An image comparison of Zerene Stacker and Helicon Focus

[rjlittlefield]

Every once in a while somebody asks for a comparison of Zerene Stacker and Helicon Focus. Usually I answer with something like this:

In favor of Zerene Stacker, it gives cleaner images (less halo & stacking mush), especially for deep high magnification stacks. If you are interested in stereo/rocking 3D views, then another difference is that Zerene's method works well for all subjects, where Helicon's method only works well with smooth opaque surfaces.

While the above is technically accurate, I'm not sure that it communicates very well. So I decided to stop being wishy washy and just show straight up what I'm talking about.

Regarding halos, here's the same stack processed through ZS PMax and HF Methods A and B. These are unretouched and unmodified, processed with default parameters in all cases, using current versions of both programs. This is the head of a fruit fly, shot through a 10X microscope objective, HERE.




Regarding stacking mush, here's the same stack processed through ZS PMax and HF Methods A and B. These are unretouched and unmodified, processed with default parameters in all cases. Subject is the papillate stigma of a maple flower, HERE.



Here's an example with issues in both stacking mush and halos. It's a pair of dandelion seeds, shown in their entirety HERE. Processed through ZS DMap and HF Methods A and B, unretouched and unmodified, processed with default parameters. (The undersides of both seeds and the OOF background between them are good places to consider.)




Regarding stereo, here's the same stack processed through ZS PMax and HF Method B (the only that supports stereo output). Again unretouched and unmodified, processed with default parameters, except that for HF I've selected the highest quality surface. Subject is a hobo spider, HERE.



If you don't immediately see the difference in these pairs, then consider the bristles on the front legs. In the HF image, some of the big bristles are simply missing, and the rest of them look like they've been greased back against the cuticle or are streaming off the legs and pointing back toward the balsa block. This is a fundamental characteristic of the 3D surface approach used by HF. If your subject really does have a smooth surface with no overlaps or discontinuities, then the HF approach works great. If the subject does have overlaps or discontinuities, then the 3D surface approach does what you see here.

I have the usual conflict of interest in this post. Because I wrote Zerene Stacker, I imagine the reader will presume that I've somehow fudged the demo to make ZS look good. I mean, the comparisons are pretty striking, at least to my eyes. But nope, no fudging -- everything is exactly as stated.

What I have done, I frankly confess, is to deliberately choose subjects that illustrate what happens outside the range where Helicon does well.

This approach goes directly counter to reviewers' common practice of choosing some easy subject that every software handles well, and then structuring the review as "well, they all do great work, so let's talk about other stuff".

The reason I choose to make the comparison like this is that these are the sorts of subjects and issues that I have faced from Day 1 in my own photography.

I've used Helicon Focus since 2004 (before it even had image alignment, let alone retouching). For a long time I recommended it as "best of breed". It's a great tool for a lot of purposes.

However, the roots of Helicon Focus appear to be firmly planted in what to me are "shallow low magnification stacks", with subject field widths of a few inches and larger. When used with deeper stacks at higher magnification, it develops the issues shown above. At the same time, the design of its retouching capability becomes less effective because depth navigation is slow and all you can do is clone from source images to a single output image.

It was exactly those considerations that prompted me, in 2008, to begin development of a new tool designed from a clean start to focus on the stuff I cared about: deep high magnification stacks.

Zerene Stacker is the result. It's not perfect, but for what I do it's a big improvement. Your mileage may vary.

Hope this helps. At least now you'll know for sure what I'm talking about.

--Rik

[Pau]

Thanks Rik for the comparation.
I do very few stacks, so I'm not at all an experienced user. I tried trial versions of both Zerene and Helicon and in quick and simple tests the difference was even higher that in your exemples here: HF didn't seem useful for microscope work.

But, What about Combine ZP?. Apart of its big economical advantage (freeware) and its big disadvantage of its less friendly user interface the results I'm getting sometimes are very good and sometimes horrible depending of the subjet and stacking method. Perhaps to include it in your test would also be very illustrative.

[ChrisLilley]

What I have done, I frankly confess, is to deliberately choose subjects that illustrate what happens outside the range where Helicon does well.

This approach goes directly counter to reviewers' common practice of choosing some easy subject that every software handles well, and then structuring the review as "well, they all do great work, so let's talk about other stuff".
--Rik

Which any competent reviewer should be doing but, increasingly, most don't do.

<rant>
Symptomatic of the general decline of incisive investigative journalism, most reviews now, as you say, give little in the way of practical and in-depth help to select between alternatives on their strengths and weaknesses. Instead they trot out some facts and figures harvested from press releases or brochures, offer vague platitudes, then make comparisons on basis of price or small differences in numerical parameters (which they have not measured themselves and often, don't really understand).
But on the plus side, the whole 'review' is completed in 30 minutes, they still get paid, and they didn't annoy any of their valued advertising clients.
</rant>

An example review of this type, comparing Photoshop, HF, CZP and ZS. No substantive comments on image quality or handling of tricky subjects, but it does discuss pricing quite a bit.

Given that I already own Photoshop CS5 and Helicon Focus Pro X64 I haven't done too much playing with CombineZP

This, apparently, counts as a comparative review.

Having said which, the criticism of ZS not offering colour-managed preview and retouching images is, I think, merited and this would be useful as an option in a future version. Particularly for those folks who choose something like ProPhotoRGB as their working space, the non-colourmanaged view will look very washed out.

[rjlittlefield]

...the criticism of ZS not offering colour-managed preview and retouching images is, I think, merited and this would be useful as an option in a future version.

I agree completely. The beast would be color managed already, and actually it was at one time, except that it turns out the Java image I/O facilities I'm using become ghastly slow for non-sRGB profiles. It actually took some extra coding to remove color management for screen I/O while still keeping it embedded properly in the files, but the benefit of making that apparently silly change was to speed up the user interface by a large amount. If memory serves, "large amount" was something like 50X -- really an overwhelming amount that made the difference between "ridiculously slow" and "works fine".

Obviously it's possible to do proper color management and maintain good speed too, witness which other software does the trick. So in the grand scheme of things it's just a matter of finding the right bits and pieces and getting them integrated with the rest of the code.

Unfortunately that process of "finding and integrating" competes for the same resources that enable other code improvements, such as automatic shooting, prevention of dust trails, reduced noise accumulation, alternate retouching brushes, and literally 100+ other line items on the long-term "to do" list.

So it's the usual story: "on the list, can't say when it'll get done".

(Edited to add: It got done in March of 2014. Options > Preferences > Color Management > "Use input file profiles for ZS screen displays".)

--Rik

[rjlittlefield]

But, What about Combine ZP?. Apart of its big economical advantage (freeware) and its big disadvantage of its less friendly user interface the results I'm getting sometimes are very good and sometimes horrible depending on the subject and stacking method. Perhaps to include it in your test would also be very illustrative.

CombineZP can definitely be talked into doing good work. Prior to Zerene, CombineZM and later CombineZP were what I used for problems where Helicon got too frustrating. And indeed, there is some common heritage between CombineZ* and Zerene, since at least two of the key algorithms used by CombineZ* are in there because I introduced Alan Hadley to them and recommended he consider incorporating them. (Obviously that was "back in the day" before I started working on Zerene.)

With proper selection of methods and parameters, and staying within its limitations such as 8-bits only, I think that CombineZP can be made to produce very similar results to Zerene Stacker. But I will let someone else spend time researching and documenting how to do that, since for me it makes little sense to spend resources teaching people how to best use a competing product whose main advantage is that it doesn't cost any money.

--Rik

[Pau]

...But I will let someone else spend time researching and documenting how to do that, since for me it makes little sense to spend resources teaching people how to best use a competing product whose main advantage is that it doesn't cost any money.

Thanks Rik for your reply, I understand you perfectly
In my very limited tests I found CZP inferior to Zerene but much better than HF

[stevekale]

I must admit I am a bit surprised to hear that ZS isn't colour managed. (I use ProPhoto as a workspace.)

Having plonked down the $ for HF I am not happy to see these comparisons though.

One question, is the functionality between Personal Edition and the Professional Edition the same? (I don't sell images.)

[rjlittlefield]

I must admit I am a bit surprised to hear that ZS isn't colour managed. (I use ProPhoto as a workspace.)

Quoting from the Zerene Stacker FAQ at http://zerenesystems.com/cms/stacker/do ... r_profiles...

How does Zerene Stacker handle color profiles?

By default, whatever color profile is used by the input files is also copied through to output files that are saved as JPEG or TIFF. However, color profile is always ignored in Zerene Stacker's own screen displays. As a result, stacks using a wide-gamut profile such as ProPhotoRGB may appear significantly different when displayed by Zerene Stacker than when displayed by Photoshop or similar tool. On some computers Zerene Stacker will also ignore the color profile of your monitor, so that even sRGB images look different than shown in Photoshop. These differences may be worrisome if you don't know what's causing them, but the key thing to remember is that the color profile will be correct in the output file.

Zerene Stacker is a stacking tool. It is not designed as a viewing tool except as needed to support retouching, and it has no capabilities to adjust contrast, brightness, or color balance. As a result (it seems to me), the lack of color management during display has pretty small effect on its operation. At least that was the case in my tests. If I'm wrong about that judgment, then more information and discussion would be most appreciated.

One question, is the functionality between Personal Edition and the Professional Edition the same? (I don't sell images.)

http://zerenesystems.com/cms/stacker/do ... ofessional

What are the differences between Personal and Professional?

There are no technical differences in the current full release; differences are entirely in the licensing terms. As major new features are added, they will generally be available only in Professional (after a free trial period in Personal, of course). We are currently beta testing two new Professional-only features: an integrated controller for the StackShot automated rail and the ability to ingest and stack images while they are being shot with a USB-connected camera. Other features will be added later, but we can't commit to a schedule. If you're eligible for Personal Edition and you don't need the features currently in beta, then we recommend to buy Personal now and upgrade later if desired.

--Rik

[ChrisLilley]

I must admit I am a bit surprised to hear that ZS isn't colour managed. (I use ProPhoto as a workspace.)

The preview images and the retouching window are not colour managed. If the input images have a colour profile, this will be copied to the output image. So subsequent processing in a colour managed application is unaffected.

[rjlittlefield]

The preview images and the retouching window are not colour managed. If the input images have a colour profile, this will be copied to the output image. So subsequent processing in a colour managed application is unaffected.

Very concise. Thanks.

--Rik

[rjlittlefield]

Alignment accuracy is another issue that may be of interest. Normally I don't run into this issue myself because I shoot mechanical stacks. But this evening I happened to go looking for something to photograph handheld. Only one thing happened to catch my eye: this fly.

6 frames shot handheld, processed with Zerene Stacker at default parameter settings and with Helicon Focus with maxed limits on alignment.



In case you're wondering about the "maxed limits", I also tried with default and with limits set to the same values as Zerene's defaults (20,20,20,20). Here are the HF results with those settings.

HF, defaults:


HF (20,20,20,20):


This exercise didn't even start out to be a comparison. I just wanted to go shoot something with my new setup, and while I had the stack in hand I thought I'd run it through both tools just to see what happened. Somehow the results were more interesting than I anticipated.

--Rik

[stevekale]

Helpful. As you can tell, I ran a little fast into HF and have not looked at ZS and didn't want to given my investment in HF. I suspect I now have to.

I don't see any need for ZS to offer colour correction or curves. (I would rather do this in RAW or the LAB colour space.) I understand that the image profile is carried through (or could even be manually reassigned). It's just odd for me to see any image programme that isn't colour managed these days and would have thought it "easy" (relative to image stacking algorithms) to incorporate.

I also have questions re RAW handling but will read the ZS website first. To my mind, it would be much more efficient to do all the alignment, enlargement, merging and "composite" editing in RAW and then hand a RAW file back to the use at the end. One thing I don't like about HF in this regard is that it does the RAW-to-RGB processing which of course doesn't allow me to use settings (contrast, colour etc) appropriate for my individual camera, lens and the image (white balance). So I have been processing all the stack images through ACR first which is more time consuming and memory intensive. I don't know anything about the underlying techniques used by stacking algorithms to understand whether it's possible to keep all this in the RAW domain.

EDIT: re full RAW ZS, one issue would be pre-stack-processing "capture" sharpening which I like to do in ACR. So replicating this in ZS would be good although again I'm not sure which domain ACR is actually doing the sharpening - RAW or in the pro forma processed RAW. I suspect it may well be the latter.

I NOW SEE IT DOESN'T HANDLE RAW (but I would have thought this area deserves some consideration - just not if it is done the HF way)

BTW I just watched the retouching videos. Very impressive. What's it like to use with a Wacom Tablet? (I'm particularly thinking of the scrolling)

[rjlittlefield]

Steve, I'm afraid you've fallen into the very common trap of elevating RAW to a pedestal that it simply doesn't deserve.

First off, please understand that RAW is not one format, but rather something like 38 different formats. As explained at http://en.wikipedia.org/wiki/Raw_image_format,

Providing a detailed and concise description of the content of raw files is highly problematic. There is no single raw format; formats can be similar or radically different. Different manufacturers use their own proprietary and typically undocumented formats, which are collectively known as raw format. Often they also change the format from one camera model to the next. Several major camera manufacturers, including Nikon, Canon and Sony, encrypt portions of the file in an attempt to prevent third-party tools from accessing them.

Second, the advantages of raw formats are due entirely to their being a faithful representation of what the camera sensor actually saw. Unfortunately, what the camera actually sees is typically a mosaic of RGBG, GRGB, or RGGB values, one color per pixel position, courtesy the Bayer filter that sits in front of the sensor to allow color information to be captured.

Information in Bayer filter format is essentially useless for most photographic purposes.

That's why the first step in digital photo processing is to carefully convert the Bayer matrix pattern that the camera actually saw, into a traditional RGB or LAB representation in which the same type of information is stored at each pixel position.

That word "carefully" is important. Converting an image from raw to RGB or LAB representations is a very complicated process. (If you want to know more about that, you can start at http://www.cybercom.net/~dcoffin/dcraw/ and drill down all the way to source code if you want to. I've done that; I don't recommend it.)

So, the trick to getting a great result when working with raw images is to use an excellent converter with appropriate parameters.

You wrote:

To my mind, it would be much more efficient to do all the alignment, enlargement, merging and "composite" editing in RAW and then hand a RAW file back to the use at the end. One thing I don't like about HF in this regard is that it does the RAW-to-RGB processing which of course doesn't allow me to use settings (contrast, colour etc) appropriate for my individual camera, lens and the image (white balance). So I have been processing all the stack images through ACR first which is more time consuming and memory intensive. I don't know anything about the underlying techniques used by stacking algorithms to understand whether it's possible to keep all this in the RAW domain.

Now perhaps you see the issues more clearly. Not only are there a bunch of different and incompatible raw formats, but almost all of them are simply terrible for operations needed for stacking, such as small changes in shift/scale/rotate needed to align stack frames. The only reasonable way to do stacking from raw is to convert first, then stack the converted results.

So, the approach taken by Helicon makes perfect sense. Conversion to RGB is required in any case, and for the utmost in user convenience, there's nothing better than to convert behind the scenes where it doesn't clutter up the workflow. Now it's true, as you've noticed, that a certain amount of control may be given up in that process, but apparently a lot of users either don't understand that or don't care about it, because direct input of raw images is a very popular feature.

At some point Zerene Stacker will probably provide direct input of raw images too. But when it does, the implementation will be just the same as Helicon's: behind-the-scenes conversion to RGB for the sake of user convenience. The recommendation for best results will continue to be the same as now: control the conversion yourself using a separate tool that's optimized for your specific purposes.

I just watched the retouching videos. Very impressive. What's it like to use with a Wacom Tablet? (I'm particularly thinking of the scrolling).

I've heard from Wacom Tablet users that it works great and they wouldn't do it any other way. For more details about how to set it up for most convenience, I would have to go ask. I do own a Wacom Tablet (bought for testing), but quite frankly I've never spent the time to get good at using it.

As you can tell, I ran a little fast into HF and have not looked at ZS and didn't want to given my investment in HF. I suspect I now have to.

I apologize for the inconvenience. Whether you switch tools is entirely up to you, of course, and the tradeoffs depend on what you want to do. At lower magnifications, HF does great work, and when it does work well, it can be faster and more convenient than Zerene Stacker. With higher magnifications and deeper stacks, the tradeoffs shift in favor of ZS because that's what it was designed from the beginning to do.

The other bright spot is that because of the difference in pricing models, as a Personal user you can get a very well featured version of Zerene Stacker for less than half what you may have paid for Helicon.

--Rik

[Harold Gough]

Rik,

I'm very new to stacking but I'm struggling with the two stereo pairs of the spider. Flattened effects on the bristles may be one thing but the second pair seem to be from a slightly lower viewpoint.

Harold

[rjlittlefield]

I'm struggling with the two stereo pairs of the spider. Flattened effects on the bristles may be one thing but the second pair seem to be from a slightly lower viewpoint.

You're right -- good catch. The difference is due to a setting I didn't notice when I captured the screenshots. Both products allow the viewpoints to be positioned arbitrarily, but the user interfaces are different and it's a little tricky to get exactly the same views. The Zerene Stacker pair shown here is strictly left and right of the central axis where the stack was actually shot. The Helicon focus pair are left and right of the central axis, and also both are from slightly below the central axis. Your description is exactly correct.

--Rik