Pages: [1]   Go Down

Author Topic: Some observations on Focus Stacking  (Read 6856 times)

mdijb

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 666
    • mdiimaging.com
Some observations on Focus Stacking
« on: July 08, 2012, 08:32:10 pm »

I finished more testing on some flower arrangements coloseups and some still lifes in an antique store

The Landscapes and Antique store images were clearly better using CS5 compared to Helicon

However with the Flower macros the results were not clear and highly variable

There were parts of the flowers that were more sharp using each program with a bias towards better results using Helicon, but not consistently so.--Frankly it was a bit confusing to evaluate

Another Side note--Using manual focus as opposed to focus peaking using the NEX 7 appeared to give better results

Since Helicon and the other programs were designed for macro work, I suspect these program will work best on this subject matter, while CS5 is better for landscapes or other subjects where the distance between camera and subject is much larger.

If you have interest in this area, I welcome your comparison results and comments

MDIJB
Logged
mdiimaging.com

Tony Jay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2965
Re: Some observations on Focus Stacking
« Reply #1 on: July 08, 2012, 11:26:34 pm »

Thanks for your observations.

I have only ever used Helicon focus but your thoughts that Ps5 may be better for landscapes gives food for thought.
I might do a little experimentation myself this weekend.

BTW I do use the Helicon utility to plan and execute the focus stack and have also recently acquired the Promote control remote that performs a similar task.

Regards

Tony Jay
Logged

bill t.

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3011
    • http://www.unit16.net
Re: Some observations on Focus Stacking
« Reply #2 on: July 08, 2012, 11:39:16 pm »

Nope, Helicon kicks CS5 butt in the landscape stacking dept!  Every time.

Here are a couple quick examples.

It's Helicon on the left, CS5 on the right.

8 deep stack.

No matter what I did, or how I ordered the stack, CS5 ALWAYS picks the least focused version of the sky.  And note the issues along the CS5 skyline.  In every area, Helicon's stacking is hugely superior to CS5, and can be adjusted to best fit the subject matter.

To be fair, however, CS5's panorama blending is second to none.
« Last Edit: July 08, 2012, 11:41:22 pm by bill t. »
Logged

KenS

  • Full Member
  • ***
  • Offline Offline
  • Posts: 131
    • Spark of Light Photography
Re: Some observations on Focus Stacking
« Reply #3 on: July 09, 2012, 11:11:44 am »

Until recently I was quite happy with both focus blending and also pano blending using CS5 for landscape photography.  However, a few months ago I started to find landscape images that CS5 focus stacking failed to blend well. These were images made of just 3 shots, but the very near forground, near the bottom of the frame had regions out of focus.  Instead, I tried Helicon Focus, purchased it, no problems anymore.

Yesterday, for the first time I found a two shot pano, made with a the Canon TS-E 17 mm lens that CS5 failed to combine (the CS5 message said it could not create a successful pano).  The shots were made on an Induro PHQ-3 pano head, the TS-E was tilted for DOF and shifted down perhaps 6 mm, but the camera itself was also pitched down (making it a more difficult case).  There was considerable overlap between the two frames.  I tried the various CS5 perspective choices, some of  which has always woked in the past.  Since CS5 failed, I decided to download the newest trial version of PTgui (9.1.3), and it worked, producing a blend I could use to produce a final image.

So, based on my experience so far, it appears that while CS5 will work much of the time, the specialized tools (Helicon Focus and PTgui) do have there place.

Ken

bill t.

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3011
    • http://www.unit16.net
Re: Some observations on Focus Stacking
« Reply #4 on: July 09, 2012, 12:14:18 pm »

PTGui is a great stitcher, no question about it.  As mentioned CS5 stitching still has trouble aligning images when the camera is tilted away from the horizontal, and with very wide lenses.  I have never failed to get a good PTGui stitch on the most difficult panos, even when I didn't shoot a technically good pano set.

But for blending the already stitched image, I often prefer CS5.  It has a sharp line blending algorithm that wiggles and winds its way between adjacent images to find the best possible match between adjacent colors and densities.  As long as I have taken the time to balance adjoining images fairly well, it gives me a crisp join with no mushy "averaged out" overlaps that I sometimes see in other blending algorithms.  But OTOH ever since version 8.x, PTGui's own native blending has worked very nicely as well.

FWIW the workflow pipeline between LR and PS is so easy that I often choose to stitch with CS5 for the less challenging panos it can handle well.  A few quick trial pano iterations via "Photo->Edit In->Merge to Panorama in Photoshop" with the CS5 "blend" option un-checked quickly shows me which images still need some work in LR.
Logged

Alan Smallbone

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 788
    • APS Photography
Re: Some observations on Focus Stacking
« Reply #5 on: July 09, 2012, 03:36:14 pm »

As a side note you might want to also try Zerene Stacker, it has some slightly different options to Helicon and in my opinion handles bokeh in the background better and the editing tools are easier, in my opinion, than Helicon. I have used them both and ended up buying Zerene, the guy behind it was not happy with Helicon and made his own focus stacker.

The Promote Controller works great for stacking. If you have an Android tablet, there is utility called DSLR controller, that has similar features to the promote only a "live view". I use it with a Motorola Xoom and seeing the live view on a 10" screen is really nice and you are able to set the starting point and ending points a little more accurately than you can with the live view on the back of the camera. The app for Android is only $9 which is a whole lot cheaper than the app that Helicon is making.

Just some alternatives.

Alan
Logged
Alan Smallbone
Orange County, CA

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Some observations on Focus Stacking
« Reply #6 on: July 09, 2012, 03:49:58 pm »

Hi,

Thanks for sharing!

Best regards
Erik

As a side note you might want to also try Zerene Stacker, it has some slightly different options to Helicon and in my opinion handles bokeh in the background better and the editing tools are easier, in my opinion, than Helicon. I have used them both and ended up buying Zerene, the guy behind it was not happy with Helicon and made his own focus stacker.

The Promote Controller works great for stacking. If you have an Android tablet, there is utility called DSLR controller, that has similar features to the promote only a "live view". I use it with a Motorola Xoom and seeing the live view on a 10" screen is really nice and you are able to set the starting point and ending points a little more accurately than you can with the live view on the back of the camera. The app for Android is only $9 which is a whole lot cheaper than the app that Helicon is making.

Just some alternatives.

Alan
Logged
Erik Kaffehr
 

mdijb

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 666
    • mdiimaging.com
Re: Some observations on Focus Stacking
« Reply #7 on: July 09, 2012, 07:33:33 pm »

Funny--I got the opposite result when I did my comparison.

What settings do you use for Helicon?

I tried some different setting but did not see much difference in the results

Additional question--Can you used Helicon remote to create the stack and then save the images and process in another program, CS5 for example?

MDIJB
Logged
mdiimaging.com

Tony Jay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2965
Re: Some observations on Focus Stacking
« Reply #8 on: July 09, 2012, 09:32:46 pm »

Additional question--Can you used Helicon remote to create the stack and then save the images and process in another program, CS5 for example?

Absolutely!

Regards

Tony Jay
Logged

bill t.

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3011
    • http://www.unit16.net
Re: Some observations on Focus Stacking
« Reply #9 on: July 10, 2012, 12:42:18 am »

For focus stacks of technically good DSLR images, do some tests beginning with Helicon Method "B" with a Radius of 4 and Smoothing of 2.  Zoom in to 100% on parts of the output image that would be made up of parts of at least 2 different focus planes.  Look at the quality of the edges, size of the mask ghost, etc.  Keep trying variations until you find something that works for you.  But B-4-2 works pretty good a lot of the time for highly detailed landscapes.

With stuff like this, a few tests are worth 1000 posts.

The attached image shows why I like Helicon Method B.  The three sections are, left to right, Helicon B, Helicon C, Helicon A.

Note that Helicon C is virtually identical to the Zerene PMax method.  I didn't grab an image from Zerene because of color management issues and because my trial period is expired so I couldn't save a color managed result.  But it looks the same as Helicon C.  Yes you get nice tight masks, and yes you can pay for it with weird artifacts elsewhere.

Note that Method B does a fab job of dealing with the weird artifacts created when Mr. Wind is blowing things around between exposures.  Note the pronounced presence of such artifacts in Methods C and A.   It's that easy.  Helicon Method B.  Don't leave home without it.

As far as getting better results from CS5 stacking, please post an example.




Logged
Pages: [1]   Go Up