Is Nikon intending to abstract the esence of focus stacking?

This area is for the discussion of what's new, what's on your mind, and general photographic topics. A place to meet, make comments on this site, and get the latest community news.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

ChrisR wrote:People will disagree on what is "significant" :).
Is this for competitive commercial use, or enjoyment?

If the photos are for web use, maximum perhaps 1600 pixels wide, then if you have a sensor 7379 wide, would it help to have 13% more?
Then would the lack of AA filter make a difference?

If you're doing microscopy, how many pixels can your optics use?
But then the silent shutter will help, in many setups.

Plus, there is a number of details we don't know about, afaik, such as how you can use flash, in all shutter modes.
Chris,

You bring up a good point about how the flash will work, and especially in silent shutter mode. Soon we should find out.

Best,

Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

All Ex,

I'll certainly provide the comparison results here when I have a chance to do the test.

Best,

Mike

All Ex
Posts: 261
Joined: Mon Jul 20, 2015 3:54 am
Location: Greece Thessaloniki

Post by All Ex »

I`m back on some other post, my stackshot controller is stacked in the costums, as soon as it arrives I`ll perform my stacks.

Talk to you in some other post as well,
All--Ex
My YouTube initial video

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Rik, All Ex,

Here's a couple small crops from a quick session with a chip to compare in camera and external generated TIFF files. The first Test Chip X is with the camera (D800E) generated TIFF (8 bit), the second is with LR converting D800E RAW into a TIFF (16 bit).

These were from a Vertical Stand with a Mitty 5X and Nikon 200 F4 tube lens. 3 Strobes (Adorama SK300II) were used, with diffusion provided by a foam cup. Step size was 20 microns and 347 images were used in the originals, but I reran with only 40 for these crops. Align was OFF in Zerene.

No processing was done before or after stacking in Zerene other than cropping and trying to get the images to fit here with PS.

Best,

Mike


Image

In camera generated TIFF (8 bit)


Image

LR generated TIFF (16 bit) from RAW

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

That looks like a good demonstration of why there's no such thing as "just converting" from raw to TIFF. Regardless of which image might be preferred, there's no doubt that they're different.

The in-camera TIFF has higher contrast and looks sharper, but has posterized highlights due to saturating the red channel. These are typical of differences in how the raw conversion parameters are set.

Also interesting, but of unknown significance, is that the two stacked outputs appear to have different geometry. As shown by the following animated comparison, the Lightroom stack appears to have been shot from a lower angle, with the chip surface tipped farther back and more of the side of the chip showing. (See lower left corner of image.)

Normally I would presume that this is due to some unexpected interaction between the alignment algorithm and differing appearances of the source images.

But in this case, because "Align was OFF in Zerene", it seems clear that computational alignment is not to blame.

Any ideas what is actually causing it?

Image

--Rik

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

rjlittlefield wrote:That looks like a good demonstration of why there's no such thing as "just converting" from raw to TIFF. Regardless of which image might be preferred, there's no doubt that they're different.

The in-camera TIFF has higher contrast and looks sharper, but has posterized highlights due to saturating the red channel. These are typical of differences in how the raw conversion parameters are set.

Also interesting, but of unknown significance, is that the two stacked outputs appear to have different geometry. As shown by the following animated comparison, the Lightroom stack appears to have been shot from a lower angle, with the chip surface tipped farther back and more of the side of the chip showing. (See lower left corner of image.)

Normally I would presume that this is due to some unexpected interaction between the alignment algorithm and differing appearances of the source images.

But in this case, because "Align was OFF in Zerene", it seems clear that computational alignment is not to blame.

Any ideas what is actually causing it?

Image

--Rik
Rik,

I noted this in the original stacks with Alignment ON, so I did a short rerun with it OFF and the slight shift was still there.

These two images were shot at significantly different times (~3hrs) and the chip was mounted to a nail with modeling clay. I suspect what happened is the clay warmed up and the chip angle dipped slightly. The images were 347 and about 6~7 seconds between images, so image capture time was quite long too.

Agree that the in camera 8 bit TIFF has higher contrast and looks sharper. Not sure these would make much difference after some post processing though. Maybe Nikon introduces some slight image manipulation in the TIFF file creation? The subject may not the best for this comparison, maybe others can do something similar with other subjects.

Anyway, you can see why I have been just allowing the camera to produce the TIFF files, rather than converting from RAW in LR.

Soon I may have to do the conversion from RAW in LR because my 64GB cards fill up quickly with the 110MB per image in camera TIFF file when I get slightly over 500 images per stacking session. I know I could just buy some 128 or 256GB cards!! With the projected D850 file sizes being ~150MB TIFF, this may come sooner than I expected!!

Best,

Mike

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Warm clay -- gotcha! Thanks for clearing that up.

--Rik

All Ex
Posts: 261
Joined: Mon Jul 20, 2015 3:54 am
Location: Greece Thessaloniki

Post by All Ex »

In my poor opinion, the in camera converted files are better even if we accept that the red channel highlights are clipped. It somehow looks more contrasty, it is clear, in my eyes that in camera conversion works better.
Mike, you gave me an idea with that portion of the ship, I hope that when my stack shot controller will "at last" leave the customs I'll be ready to perform a stack comparing my new rig escorted by my new Mitutoyo 5X & my also new JML 21mm.
In the beginning used with, as tube lens my Nikon 200mm MF Ai lens.

Thank`s for sharing your photos, Mike.

:)
All--Ex
My YouTube initial video

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

All Ex

Don't know what camera body you are using, but the old Nikon 200mm F4 "Q" type film lenses work well with the Mitty 5X and other Mitty's. It does vignette slightly on the FX full frame sensor, sometime I like that "look" though. No vignetting on DX sensor.

These are on eBay for very low prices, I got one for $20 and another for $30. The $20 ($12 + $8 shipping) one had the aperture blades stuck and both needed AI conversion. I followed Richard Haw's superb site and modified both for modern F mount, and eventually fixed the stuck aperture blades.

Edit. Don't know about your 200mm but the "Q" versions have a tendency to creep away from inf focus when used vertically with a heavy Mitty hanging off the front. I spent lots of time and wasted a bunch of stacking sessions until I found this out :oops:

Anyway, there's lots of low cost options for tube lenses with these old Nikon 200mm types, Vivitar (Komine version) & Zeiss 135mm and the Raynox 150 & 250 (208mm & 125mm). All seem to work really well thanks to the prior works of Rik and Lou evaluating such and letting us know about these bargains!!

Best,

Mike

All Ex
Posts: 261
Joined: Mon Jul 20, 2015 3:54 am
Location: Greece Thessaloniki

Post by All Ex »

Mike,
I`ve spotted the vignetting in my Nikon Ai 200mm lens & the Ais 135mm one I have, with my other Nikon objectives.
I`ve already constructed two other tubes based on the 250 DCR Raynox and the 150mm Rodenstock-apo Gerogon that doesn't vignette so.
I`m using a D800 and I deal with the little blurriness it produces with a hint of sharpening inside Photoshop.
Never quitting the idea to get a D800E (in a reasonable shutter count) if I find it in a good enough prize and if I manage to afford it.
All--Ex
My YouTube initial video

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

All Ex wrote:In my poor opinion, the in camera converted files are better even if we accept that the red channel highlights are clipped. It somehow looks more contrasty, it is clear, in my eyes that in camera conversion works better.
Of course what you're really saying is that you prefer the in-camera conversion over Lightroom's conversion with whatever settings mawyatt happened to use.

If you don't pay attention to raw conversion, then it's probably a bad idea to shoot raw.

--Rik

All Ex
Posts: 261
Joined: Mon Jul 20, 2015 3:54 am
Location: Greece Thessaloniki

Post by All Ex »

In that point, it is clear that we should use tiffs.
You suggest to shoot raw and then convert them to tiff ones in an external program (I have Photoshop and don`t Lightroom)
I understand that clipping is something outside the borders, but will it be the same in the case of 14 bit raw or something like that?
Last edited by All Ex on Sat Aug 26, 2017 10:52 am, edited 2 times in total.
All--Ex
My YouTube initial video

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Rik,

The TIFFs were created in LR in sRGB Color Space and 16 bit channels, no other processing was done to these files in LR (that I know about?). I just imported the RAW files from the cards and exported them as TIFFs, nothing else.

The in camera (D800E) TIFF files have nothing done to them (that I know about?) other than the specific card they are using. The camera is set to sRGB Color Space, Custom WB and Vivid Picture Control. ISO is L1.0 (lowest ISO on D800E) and shutter is 1/200. Same setting for TIFF and RAW image captures.

Would be nice if the camera could create the RAW and TIFF files simultaneously like it does with the RAW and JPEGs.

Best,

Mike
Last edited by mawyatt on Sat Aug 26, 2017 10:54 am, edited 1 time in total.

billjanes1
Posts: 91
Joined: Fri Dec 30, 2016 1:59 pm
Location: Lake Forest, IL, USA

Post by billjanes1 »

ChrisR wrote:It's not a camera issue, but no, ZS doesn't accept Raw. Helicon sort-of does, but wouldn't you prefer to use your favourite Raw >Tiff processing program?

Stacking time in ZS depends on MP, not file size. My new fast-ish but not unusual PC should stack PMax at about 12-15 frames of 45MP per minute. That's OK by me.
I find it convenient to export 16 bit TIFs to Zerene from Lightroom. The TIFs are stored in a temporary directory and there is no need to save them, since they can be easily replicated from LR. TIFs from my D800e are over 100 MB so storing 100 36 MP TIF images would take up quite a bit disk space.

100 image stacks from my D800e are reasonably handled by Zerene on my 2 year old middling windows machine, so I should be able to handle D850 images when I get that camera.

Bill

All Ex
Posts: 261
Joined: Mon Jul 20, 2015 3:54 am
Location: Greece Thessaloniki

Post by All Ex »

Is it preferable to do the same procedure with 14-bit raws, will it have any visible difference?
I think that ZS has the option to transfer the files in a specific directory if the camera is tethered with its USB3 port.
All--Ex
My YouTube initial video

Post Reply Previous topicNext topic