Best Focus-Stacker and Why?

John Koerner wrote in post #12884043
Well, we most definitely are not in agreement here!

The way I look at it is, Time is Money, and the more time anyone has to spend "correcting" what the software should have done right to begin with, the less that person really needs the product. (Taken to the extreme, if you had to do everthing by hand, then you wouldn't need the product at all.)

By contrast, the more you are able to "push a button" and get everything you want to get done, done right, the more that product becomes a useful tool that helps you leverage your time, not waste your time cleaning-up the product's botch-job.

This basic goal of a stacking program is exactly why I felt Adobe CS5 was the superior product ... because it did the job that the other products should have done ... so I don't have to waste my time doing the clean-up work for them. I mean, just look at my images on the blog post and imagine the time it would have taken to correct the botch-jobs of the other products ... no thanks!

John,

No, on the contrary, we do agree... I mean I agree with your statement.

Remember that I shoot handheld, so my frames do not really allign perfectly in the first place! Sometimes the angle is slightly off, perspective is changed, things like that. So I have to do a lot of retouching. I am a victim of my own shooting technique, not the program I use.

I totally agree with you, that in your case, since the shots were from a tripod, they should have all perfectly alligned and there should be no need or a very minimal need of retouching.

I had noticed a while back that Zerene changes colours of certain shots because I was trying to stack two shots, not for focus, but to use the background of one that was more properly exposed. I know this can be done in PS, but I work very easily in Zerene (and sadly have not a whole lot of PS skills), so I thought this could be a good idea. It was not, it did not work at all. Of course I realized that I was trying to use a program for something it was not designed to do, so I did not think much of it at the time.

Martin

Martin G. wrote in post #12884294
John,
No, on the contrary, we do agree... I mean I agree with your statement.

Exactly Martin, the tools that we spend our money on should save us time (and, better yet, leverage our time), not cost us time.

Martin G. wrote in post #12884294
Remember that I shoot handheld, so my frames do not really allign perfectly in the first place! Sometimes the angle is slightly off, perspective is changed, things like that. So I have to do a lot of retouching. I am a victim of my own shooting technique, not the program I use.

Well, we have already filled several pages debating tripod use (LOL ), but even tripod use out in the field is not as "perfect" as it can sometimes be in the studio (which itself has has its own problems when dealing with live critters). For example, once again my girl took photos of me taking the photos of this Tall Ironweed, while we were driving down a long dirt road, which might help folks see the whole picture:

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

Martin G. wrote in post #12884294
John,
I totally agree with you, that in your case, since the shots were from a tripod, they should have all perfectly alligned and there should be no need or a very minimal need of retouching.

Exactly, but think about it like this: if the Zerene program made me have to re-touch so much photographing off of a tripod ... I can only imagine the kind of work it is forcing you to do, hand-holding

Or, conversely, if Adobe CS5 worked so well with such a multi-tipped difficult image as what I selected, imagine how much time it would save you in post-process as well ...

Martin G. wrote in post #12884294
John,
I had noticed a while back that Zerene changes colours of certain shots because I was trying to stack two shots, not for focus, but to use the background of one that was more properly exposed. I know this can be done in PS, but I work very easily in Zerene (and sadly have not a whole lot of PS skills), so I thought this could be a good idea. It was not, it did not work at all. Of course I realized that I was trying to use a program for something it was not designed to do, so I did not think much of it at the time.
Martin.

Interesting.

I do not have enough experience stacking images to speak with authority on the subject, but it occurs to me that "stacking" something like Fly's Eyes would be easier for a stacking program to process (because of the relatively-even geometrical patterns of their eye shapes) ... than it would be for such programs to render random, "curling tip" images such as what I chose for my experiment.

And then when you start talking bokeh ... which are complex swirls of smooth colors all blended delicately and perfectly by nature ... a program's ability to handle this kind of smooth, graduated color complexity properly through a multiple image stack is critical. Processing a bug on a green leaf is one thing, but processing a multi-tendriled flower swimming in a bokeh of ever-shifting "light-mist" is quite another.

Well, when I worked with and compared the programs, in just looking at the Zerene interface, and especially through a color-calibrated monitor, it couldn't even display the color of my images properly on a color-calibrated screen. (I did read the provisio that the program displayed only sRGB by default, but it was quite a shock to be used to seeing all of my images in vivid color in Photoshop, only to see them displayed as grainy and "off" as I worked with them on the Zerene Stacker. And, even though they "claimed" the photos would be rendered "normally" on output, despite the sRGB viewing limitations, when compared to the Adobe CS5-rendered images, the Zerene-rendered images most assuredly were not.)

That is one thing I have to say on behalf of the Helicon-Focus program, BTW, is that its user interface was far better than the Zerene's "minimalist" interface. The Helicon program "displayed" the full colors accurately on my monitor, and was accented by a dark gray trim like Lightroom 3 (which is ideal for photography), so it really was nicer to work with while loading images ... while the Zerene program was like using a no-frills trial beta program by comparison.

In the end, color is where I think Adobe really offers more to its image-rendering than either of these other companies, and that is because Adobe has world-class color management capability. These other programs just dropped the ball entirely on the accurate color/bokeh handling compared to Adobe, at least in my experience, but what was really surprising is they likewise fell short even in the "stacking accuracy" of their namesake "focusing" ability!

I think just for fun I will run a few more experiments and see what happens ... with some more particularly tough images (perhaps some colorful spiders ... taken in optimal natural light), to make sure there is a full complexity of color ranges and "random legs" that these programs have to "align right and render" ... so we will see what happens

Cheers!

Jack


.

John,

I guess I would need to try PS to see.

If you have time, one thing I would like to see is how PS handles noise in stacks. Zerene "stacks" the noise. I take most of my shots at ISO 800 and while with proper control of exposure, it is not a problem for one single image, say you stack 6 images, you end up with 6 times the noise. So, what I always do is retouch the whole background to use only a single frame and avoid extra noise.

I would be curious to see what happens with an "unretouched" stack in PS as far as the background goes.

Martin

Just a quick comment on focus stacking. There really is not a focus stacking programme out there that will magically stack your shots without any problems. The best one for you is the one that stacks the type of shots you take with the least problems.
Interestingly the guy here http://www.macrostop.c​om/ used to use CS5 for stacking but then started using zerene stacker for a number of reasons.
I tend to use zerene stacker for my own handheld stacks but freely admit i have never tried CS5.
Most of the people that do very extensive stacks (50 or more high mag shots) find zerene produces less artifacts than helicon. I gather from comments others have made that CS5 tends to start getting rather slow with stacks of more than about 10 shots.

Brian v.

Brian, I appreciate your comments and you may indeed be correct about PS being overwhelmed with the bulked-up files. But could it be due to inadequate computing power...how much RAM do you need for those big, multi-image jobs using zerene?
Your link provides some interesting reading...always good to get other opinions.

chauncey wrote in post #12886969
Brian, I appreciate your comments and you may indeed be correct about PS being overwhelmed with the bulked-up files. But could it be due to inadequate computing power...how much RAM do you need for those big, multi-image jobs using zerene?
Your link provides some interesting reading...always good to get other opinions.

Well I've only got 2gb ram with what is now a fairly old winxp dual core system and have not had any problems stacking upto 20 shots with zerene. I do however find the repair tool a bit clunky to use in zerene which may be due to memory.

As I tried to imply in the original post none of the stacking progs are foolproof and whilst for me zerene seems to stack the best out of helicon, zerene and the combine series, there are still times when I end up stacking by hand PS.
Brian v.

Geez Brian, you really gotta consider upgrading...but what's impressive is that you can cough out the images that you do with that...ah, antiquated, old stuff.

Martin G. wrote in post #12885447
John,
I guess I would need to try PS to see.
If you have time, one thing I would like to see is how PS handles noise in stacks. Zerene "stacks" the noise. I take most of my shots at ISO 800 and while with proper control of exposure, it is not a problem for one single image, say you stack 6 images, you end up with 6 times the noise. So, what I always do is retouch the whole background to use only a single frame and avoid extra noise.
I would be curious to see what happens with an "unretouched" stack in PS as far as the background goes.
Martin

I have stacked a few ISO 800 files with PS, and have never noticed a change of background, especially not like Helicon-Focus, which destroyed my background.

Zerene also did not handle the colors as well as PS, so I would image they wouldn't handle noise as well either, but I will test it and see first part of next week.

Jack

PS: Heard back from Rik, and he agreed his program did not handle that stack as well as PS, and he asked me to forward my original files to see if he could personally "map" the image and try to do better himself. If he succeeds in making the image successful, he plans on using my image as a tutorial on "what to do" for problem "ends" shots like that ... and if he doesn't succeed, then he needs to make his program better, and plans to try. As always, he went above and beyond the call of duty to be helpful, and views this as a learning experience for himself as well.

LordV wrote in post #12886045
Just a quick comment on focus stacking. There really is not a focus stacking programme out there that will magically stack your shots without any problems. The best one for you is the one that stacks the type of shots you take with the least problems.

Well, then for me Photoshop has proven to be "the best one," by your own definition (as well as my own).

No one is requesting "magic," but the entire idea of a stacking program is to get it aligned properly, without losing quality, with the click of a button.

The more work you are forced to do after you "click," the worse the program is doing its job ... while the less work you are forced to do after you "click," the better the program is doing its job.

LordV wrote in post #12886045
Interestingly the guy here http://www.macrostop.c​om/ used to use CS5 for stacking but then started using zerene stacker for a number of reasons.
I tend to use zerene stacker for my own handheld stacks but freely admit i have never tried CS5.

I am going to try Zerene next to Photoshop and Helicon for 10 straight tough images, and let the chips fall where they may ...

LordV wrote in post #12886045
Most of the people that do very extensive stacks (50 or more high mag shots) find zerene produces less artifacts than helicon. I gather from comments others have made that CS5 tends to start getting rather slow with stacks of more than about 10 shots.
Brian v.

I can imagine that Zerene works better than Helicon, but as far as Adobe Photoshop being "slow" is concerned, 3 things come to mind:

1. Many people do not have the system resources to process that many images at once properly, so what they call "slow" is really their own system's inadequacy to handle the job;

2. Because Adobe is actually retaining all of the color detail better, this is going to take a whole lot more time to process than a program that is "dropping" the color tones and detail. If anyone is processing 30 images and his computer does not slow down a bit, then he either has one helluva fast system, or his images are not being processed properly and with ALL of the detail/colors being processed with true precision and accuracy;

3. I guess it all depends on what a person's definition of "slow" is. The Zerene stacker took about 10-15 seconds to process my 5 detailed images ... and it would have taken me about 4 hours to cut every single artifact out of that image, to clean up the blurs, plus adjust the colors, etc. Meanwhile, the Adobe program took maybe 2 minutes to process, and I might have had 2-5 more minutes of processing to do. So yes, maybe the Zerene product is spitting-out a "result" faster after you click on the "process" button ... because it doesn't really process all of the details as precisely ... but if you have to spend a couple of hours doctoring those "results" can you honestly say you're "saving time?"

To me, having to wait a couple of minutes longer for the Adobe program to process is a small price to pay for saving me a nightmare's worth of "corrections" after it's finished ... so, again, it's all the time that it takes to process the image that needs to be taken into account, not just how long it takes after you click the button.

Jack


.

Got to be the human Brain for the algorithm and any editor with layers.

Another thing to consider is the manner in which you take your individual images...if you're moving your camera, either using a monopod or by using focusing rails, you are altering the perspective, albeit minutely, which might help to explain why the various soft-wares will cough out imperfections.
I might suggest that one might be better served by using a tripod and, by using 10X zoomed LV, focus on various areas of the image and using those images as a basis for stacking...no altering of perspective.

I did make some tests one time Zerenne, versus PS CS5 with 100 shots. I also tried Helicon, but bought Zerenne as it is supposed to be US made.
I got something I can use with Zerenne, with CS5 I got a big blurry image.
So now I use CS5 for up to 5/10 images, otherwise I use Zerenne.
All the tests were made using an heavy setup and the advance of the subject was made using micrometeers.
My 2 cents...

chauncey wrote in post #12891661
Another thing to consider is the manner in which you take your individual images...if you're moving your camera, either using a monopod or by using focusing rails, you are altering the perspective, albeit minutely, which might help to explain why the various soft-wares will cough out imperfections.
I might suggest that one might be better served by using a tripod and, by using 10X zoomed LV, focus on various areas of the image and using those images as a basis for stacking...no altering of perspective.

Excellent point.

By reading the literature, stacks are best made when performed incrementally ... from front-to-back ... rather than being made randomly (a front shot, a side shot, a back shot, another side shot, etc.).

Gradually shooting your focus efforts from front-to-back will help the program "see" what you're trying to do better than randomly picking "spots" on the subject to focus, with no rhyme or reason to the sequence...

Jack


.

Well, I ran a much more intensive and exhaustive test on these 3 programs again, and once again the Adobe CS5 Extended prevailed every time.

I ran tests on simple 2-stack images on basic images, to multiple stacks on complex images, and Adobe CS5 simply shined compared to the other stacking programs IMO.


((( Full and Extremely Detailed Report Here )))


Jack


.

John,

Have you been stacking with Dmap in Zerene? I could swear that is what you have been doing judging by your results.

Seriously, I never obtained such horrible pictures in Zerene as the ones you have shown in your tests, except when I had tried Dmap, they look like typical Dmap stacks. I always stack in Dmax since I never understood how to work out Dmap (and I seem not to be the only one) and it always produced images like the ones you have shown, seems like only Rik understands the obscure science of Dmap! lol That is why I stick to Dmax.

Can you enlighten me?

Martin