is there something similar to the EOS Utility but for the Panasonic Lumix G5

-JW:
You mean if any Canon (or other brands) can do this? Continuous shooting? Probably yes. But as Franck added, this may be limited by the flash card speed, they tend to lower the write rate after some shoots (to prevent card over heating).

This Panasonic G5 having an advantage. As it allow to disable the mechanical shutter, it will prevent those vibrations. On that video, this Canon seems to produce mechanical noises, and sure would blur the images by the shutter produced vibrations:



That could be an advantage of taking 2 or 4k videos, if the camera supports this: no vibrations added by the shutter.

JWW
Posts: 152
Joined: Tue May 16, 2017 9:24 pm
Contact:

Re: Focus stacking from a video

#23 Post by JWW » Mon Mar 11, 2024 4:36 pm

bkt wrote:
Sun Mar 10, 2024 9:25 pm
JWW wrote:
Sun Mar 10, 2024 5:45 pm
FRanck,

Very nice. Is there something similar to the EOS Utility but for the Panasonic Lumix G5

-JW:
You mean if any Canon (or other brands) can do this? Continuous shooting?
No, I was referring to one of your links where you were using the Canon EOS Utility on your Canon camera. I was wondering if there was any similar tethering software for Panasonic, such as the EOS Utility. I really like the EOS Utility.

-JW:

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#24 Post by fdupre » Mon Mar 11, 2024 6:40 pm

JWW wrote:
Sun Mar 10, 2024 5:45 pm
FRanck,

Very nice. Is there something similar to the EOS Utility but for the Panasonic Lumix G5

-JW:
Hello all,

JWW : Thanks JWW :)
I also like the Canon EOS Utility software.
Concerning the Panasonic G5, to my knowledge, unfortunately there is no software allowing you to control it from a PC.
Here is the "LUMIX Tether" software offered by Panasonic to control APNs via a PC but the G5 is not compatible (see the list) :(
https://av.jpn.support.panasonic.com/su ... ether.html

Regarding the vibrations due to the mechanical shutter, frankly, I did tests with the G5 and a Canon 1200D and I did not see any really significant blur on the Canon 1200D compared to the G5 (at resolution identical).
I even read (I'll have to find the article)that with live view mode and the Canon 1200D (and certain other Canon cameras), which happens when you use the Canon EOS Utility software as in the example below:
poste_travailANG.jpg
poste_travailANG.jpg (174.82 KiB) Viewed 1996 times
the vibrations were very very attenuated or even non-existent.

I must "admit" that I did not push my tests with different exposure times / ISO, etc.
I think that in fact, the mechanical shutter must certainly cause vibrations and that is logical, but hey, it didn't bother me that much. Maybe this is due to my way of managing sharpness (see my post above: viewtopic.php?p=138483#p138483

Kind regards,
FRanck
Last edited by fdupre on Tue Mar 12, 2024 10:32 am, edited 2 times in total.
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#25 Post by bkt » Mon Mar 11, 2024 9:10 pm

@JWW maybe have a look at what is available on Android or IOS. I've a now old Pana LX100 that can be managed via wifi, a now old Nikon manageable via USB. Check also third party apps in stores, which could do more... This could be used for minimal cameras controls, and for remote control/shootings, for less vibrations.

Shooting at low speeds was a real problem with the mirrors of cameras, moving up/down - clac clac, the noise, and up to high vibrations due to the mechanic. It is maybe less an issue with the shutter moving only. For low speeds with DSLR, it needed to raise the mirror, then waiting some time for the hardware (tripod, telescope, macro rail, ..) to settle, or needed to use Live View mode (mirror always up), then only the images could be of the best quality. Or a solution was to use a timer before shoot (pressing the button usually makes the hardware move), plus very low speeds/long exposures, for averaging (this is still useful for night shots using a tripod; or for averaging pixels noise using lower ISO settings).

According to what one could expect from a microscope also, mirrors and shutters vibration could more or less blur or alter the image quality. Depending on the resolution.

At some point, the limitation is in the optical system. Then it isn't needed to tweak much more at the camera side, which would then only accurately reproduce what the system/optics achieve.

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#26 Post by fdupre » Mon Mar 11, 2024 9:43 pm

Your explanations are very interesting and well explained. I completely agree with you bkt. :)
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#27 Post by bkt » Tue Mar 12, 2024 10:08 am

I ordered some parts, I may do tests in the coming weeks, with powerful 3W LEDs and collimators. The point with them: never look at such a luminous light sources.

According to the optics and what could be expected from them, a combination taking videos should ease the process:
- using the fast video stacking solution to tweak the image, alignments, positioning, illumination, background, colors... takes some minutes per stack only
- once it is satisfying, switching to the old and slow method, using pictures, maybe with steps, ISO 100 and long exposures, adapted to the light sources

Not all publish high resolution images. The some I've seen for Lomo 3.7 being of very good quality.

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#28 Post by bkt » Wed Mar 13, 2024 8:46 pm

Hi, I noticed this exists. Could allow much better than ISO 800 1/20. 6W (3+3?) 6500k spot lights, don't know how safe this is (either looking into it or by reflections)...
6W6500k.JPG
6W6500k.JPG (29.31 KiB) Viewed 1778 times

User avatar
fdupre
Posts: 282
Joined: Fri Jan 21, 2022 10:54 am
Location: Europe / France / Paris

Re: Focus stacking from a video

#29 Post by fdupre » Fri Mar 15, 2024 8:36 am

do you have a link?
Equipment : binocular magnifier Bresser Biorit ICD x20 x40 / Microscope "Bresser" science infinity / Home macro bench
YouTube channel : https://www.youtube.com/channel/UCTbdg7 ... J_VpBU54YQ
first name : Franck

bkt
Posts: 103
Joined: Tue Feb 06, 2024 4:11 pm

Re: Focus stacking from a video

#30 Post by bkt » Fri Mar 15, 2024 10:08 am

fdupre wrote:
Fri Mar 15, 2024 8:36 am
do you have a link?
For the LEDs? You will find some offers searching for: 6w gooseneck for microscope

Post Reply