Switch from Photoshop to Gimp: Tips From a Pro | Hacker News Reader

Switch from Photoshop to Gimp: Tips From a Pro

rileybrandt.com

259 points_tznn125 comments

This was posted to reddit a couple weeks back, and I'll repeat the same comment I made there.

"I spend about 90% of my time in Lightroom and only 10% in Photoshop."

This is why he was able to do this. He isn't using the functionality that Photoshop provides, he's using the functionality that Photoshop Essentials provides. Look at his parity instructions - every feature is found in Photoshop Essentials. If that's all you use, then great, by all means switch over to Gimp. The bigger the userbase it has, the better off it will be.

But don't think that it's going to replace Photoshop in the near future. Non-destructive editing is the biggest thing GIMP is missing in my opinion, but supposedly the move to GEGL will allow them to start development on this ( https://mail.gnome.org/archives/gimp-user-list/2013-December... ).

Smart objects are another huge feature it's missing - not only the ability to downscale losslessly, but also the ability to edit all replicates at once. This is huge for designers. If I'm defining a user control for a design mockup, and I need to make a change to that control, in Photoshop all I have to do is edit the singular smart object source. The source edits will propagate to all the copies automatically. In Gimp I have to do this by hand. There are some minor features too that bug me. The inability to add a mask to layer groups is a big one for me. Layer effects (while often overused and gaudy) can be really helpful for design work - need to change the color of an icon that's raster art? Just drop a color overlay on it. If you have style swatches, it can be really easy to do fast mockups using this. This in conjunction with Layer Comps (also something missing in Gimp right now) can really help in switching between two or more alts. A great way to see what Gimp is currently missing in comparison to Photoshop is to look at the development roadmap ( http://wiki.gimp.org/index.php/Roadmap ). If some of your heavily used features are on that list, it might not be worth switching over to Gimp. If you don't see yourself as a user of those features though, give it a shot.

Worse, lightroom is an adobe application. I no longer use Photoshop because I use another Adobe tool more often... Makes the switch to GIMP kind of pointless. Not that I should talk, I hardly use Photoshop any more either, but that's because I live in Illustrator.

My problem with open source graphic tools remains the same. I'm in the print world, either it supports Pantone or it's not that useful of a tool.

I hear this stuff all the time. So this is my own canned response:

https://news.ycombinator.com/item?id=6675258

> it has a ways to go before it will actually be a competitor. And it doesn't have a few features that I find incredibly invaluable as an Evernote user:

[list of features snipped out]

As an Octave developer, I frequently hear exactly the same things about Octave vs Matlab, and if I listen too closely, I find it disheartening. Why work on something that will forever suck and not be a competitor, and can't implement the full list of features because we don't have the giant budget of our non-free competitor?

But then I remember that people are typically more flexible than they appear to be when they write these lists, and you'll discover that they may do without some of the features that they list, at least in some circumstances. And that there exist many other people with different lists of sine qua nons who will use the free as in freedom alternative you're working on because they have different needs.

So, to the OpenNote developers and to anyone else implementing a FAIF replacement I say this: don't let these giant lists of features and suggestions about how you'll never be competitive grind you down. There are many people out there who will appreciate your work, sometimes even the people who compiled these lists. Look at the lists, see what you can implement, and don't be disheartened by the parts that seem impossible to implement. Who knows, maybe some day someone will come along and help you implement the parts that seem so hopeless to you.

Keep your chin up, free software devs!

Octave is awesome! Keep up the good work. At university everyone was begging for Matlab instances from the licence server, while I had 6 instances of Octave open :)

Instead of the GIMP limping along behind PhotoShop, both dragging their decades-old legacy baggage, monolithic architectures, 1990s pricing models, etc., into the foreseeable future, I wish we had something more like Octave, SQLite, or Python: a powerful, faceless dev platform on which to build graphics-processing apps.

Basically, a runtime that holds a GPU-optimal image model (raster layers, vector layers, non-destructive edit management, multi-session undo, etc.), a set of APIs for sending memory-protected commands to the model, and a growing standard library of routines that have proven useful (network comms for offloading some processes or results to "the cloud", input/output codecs, CSS3, whatever). It would have a relatively easy to use scripting language with some built-in data types such as layer, region, vectorPolygon, gradient, font, etc., plus the usual lists, dicts, first-class functions, etc. Scripts that get a lot of use could be reimplemented in C, etc.

Developers could wrap whatever GUIs they wanted around this engine, creating easy consumer apps such as a slideshow or a camera-to-computer file importer with a magnifying glass and a keep/delete shortcut, all the way up to building clones of PhotoShop, Illustrator, resurrecting Fireworks, and so on.

Maybe someone would build a GUI builder for it that let you drag image rectangles and buttons into a window and script the whole thing, so non-programmers could create lens comparison apps, blink comparators for comet hunting....

I'd so much rather have a platform like this, where the hardest part--the image engine--would be done by experts, and the apps themselves, the feature sets, the UIs, could then be built by hordes of amateurs.

> I'd so much rather have a platform like this, where the hardest part--the image engine--would be done by experts, and the apps themselves, the feature sets, the UIs, could then be built by hordes of amateurs.

The naïveté of a budding hacker.

Great UIs are done by experts too, young padawan.

Great UIs are done by experts too, young padawan.

Yes, they are, but that doesn't preclude building the underlying data model and functionality as a common platform with shared tools. It also doesn't mean that building that platform and implementing a comprehensive range of functionality isn't a much more demanding task than building a UI on top if you've already got a good foundation.

No doubt there is plenty of hard work to go around, but I think the idea of building more software with this kind of strategy has a lot of potential. It could save on the donkey work for everyone, leaving different developers to focus on the interesting/useful/distinctive aspects for different projects, while retaining some degree of compatibility and robust basic functionality.

I wasn't disagreeing with your vision, but rather with your dismissal of UIs as "amateur work".

And I'm not even a experience designer :)

It will make more sense if you know that, by "amateur", I mean someone who knows something about image processing, but not enough to build an image processor--someone whose expertise is in something else: photography, public presentation, UI design, etc.

I should have worded my final sentence differently, but in the sense that someone who knows enough about database engines to be able to use one but is an "amateur" at creating them can still use SQLite to build an app incorporating a database engine, my "vision" would let developers whose expertise is not in image processing engine implementation to write apps that incorporated a sophisticated image processor.

Please note that I wasn't the person who you replied to before. In fact, part of what I do professionally involves UI work, so clearly I don't regard UI design as merely work for amateurs. On the other hand, I also find that 90% of the skill in building a good UI has very little to do with knowing how to make dialog boxes and toolbars appear. The functionality behind those UI elements, how that functionality is organised, and presenting everything as clearly and ideally as simply as possible is often far more important, IMHO.

I am quite a user of imagemagick, or imagick in php. This is far more useful than Photoshop or Gimp when it comes to cropping images and applying basic filters such as a sharpen on thumbnail images. I also use Gd2.

You would be surprised at what can be done in code. You would also be surprised at how much of the stuff that can be done in code is done the time consuming way by people with expensive Photoshop rigs.

So, to a certain extend a lot of what you desire can already be done with code. However there is not one single 'graphic artist' degree where such possibilities could be entertained, plus programmers are not allowed to touch images under any circumstances whatsoever. There is a rift valley between programmers and graphic artists, a deep cultural one.

Something like a programming language for composing graphics? Racket comes awfully close to that ideal.

http://docs.racket-lang.org/quick/index.html

I laid out a conference poster this way.

The poster: http://temp.sneakygcr.net/cvpr2012-mughunt-demo-poster.pdf

The source: https://gist.github.com/gcr/9854495#file-poster-rkt-L227

Unfortunately I gave in and made the center figure in Inkscape.

I just want to let you know that I did my entire masters thesis/project using Octave (sound signal processing). Please keep up the good work. :)

>He isn't using the functionality that Photoshop provides

And since that is the same for the vast majority of photoshop users, he provided a super helpful guide for them. It does not say "how everyone can switch to gimp".

You have a long road to hoe before you claim so emphatically that that's true for the "vast majority" of Photoshop users. Photoshop is the 90/10 app to end all 90/10 apps.

I do very straightforward digital art and manipulation and I'd put my head in an oven before trying to do it in GIMP, and I am not using anything complex.

I didn't say you would want to switch. I said most people use little photoshop functionality. If you think that is controversial, then I think you are delusional.

GIMP doesn't support all of most people's ten percent, almost regardless of involvement. That is very far from delusional.

This is like using Vim-mode to switch to Emacs, or installing Lindows. If people could accept the trivial conventions of the software you're moving to ("How can I use GIMP? The hotkeys are different!") and stop treating it like a free badly-executed Photoshop clone, they might realize that it's really nice to work in.

My problems with GIMP were the CMYK stuff and a few terminology issues. I really love the GIMP interface, and I'm not the only person who finds the Photoshop interface awful.

I think that people just hate to have to learn anything new, and are willing to pay rents for the rest of their lives to fend that moment away where they might have to spend up to a week or more working in something that they're unfamiliar with. It's a bit like the fortune web developers pay to not have to figure out unix, or Apache.

Or like using a guitar tuner to tune each string separately. Maybe I'm just an angry guy:)

It takes a lot of effort to change/unlearn muscle memory. Hell, just ctrl-f vs. ctrl-h, and F5 vs F8 etc is pretty thought consuming, let alone years of using a given application and switching to another.

The thing you're missing is that the differences between GIMP and Photoshop are not just user interface idioms. I wish it were that simple (I'd love to switch to GIMP if I could). It's not a matter of "learning something new"; it's the fact that without real non-destructive editing features, GIMP is basically just an enhanced version of MS Paint in the hands of serious photographers and retouchers. Non-destructive editing is not a "nice to have" feature. It's quite literally a fundamental shift in methodology - and one that apparently the GIMP devs seem to be ignoring.

I know several professional photographers/retouchers, and for them, a typical Photoshop document has a single source layer (the original photo) on the bottom, and hundreds of adjustment layers (organized into folders) on top to create the final image. There's just no way to approach that power with GIMP.

Shouldn't you know what you want to to, if you're a professional?

If you don't know what you want to accomplish, you need to experiment. This is when those non-destructive features come in handy, but if you know the goal and how to achieve it, they aren't necessary.

Non-destructive editing isn't about knowing or not knowing what you want.

If you're a professional coder, shouldn't you know exactly what you want to accomplish without needing to experiment? Why not then write the final version of your code all at once? Why use revision control?

You should know what you want to do, of course, but you are not usually the client unless you're a hobbyist (or a fine-art type, who is essentially a hobbyist who takes money in applause). Non-destructive editing is not optional. (A photographic file might typically have several tens of layers, and that's at the low end of the scale. Hundreds is not uncommon.)

This post is a really poignant (if unintentional) example of how a lot of programmers/non-arty types have a fundamental misunderstanding about how graphic design works.

"Just do it right the first time" is such a maligned attitude when directed towards coders ("No, boss. I need to make revisions!"). Interesting that the perspective isn't being applied equally.

Shouldn't you know what you want to do, if you're a professional? If you don't know what you want to do, then you need git or other version control. If you know what you want to do, why do you need version control or patches?

Andrew Clarke talks a bit about this: http://www.smashingmagazine.com/2014/03/25/a-modern-designer...

"Shit's Easy Syndrome": http://steve-yegge.blogspot.com/2009/04/have-you-ever-legali...

I'm not saying this stuff is easy.

I'm saying professionals (people who are skilled in non-easy stuff) know what they want and just do it.

Probably there are just many people out there who using PS and aren't professionals, which is okay...

I note that it has taken you - so far - two attempts to make your point ;)

zing!

No, you shouldn't. That's spoken like someone who has never worked in graphic design.

For one, professional graphic design is still an art. An applied art, but an art nonetheless. And as in all art, experimentation and happy accidents are essential. (And of course Photoshop/GIMP are also used by digital artists, not just graphic designers).

Second, (and I feel like I'm stating something beyond obvious), people make mistakes and people change their minds, even if they are professionals. Here, non-desctructive saves many hours.

Third, you obviously forgot that clients also change their minds, and can demand this or that change. Again, here, non-desctructive saves many hours.

Fourth, it's not different than say, the ability to revert a repo to a known state, that a SCM system gives, something that professional programmers consider essential.

What you say basically amounts to: who needs SCM and good prototyping features like a REPL. A professional should just start coding and already know his goal and how to achieve it (rolls eyes).

>> and one that apparently the GIMP devs seem to be ignoring

If you can code then contribute the feature, if not then hire someone to write it and contribute it and if neither then at least file a bug/feature request.

Non-destructive editing has been on the list of GIMP feature requests and roadmaps for many years. I may be wrong, but it sounds like the problem is more a misalignment of project goals with target user requirements rather than a sheer lack of development resources at this point (since so many other features have been implemented in the meantime).

I could be wrong too however we are both just stating opinions and mine is proven to work eventually so why am I being downvoted?

Because "code it if you know how or pay someone to code it" is not an answer when talking about software and lack of features, even if it's OSS.

OSS software is built by its community of developers, not by its community of users (even if those overlap a little, the second is many orders of magnitude bigger). So when someone criticizes it that it lacks feature X, telling them to "code it" themselves is not really helpful. They already know they have this option. If they had the knowledge, inclination, interest and time, they would have done it already. Not to mention that even if they tried, community politics and roadmaps might be against their wanted feature.

Can we accept that someone is free to criticize a piece of OSS software without being obliged to fix it himself?

But these kind of critics about known problems don't add any value.

Or, in the real world, where people just need to get shit done and do not know how to program, I'll just buy photoshop. Why in the world would I spend more, in time or dollars, to add a feature to Gimp which already exists in Photoshop?

Did you read my suggestion as dictation? It was just a suggestion, nothing more, if you prefer PS, go for it, no worries.

I am saying that yours is a poor and unrealistic suggestion.

I'm not trying to deny that GIMP has technical deficits compared to current versions of Photoshop - I'm certainly not enough of a Photoshop power-user to say that. I am annoyed about the amount of time wasted in making GIMP look like Photoshop rather than attacking the feature deficits. (Of course, those developers are just scratching their own itch, and we should all be grateful to them for it, and if I care about it so much, I should be coding it.)

If I were a Photoshop user, and GIMP looked just like Photoshop but worked differently, it would leave a really sour taste in my mouth. Additionally, Photoshop can change interface elements at a whim, so it's a moving target of dubious benefit.

>GIMP is basically just an enhanced version of MS Paint

GIMP is 10000x the tools that people used to do wonderful graphic work over the history of photography. If you're willing to pay thousands in order not to devise a workflow that doesn't need "real non-destructive editing features," the GIMP is currently not for you.

> If you're willing to pay thousands

Photoshop CS6 is $700. If you are doing professional work that is a trivial amount of money.

I don't think you can just buy photoshop anymore.

You have to subscribe to it for a monthly fee

http://www.adobe.com/products/catalog/cs6._sl_id-contentfilt...

You can still get CS6, but CC, which was equivalent to CS6 when it started, is coming up on a year old, and you pay 20 dollars a month to have access to it. CC will continue to get updates. When you stop paying your 20 dollars, you stop having Photoshop at the end of the month.

There was actually a $10/month offer for Photoshop CC and Lightroom CC (expires 31st March) - I got bought it for my other half. At $120/year it's cheaper than buying Lightroom in full here in Australia.

The frustrating part is that you can't pay for the full year up-front for most Creative Cloud plans.

Ex-Gimp User here. The article does point out that most photographer workflows are in their raw editors. You get Non Destructive Editing with tools like Darktable in Linux. Non destructive editing helps to improve your productivity. It's similar to making code changes knowing you have a revision system behind and to when you don't. That being said non destructive doesn't make you a better retoucher, it's there to save your ass, which happens less often for the seasoned professional. Professionals were productive long before non destructive editing. Gimp is used professionally, even though we hear little about that. Krita and Blender have all been used on big budget movies as well. Gimps biggest problem IMHO ( and strictly from my experience) is usability. This post isn't going to turn into a typical gimp ui rant even though gimp does makes my head hurt; a lot but I don't think I'll be going back to gimp unless I really have to. I have used since it since 2001 and seeing it lag behind other open source editors makes me sometimes wonder why usability is an after thought.

Another huge problem is that if I'm working with a "real" designer its exceedingly unlikely that they are going to have GIMP or be willing to learn it, and without pixel perfect rendering of PSDs (which it doesn't have) if I need to make buttons, etc I have to use Photoshop.

"Usability" never seems to mean anything specific, but rather whatever the speaker has in mind as a personal taste. In this case, my best guess is that you mean "UI works exactly like Photoshop"

I use Gimp professionally for UI and web work, and it does have some UI difficulties. The biggest for me is the haphazard arrangement of tool icons. It sometimes takes me ages to find the right icon because they aren't categorized and don't have a coherent style.

In Windows:

  Edit -> Preferences -> Toolbox -> Tools Configuration

Order the icons as you like. Hide the ones you don't need.

And also learn key shortcuts.

It's not just the default ordering that's the issue, it's the lack of grouping. And as a Vim user, I'm no stranger to keyboard accelerators. Unfortunately I don't use GIMP frequently enough for all of the keyboard shortcuts to stick (were some of them assigned from nouns in a language other than English?).

Thanks for the reordering tip, though.

Perhaps off subject but I think there is a need for a non-destructive, functional script-based image editor. I find, as a programmer, I prefer using ImageMagick when I can. Wouldn't it be nice to have the intuition of a UI backed by scripting? Imagine that at each step in your function, you can tweak it using traditional ui tools and the "tweaks" are automatically reflected in the script. To get the final export, you simply use the original image as the argument of your function.

Or perhaps such a thing already exists. If so, I'd like to see it.

One can script both photoshop and gimp. I spent a year and a half writing scripts against Photoshop and gimp has a batch mode in a CLI.

One used to be able to record a macro in Photoshop and export it to a script that one could call from a visual studio project. The code it produced was barely human readable but it was helpful for discovery.

I'm totally ignorant in this field.

If you're worried about messing up a file, why not just make a copy of it before you edit it?

over 9000 files names like myfile.backup.old.xcf myfile.other_edit.xcf myfile.tried-some-filter-10.xcf

It's not about a fear of losing the original, but losing each step in the process (which may be thousands of steps), having flexibility, and maintaining pixel information/quality throughout multiple edits.

Another way to think about it (in terms of programming) would be editing some code in an existing project, compiling it, and then losing the source code for your edits. Sure, you could keep patching the compiled version (not very easy) or use some kind of instrumentation to wrap/modify existing behavior, but if you can't easily modify the source of your past work (or just a single revision along the way), then all you're doing is piling more patches (in compiled form) on top of old work - not very efficient, and you lose flexibility.

In a non-destructive workflow, you may (for example) adjust the color balance of a single object in the scene, keep working, and then realize that the color adjustments are out of place. Since that adjustment is preserved and editable at any time independently of all other edits, then it only requires an adjustment of the parameters - and you don't lose any information.

If you were to do that kind of edit directly to the pixels (even if they're on a separate layer for the object in question), then each time you adjust color/levels/etc, you're losing information that was originally there. Similar to resizing an image very small and then trying to enlarge it to original size, destructive color edits will start producing nasty artifacts like color banding (you'll start seeing a "combing" effect in the histogram when adjusting color even after a single previous color adjustment).

Well, you're right. That sounds awesome.

Is there a good reason gimp doesn't support that? You make it sound like common sense.

Never used version control on your code?

> Non-destructive editing is not a "nice to have" feature. It's quite literally a fundamental shift in methodology - and one that apparently the GIMP devs seem to be ignoring.

It's on the http://wiki.gimp.org/index.php/Roadmap , UI designer Peter Sikkings has written a lot about how one might do it http://blog.mmiworks.net/search/label/GIMP%20redux and one of the main selling points of the move to GEGL is non-destructive editing. And all the focus now is on the move to GEGL.

Especially odd considering there is a free photoshop clone called Paint.net for windows. Its like a really old version of photoshop.

Funny, I've used many image editing software using different paradigm and Gimp manages to cut me at every corner.

Maybe someday the gIMP will be even close to Photoshop. Heck, maybe it is for some people's needs. But, I'm not even an artist and the gIMP doesn't come close to meeting my needs.

"No decent OpenType typography, no layer styles, no smart objects, no dice." (from the comments on the OP)

If all I need to do is crop or scale an image than sure, I might get by with gIMP though in that case I'd arguably get by better with something simpler than gIMP.

But, I actually use vector layers with layer styles ALL THE TIME. I actually use text layers with layer styles ALL THE TIME. I actually use non destructive adjustment layers ALL THE TIME.

Photoshop layer styles are like CSS. You can declare your styles and then edit vectors or text and the styles apply dynamically.

AFAIK gIMP has no equivalents. Those are not minor features. They're what set Photoshop apart.

> But, I actually use vector layers with layer styles ALL THE TIME. I actually use text layers with layer styles ALL THE TIME. I actually use non destructive adjustment layers ALL THE TIME.

Just use Inkscape and call it a day?

Not great advice for photo editing, which is a huge use case for nondestructive adjustment layers.

I think you misanderstood. He does not only use vectors, he just always needs to have vectors on his projects. Inkscape won't be a solution.

Apart from that, Inkscape seems to be a good solution for pure vectorial work. But my non-techies friends hate the UI. It is not that it is different of Illustrator, they just found it ugly and unpractical.

I agree on some points, but for the non destructive photo adjustment there is an excellent open source alternative caleld darktable which you should check out if you haven't: http://www.darktable.org/

Arguably, they don't support none of the vector layers, but thats something you could do in Inkscape. Agreed that it might be more convenient to stay in one tool, though.

Darktable is like Lightroom; it has the same editing limitations as Lightroom (or Adobe Camera Raw, for that matter). It's nothing at all like what non-destructive editing in Photoshop means.

I made a free online photoshop thing:

http://yangcanvas.com/paint

I use GIMP all the time for editing and its interface feels great to me. Mostly just like this silly thing I made as an avatar http://i.imgur.com/4xFBVus.png or this color swap I did http://i.imgur.com/S7QH2EA.png , or cropping out something to put it on a different background, changing the contrast/saturation/brightness/etc to make a picture look better, sometimes pixel art, or turning a series of images into a GIF. I used to use photoshop and I remember it not really being easier or harder. It did have moments where it was more convenient, such as automatically sizing a new image to the size of the image in your clipboard.

The new single-window mode helps a lot.

If you're using Photoshop for the purpose it was designed for (editing photos, not illustration/painting from scratch), then GIMP is absolutely inferior in its current form.

If you're making edits directly to source pixel layers (or even making duplicate flattened layers and working on those), you're doing it wrong.

Photoshop is designed for non-destructive editing. An essential part of that is adjustment layers. Many serious photographers will never switch to GIMP until it can match the non-destructive features of Photoshop.

Editing photos without adjustment layers is like writing code without revision control.

I'd have to agree with this. I've never used Photoshop, but the destructiveness of GIMP is a pain. Sure, .xcf supports multiple layers, but I can't rotate a layer without doing so destructively.

If you're making edits directly to source pixel layers (or even making duplicate flattened layers and working on those), you're doing it wrong.

That seems like a rather absolutist statement. I work with PS CC and its great features all day at work but I guess I still remember the fantastic work that was done before those features existed.

I have colleagues who still edit their photos using the apparently-despised techniques you mentioned, and despite "doing it wrong" they manage to make a great living off their photography work.

When I was teaching PS at my local college, I allowed my students to try out alternative software packages for extra credit. Some of my brightest students walked away from those classes happy that they could do everything they needed to do with GIMP, even though they knew all about non-destructive editing.

Agreed. Some gifted people create stunning works of art with unconventional or limited tools. Tatsuo Horiuchi uses Excel to create beautiful paintings.

I guess my point is that it's hard to argue against using non-destructive editing if it's available and you understand it. Can you create great things without it? Of course.

Bringing the analogy back to programming: some people create useful/beautiful apps without using any form of revision control, and that's fine for them. But if you know how to use revision control and have access to it, then is there really any strong reason not to use it, especially if you're working professionally? It helps increase productivity far more than not using it. For that reason, I'd definitely tell other developers that if they aren't using revision control for professional work, they're doing it wrong.

If you were to pick two professional photographers or retouchers at random - one who uses non-destructive editing and the other who doesn't, I'd be willing to bet the one who uses it (all else being equal) produces more consistent, high quality results and with greater efficiency than the one who doesn't.

Photoshop might be called photoshop, but I don't get why anybody would edit photos with Photoshop, Gimp, or any other similar editor rather than a decent RAW editor, be it Lightroom, Darktable, Aperture, or AfterShot.

Because editing photos goes way beyond adjusting color balance or overall brightness?

This reminds me of the endless bug reports we're getting right now about the Octave GUI not being indistinguishable from Matlab. Makes me wish people weren't so inflexible about the tiniest UI differences. It's difficult to please the converts.

> Makes me wish people weren't so inflexible about the tiniest UI differences. It's difficult to please the converts.

I replaced Photoshop with Gimp years ago, but I still find Gimp incredibly painful to use. It accounts for a tiny proportion of my software use, but a massive proportion of my swearing. I use to think it was just down to being familiar with Photoshop, but I think I can only excuse it on that basis for so long.

I had the opposite experience so I can confirm it is Gimp's interface. I started with Gimp and found it painful to use. I spent hours and hours reading tutorials for basic actions. Photoshop on the other hand was easy to jump into and start using just by reading labels. You still have to read/watch tutorials but things are much better labeled and organized and you find buttons/menus where you expect them.

Yeah, is it just me or are InkScape and GIMP both set up with rather unhelpful defaults? I didn't run into such problems with Huygen.

Inkscape is actually very good, and certainly comparable to Adobe Illustrator, if not superior nowadays. The main problem I see people have with Inkscape is when they have a MacOS computer, the only available build is from 3 years ago.

I use InkScape on the Mac. To get the most recent builds running, you need install XQuartz which is not bundled with recent OS X versions.

Even the latest XQuartz install appears to be 3 years old. Were you able to find a more recent one? I would be very interested if such a build existed.

I don't know of a standalone installer, but Macports has the current stable version (0.48.4).

I love InkScape, but it's a little bit crashy when doing stuff like adding multiple outline borders around fonts. And the extensions are all very mysterious, slow and crash-prone when they generate insane amounts of geometry if you happen to guess the meaning of one of the parameters wrong.

Some parts of the UI are very rough to work with. The gradient editor, for instance. Sure I can use it to make any kind of gradient I want, but navigating your colour stops through a pull-down menu? Then also not marking the current colour stop in the preview gradient, means you can't see "where you are" in your gradient as the pull-down is collapsed. Makes adding colour stops very unintuitive cause you won't know where they'll appear (before or after the current point) ... there's more, but it's all this unnecessary friction for something that should be quite simple, and has been done in various better or worse ways lots of times before.

Inkscape is not the same history as GIMP vs Photoshop, Inkscape is actually very very good in what it does, one doesn't even need to compare it to AI.

It's often hard to tell the difference between bad UI and unfamiliar UI. Seems like most people agree on which side Gimp lands on. No idea which side Octave's GUI is on.

In the past GIMP was the winner of my personal disproportionate swearing award. That all changed when I started to use Calibre for ebook management. Calibre is far and away, the worst UI for "best in class open source software."

It's not that people are inflexible, it's that if you're going to provide an alternative, it should slip itself into someone's workflow and processes with almost no inertia. When someone is accustomed to hitting a keystroke to achieve a certain effect, or someone is using to having a widget in the left corner, any change to that process is going to cause discomfort, and most people don't need that.

A lot of people just never seem to become accustomed to anything in particular, so these changes don't bother them. Nothing wrong with that (it takes all sorts), except that I've a strong suspicion that it is these people who then keep changing the keyboard shortcuts, moving the icons around, poking at everything so it looks totally different, and then removing the macro functionality. Because I guess they just have no idea how traumatic this sort of thing is.

If you use Windows, consider Paint.NET (http://www.getpaint.net) instead of GIMP. It's much quicker (particularly the 4.0 beta builds) and has a much more Photoshop-like UI out of the box.

Sure, but it doesn't do nearly as much. Especially when creating graphics from scratch

I like Paint.Net a lot on the windows side... for Mac/Linux alternative, look into Pinta http://pinta-project.com/ ... it's essentially a cross-platform port of Paint.Net

They do the vast majority of what most people need in a simple photo editor.

Looks like a clone rather than a port (as paint.net isn't open source you can't really port it), but I'll take a look.

Their github (https://github.com/PintaProject/Pinta) makes it look like the project may be abandoned.

It's a port of Paint.net 3.3.6 from .NET to Mono. Paint.net used to be MIT-licensed, and 3.3.6 was the last version that was released open-source.

Besides, the UI in Paint.net 4.0 has been rewritten to use WPF. That makes it essentially impossible to port to Mono.

For OSX, Pixelmator[0] is a great tool for the price ($30). I've used Paint.NET a lot, but since being on OSX a lot more for my job, Pixelmator is definitely a step-above Paint.NET. But as always, Photoshop is the best out there.

[0] http://www.pixelmator.com/

Krita may have a different aim on paper, but that just looks like a way to avoid conflict with the GIMP devs until it's clear to absolutely everyone that they've replaced it.

One of the real shockers with the GIMP is how badly it plays with Ubuntu's Unity, where the menubar will get emptied whenever you change something in a non-image editor window.

Krita is a great program, and maybe a future GIMP replacement, but for now, it is strongly focused on Digital Painting, and it does a great job in that field.

Curses! I've been seeking a free program with PaintDotNet levels of functionality which works on Mac for ages (We have MacBook Airs at work, while I use PaintDotNet for game dev at home). GIMP is really difficult to draw anything with, even the simplest things, with good cause - it's not a drawing tool.

I thought Krita might fit the bill, but apparently it only has support for Windows and Linux. Much sadness.

EDIT: never mind, just seen the Pixelmator rec lower down - problem solved I guess!

While I never used photoshop, I also never quite liked the gimp.

I actually sometimes use blender (yes, the 3D software) to edit photos. It's not the same paradigm as photoshop or the gimp, but it lets me do advanced color grading very easily. I can use textured 3D objects for 2D Animations or non-destructive editing. As an added bonus it works for videos and the compositing setup can be reused for multiple photos. Probably if I were really adventurous I could use more of the paint mode and texture paint mode to get even closer to photoshop. But then again I'm not a pro...

I've always disliked GIMP's floating windows (especially with a tiling window manager) but never knew of the single-window mode. That is a great tip!

Don't worry, it's pretty new. http://www.gimp.org/release-notes/gimp-2.8.html

I was a Photoshop user for a long time, but haven't looked back since I moved to Linux full time a few years back. The challenge was to use libre software for everything and reduce my business' expenses. It's been 4 years since then and I don't miss Photoshop. I realized in the process the I was terribly misusing it. It was my goto tool for everything, from web design, to photo retouch, to vector, icons...

I have a background in 3D & VFX but my work today is mostly web development, and opensource software covers all my needs. MyPaint for free drawing like brainstorming, illustrations, and blueprints. Inkscape for vector graphics like icons, logos, layouts, mockups. Darktable for simple photo editing. GIMP for heavy photo retouch, and well, it's still my goto tool for those quick screenshots, image cropping jobs. And Blender for 3D work of course. I hardly do print, but when I do I use Inkscape, HTML and CSS.

TLDR; Photoshop isn't the best for every task. There are free software alternatives for everything. Using them all together provides me with all that I need, so I no longer miss Photoshop.

As a developer, I've often felt that if I can't script something, embed it into my own software or service, or even redistribute it, then I haven't really done it with software. This is an unrealistic goal if I was in DTP work still, but by setting this bar, I've been amazed about how my understanding of the context of a problem has broadened my horizons. For instance I feel I now have a great grasp on Nurbs, Voxels, and 3d meshes in general from having done it all with Clojure instead of Blender. Batch processing with Imagemagick is now so second nature to me I can't stomach the idea of having to load up PS or even Gimp.

I think when the process involves subtle interactive hand touches and nuances, then an interactive editor will always win out. However rudimentary, discrete "operations" are things computers should do well, and PS and Gimp fall short in my opinion. Gimp being slightly better sort of with its open scripting.

Photoshop got full scripting support in 2003 with Photoshop CS. Before that you had to install a free plugin and it didn't support everything but it was still pretty powerful, I don't remember exactly when that came out but I think it was first for version 6 in 2000.

Photoshop has also had full recordable macro support ("Actions") since 1996. Many photographers I've worked with have developed huge suites of actions they've created that enable them to automate much of their workflow despite having no coding ability whatsoever.

http://www.adobe.com/devnet/photoshop/scripting.html

Yeah I guess I was thinking about that when I wrote this, but for instance with Python I can instantiate a batch of transformations, and integrate other libraries, other programs, and share inputs and outputs among all of them. I'd be excited if there was a CS scripting whereby I could take an AI file, apply layer effects, and then position that in a 3D transform in AE, and render an MP4 all in one script.

I can do that with Linux and open source utilities, but I can't with CS. Also AI is horribly leaky in memory when dealing with advanced computational design where you have billions of points, whereas open source SVG renders take such a thing in stride.

For web developers and web designers, GIMP is a great tool.

Best of all it is free.

While I'd gladly pay ten or twenty bucks for cropping and texting images for my web properties, I'd never pay hundreds for photoshop and their invasive anti-piracy software.

I think it is a mistake to measure GIMP against a standard set by Photoshop. If a person wants a program to work like Photoshop, they should probably use Photoshop. The standard for GIMP should be whether you can use it to accomplish your goal.

You might as well publish a guide on how to make a circular table saw cut like a band saw. That's completely useless both to anyone who doesn't already know how to use a band saw, and to anyone who already satisfies all their wood-cutting requirements with the circular saw. They are different tools; they don't need to work the same way.

I only use Photoshop to slice the web site designs my coworkers send me. I would have switched to GIMP a long time ago if they all used it, too. GIMP doesn't read their PSD files very well.

I'd recommend you switch to Sketch[1] and/or Acorn[2] instead if you're on MacOS. I love Gimp, and was a long time user of Inkscape after growing tired of Adobe's products, but the UX simply is too frustrating.

[1]: http://www.flyingmeat.com/acorn/‎ [2]: http://www.bohemiancoding.com/sketch/

I suppose it depends on what you consider a "pro" or not, and what you are using Photoshop for to begin with.

I started out 20 years ago as a retoucher and digital pre-press. 20 years later, I'm back in the publishing game, but publishing on the tablet. The magazine I am producing as the "eat your own dog food" test run of our bootstrapped tablet publishing tools (see https://www.youtube.com/watch?v=22qqsFHH1HY or https://www.youtube.com/watch?v=YXlHFhbqzHU for more info) is almost entirely photography and video. Guess who is doing the "pre-press" work? Yeah, that's me.

So let me run you through my typical retouching workflow I use to retouch a fashion photo (discussion about perpetrating negative stereo types of female body types ignored for the sake of this conversation):

    - Open the image in Photoshop
    - Duplicate the background to a new layer and convert to smart object
    - Open up the smart object into a new document
    - Liquify, patch tool, liquify, patch tool
    - Hair, pimple removal with clone stamp
    - Masked adjustment layers for very slight color correction, typically make up and teeth and eye whitening
    - Save smart object
    - Go back to original document
    - Duplicate smart object twice
    - Highpass filter + overlay blend mode on top most smart object copy
    - Smart blur on smart object copy beneath high pass
    - Mask the blurred smart object and start painting in skin smoothing on the channel mask
    - Now at this point, I might have gone back to the smart object and made some more clean up and liquify tweaks
    - Manual corrections done, I move onto color grading and color correction

All of this is done with a wacom tablet, which afaik, GIMP still has issues with (usable but not on the same level as photoshop). Liquify tool in gimp is a horrible joke. No smart objects.

Can you do this process with GIMP? You can. I have. It's not as easy or straight forward as it should be though. It can't be stressed how important Smart Objects are to the modern photoshop workflow.

So as someone who uses Photoshop not only for app design, but actual honest to goodness professional retouching, I have to say if this guy is only using Photoshop 10% of the time, he's probably not even the target market for it. Will GIMP ever make it to actual competion for Photoshop on the professional level? I sure hope so, but not anytime soon. Photoshop CC (I've been using Photoshop since 2.5 on a Mac Quadra 950) is the best version ever. I know I sound like an Adobe shill, but I have extensive expertise that goes back decades now.

Now you damn kids get off my lawn.

Man, breath of fresh air, have never bothered tweaking GIMP as out of the box it's usable but frustrating compared to photoshop or fireworks (yes, I've got the latter running in a Windows VM).

On Fedora here so GIMP is what I've been using for quick image optimizations when I can't be bothered to fire up a virtual machine.

Am definitely going to integrate some of the suggested tips, thanks to whoever posted this...

Cached version http://webcache.googleusercontent.com/search?q=cache%3Awww.r...

GIMP is lighter-weight. I use it all the time for web work. If it meets your needs, it's definitely preferable to PS.

I expected this to be a joke of some sort.

The only way to get something that works like Photoshop.....

I bet you can guess the answer!

It's a bit like all those people who try to get a PC to run OSX. Just buy a damn Mac for goodness sake!

Gah, I don't understand why everyone doesn't just have infinite money for goodness sake!

I've been paying the ~$20/month for awhile now. I'm totally happy with that, and I don't think that's an unreasonable expense, especially if you're using it professionally.

I haven't tried it myself so can't vouch for it but GimpShop (http://www.gimpshop.com/) fills a similar niche.

I registered only to warn everyone even do not try gimpshop - it will install findwide mailware.

Seriously... it's been years. You'd think people stop recommending it.

I'm glad you posted that before I installed it.

Every time someone posts about this I always see some post complaining about it. I therefore wouldn't suggest it even if I personally haven't tried it.

I originally created Gimpshop, but I'm not the jerk who owns that domain and added adware & spyware to the source. Sorry about that. I hate that this guy is out there making my fun little project into an abomination.

Man, that sucks. Do you have a project site for your real source code?

I don't have a project site for it. I became discouraged after this whole ordeal and I let it slip away into obscurity. Also, when I put Gimpshop out there, I didn't really know how to code, so updating the source to stay current was a huge undertaking for me. Since then, I've become a self taught programmer. I know Ruby fairly well. I've built a couple of non-trivial Rails sites and within the last 6 months, I published my first iOS app. Gimpshop was a fun little 'prank' that got bigger than I ever expected. Sad what it has become, though.

Thanks for clarifying - I always suspected this to be the case, but I wasn't sure.

Reply on news.ycombinator.com