Show HN: Polarr.co, a WebGL photo editor | Hacker News Reader

Show HN: Polarr.co, a WebGL photo editor

polarr.co

167 pointsChrisCinelli97 comments

I hate to shit on your parade because this is pretty dope, but the title is disingenuous at best.

But I use Lightroom every day, and you might be faster but you aren't working at the same level of quality.

I loaded up a 16-bit TIFF and got an 8-bit JPEG as a result, which leads me to assume that you are downsampling to 8-bit to work.

Also, the results in polarr are way harsher then Lightroom, which is pretty common for non Adobe products. GIMP is especially harsh in contrast to Photoshop.

But even before getting into the color correcting stuff, you're already shifting color (I'm assuming you are ignoring the embedded color profile in the TIFF):

http://i.imgur.com/QKb42l5.png

Also, Lightroom is way faster making adjustments. I can see polarr "thinking" when I make adjustments (it gets pixellated and then clears up), while Lightroom does them in real time.

I couldn't get your curves to work.

No split toning.

Lens correction should be automatic and profile based. Lightroom will automatically fix distortions based on Camera and Lens model. Same with basic color calibration.

Again, I don't want to be discouraging, but I find the hyperbole kind of frustrating.

Awesome comment.

TIFF is not officially supported now and so far we can only output 8bit JPEG in html canvas through downsampling.

The downsampling is process "the thinking" part is controlled to be not hurting the user experience and it's different depending on your machine. For faster machines we can remove the thinking part but it would take up more memory (right now the footprint during processing is less than 20mb, consider this is a web app)

I agree Adobe is doing better in interpolation (the harshness) and we are working to improve that.

What OS and browser you're using? The curves and split toning works well on Mac Desktops.

We only have two developers working on this right now. Do you know where we can get lens correction profiles, is it publicly licensed?

Using Safari 8 on Yosemite.

Clicking on the curve points doesn't do anything. I just retried it thinking maybe I had done something wrong, but still no love.

Re: correction profiles, I'm not sure. Darktable has some, though not as many as Lightroom, so there must be some public source for some of them. This is actually a pretty big deal (for me at least). I'm not a photographer, but I publish a digital/ipad magazine and do most of the retouching work.

http://www.darktable.org

(Not a fan of darktable, but they do some things pretty well).

Mouse move event on scaled element on Safari Yosemite seems to be broken, we're looking for a fix now.

IIRC the lens profile format adobe uses for Camera Raw and Light Room is human readable and editable so you may be able to use it directly.

If I was developing this, I would be stoked with this kind of feedback.

What is supposed to be faster? If there is downsampling to 8bit then of course it is easier. If you can't load the full res image then it's easier (and so on and so forth).

Impressive, sure. But faster? Lightroom?

Try loading two 24mb 8 bit JPEG files into Polarr and Lightroom. Then try adjust knobs such as exposure and temperature. The rendition in Polarr is apparently faster due to its use of downsampling on the fly but still it uses fully sampled images after the user stops the interaction. I believe Lightroom is doing downsampling in one way or the other.

Awesome.

Like I said, impressive, but you'll catch a lot of flak for an otherwise impressive application by using the lightroom hyperbole. I don't think many people use lightroom to edit 8bit jpgs (changing exposure on a jpg is a bit disappointing!)

Unlike some other fortunate users who reached the editor, I never did: http://tuxdiary.com/2014/10/28/polarr/

I did re-visit after a rather harsh review because of the loading issues and I am pleased with what Polarr has to offer. The review is updated now! :)

Dev of Polarr here.

I think the biggest thing that's missing from the current release is RAW image support which is technically difficult to do. We're right now having some success of extracting the thumbnail of RAW and do it on the client side only using JS, but editing the RAW is a lot harder and we're thinking about using some WebGL texture blending. Any thought/comment on this is welcomed and appreciated. (Not sure how Adobe is implementing their RAW editing)

-- The main idea of the company is in fact automated photo retouching https://www.polarr.co/box (please use desktop, large videos), and we applied to YC with the idea.

How are you going to solve the performance issues if JS's type is only 32 bit? I could see the possibility of mixing multiple JPEGs but extracting that probably takes you a minute in processing

JS numbers are doubles (64-bit floating-point numbers).

Do you use a three level hiearchy on the image? I assume you guys first down sample to full resolution preview(like 600x400) and a low resolution preview. When editing, filter applys on the low resolution preview and then full resolution preview (1.5second latency). When exporting, you apply all the filters on the original image. Is it the case?

Came here just to see about RAW support -- I think that's the only thing stopping this from being a legit competitor to lightroom, which says a lot so nice work guys! One other thing I would miss as well is the ability to auto-correct distortion through the lens profile. This is super quick and easy in lightroom and it auto-detects your lens from the image metadata. This typically reduces the vignette, so the ability to add that back in is nice as well.

On a totally separate note, I feel like you guys are positioned in a spot to solve the photographer storage problem as well, which is detailed remarkably well here: http://paulstamatiou.com/storage-for-photographers/ -- a service like this is something that I know I would pay for and I'm sure it's the same for a number of others.

> On a totally separate note, I feel like you guys are positioned in a spot to solve the photographer storage problem as well, which is detailed remarkably well here: http://paulstamatiou.com/storage-for-photographers/ -- a service like this is something that I know I would pay for and I'm sure it's the same for a number of others.

The workflow feels a bit contrived and the whole problem based only on the cramped space on the SSD disk. I find storing previews on SSD and then the RAW files in the catalog on a huge mechanical drive works perfectly. I can't shoot pics fast enough to cope with the drop in storage cost on mechanical You get 1Tb drives in cereal boxes these days.

Ths library is then just backed with a good backup solution (NB: NOT just a sync provider like DropBox/OneDrive/...). I use crashplan, but you could use BackBlaze or any stand alone backup package to a sync provider. I also sync to my NAS, as per the "store everything in 3 places" theorem.

A problem I would like to solve (well) though is the first part of the article, where he shares pictures with his relatives. Email SUCKS for this. I use zenfolio, but creating galleries just to share a few pics is tedious. I'd like something more integrated with lightroom, and less tedious/crappy than zenfolio (better raw support, simpler/better style support etc).

Have you thought about compiling dcraw to javascript with emscripten? I'm not sure what the performance would be like, but I've seen this work in the past.

I'd try rawspeed instead of dcraw. Much faster and cleaner code and the latest version supports almost all cameras.

Thanks! Will try this out!

rawspeed won't let you extract the thumbnail. Also, the reason I suggested dcraw is because it's one C file and that makes things a lot easier, especially when compiling to javascript.

If you want to do raw processing then you don't really care about the thumbnails. No idea about the javascript compile process.

We tried it. It works with modification to emscripten and compiled code. Performance is acceptable and I think there is a much larger space using GPU to do the decoding as well. But I think the user won't just want to edit the thumbnail though I think many of them didn't know that.

>I think there is a much larger space using GPU to do the decoding as well.

I doubt that very much. Raw processing is very fast these days and I don't think you could make it faster with the GPU given the memory copying overhead. Look at darktable for an example of how the GPU is used intensively for the pipeline but the raw processing is all done by the CPU and is very fast.

Could you make it faster still by compiling to asm.js? I'm not sure how it would apply to you, but I know Mozilla has been bragging about how WebGL games are faster with asm.js.

The code is GPU intensive but not quite CPU intensive so in this scenario I believe no. Certain WebGL games need their engine compiled into JS.

The Polarr Editor is great and is a already more useful than Lightroom. What is relly compelling as $1B company is the vision in https://www.polarr.co/box

The point of using lightroom is Cataloging and RAW photo development. I can't see what's useful since the hyperbole mentions Lightroom. It's a program to develop RAW's (Even the name of the application is a pun on that)

Polarr dev here. we're very interested in bringing raw in the web and if there is a possibility of decoding and editing raw properly, we're either the first to do it or it will become some sort of commodity supported from the browser layer.

Great, a good low-cost feature in the meantime would be useful messages when trying to import unsupported or corrupt files, seems nothing happens now?

FWIW the other key feature missing for me is the histogram.[1] Pretty much everything I change in Lightroom, I'm keeping an eye on the histogram to make sure I'm not (unintentionally) blowing things out or losing dynamic range. Being able to drag the histogram itself is a bonus, but being able to see it at all is key.

Oh yeah, and noise reduction. With low-light photos lack of noise reduction is a dealbreaker. Maybe that becomes more important with RAW?

[1] https://www.google.com/search?q=histogram+lightroom&tbm=isch

Thanks for recommending all these features. Viewing / dragging histogram will be in our dev pipeline soon. There are some generic noise reduction algorithms that do not perform very well so we want to implement a few from the recent research papers. this is in our pipeline too.

- Update, we just added the previously hidden de-noise function there, it is not the most efficient but algorithmically correct, hope this helps

Don't know if it may be useful for you, but Trovebox dealt with RAW import, you might want to see what they're doing and if you can take advantage of their work

https://github.com/photo/frontend https://github.com/photo/frontend/issues/1400

The front-page of the website is extremely laggy for me. I'm using the latest Firefox with Arch Linux. The framerate drops down to about 1 or less.

EDIT: The app itself is fast.

Yeah the home page is a disaster, super laggy and the photos are even revolving over most of the 'read more' text so I can't even read about the features. Please think about redesigning it.

It's using very heavy CSS3 transition for the homepage that causes the frame drop. Will detect that and avoid that from happening. -dev

Yes, for me it freezes my Firefox browser for a time and in chrome the startup screen is better but the application has unusable performance: Scrolling in the list of filters at the left-side is very laggy and applying a filter freezes both browsers for a few seconds. I even closed other tabs and applications but no change in performance.

I don't see it, the app works, but it's just a bunch of horrible filters a-la-instagram. They work fast but they are still nothing like Lightroom which does a lot more than ruin photos with filters. All the ones I tried degraded an image, not once did I see any improvement, it's either blowing out highlights, introducing color shifts, or expanding the range to lower contrast or compressing the range and crushing the contrast.

I guess it's cool that you can do this kind of thing, I'm sure Chromebook users will be happy, but photographically speaking, gag.

We hope to improve this by having a array of filters diverse enough to capture other processes in https://www.polarr.co/lookbook

The right hand side panel pretty much does what Lightroom is doing mostly and you're in full control of which direction of your photo is moving into.

Great technology. I'm maybe a bit stuck in my RAW file processing ways to find a use for this myself, but I am impressed with it's speed.

A couple of things I'd like:

- Proper support for high-res images - it might be in there, but the demo flower image is web res - The hover-to-select mechanism from Lightroom. You only need to hover over a control to adjust it with the keys. Appreciate the tab support though. - Finer/coarser value adjust with the arrow keys. LR allows for modifiers to increase/decrease the value change.

Bigger asks: - Find some way of emulating LR's automatic lens correction. Using this still feels like magic. - Spot removal is pretty important - Other local adjustments would obviously be nice.

It might be nice to offer this engine as a technology plug-in sort of thing. I know at least a couple of clients who would love to be able to use this level of adjustment on the photos they upload to their Wordpress sites. I imagine there'd be some demand for this plugged in as an API'd service.

Nice suggestion on the hover to edit, will add that! This post is turning into a mega UX/UI testing ground which I'm really glad.

Local adjustment is working right now in Polarr. Spot removal is a very different game but we will see.

Neat webapp, quite impressive.

Still, I’m disappointed at how lazy the UI design of nearly all image editors is. They take off-the-shelf algorithms/models, implement them in code, and then slap a slider widget in for every parameter in the model. A big wall of sliders takes basically zero thought or effort from a UI design perspective, which is why it’s so popular, but other than being ubiquitous, it’s generally a mediocre UI in comparison to something that has had real thought put into it.

(But again, that’s nothing unique to this app. Adobe Lightroom takes the same lazy way out. Several of the tools that have been in Photoshop since before 1995 are a bit more powerful/flexible, but everything they’ve added in the past 20 years seems to follow the wall-of-sliders model. Some expensive professional video color grading applications are a bit better, but still nothing too special.)

I reckon with the the lack of UI innovation. On tablets and photos there are quite some newer designs in terms of adjusting the knobs and values while leveraging gestures.

Waterfall design is a lazy starting point. The innovation in this app is making a usable WebGL editor that handles multiple editing sessions without suffering serious memory issues and crashes.

Do you have suggestions? When I look at applications oriented towards prosumers and professionals like this, photoshop, etc. I think the point is to give complete control to the user.

I don't know anything about photo editing.

One can always take inspiration from audio software and audio plugins, some of those have quite slick UI ideas.

Oh please, no. The fact that every VST maker creates their own custom GUI from scratch was the best and worst thing to happen to VSTs. All of a sudden I have a gazillion different plugins each with their own idea of how user interaction should work, meaning I learn a new interface for every single plugin.

At least a wall of homogenous sliders doesn't require me to learn how the widgets work.

Sure, but one doesn't have to reinvent the slider every time to do more than just "list them". Grouping elements, maybe even color coding groups, graphical indicators along with numerical values, all those things go a long way, and don't preclude standardization of the input elements themselves. Maybe they'd need to be more customizable.

> At least a wall of homogenous sliders doesn't require me to learn how the widgets work.

That wall is available in many if not all VST hosts, you can still work with the raw parameter list; and I have to say, I for one do not prefer it, at least not for any plugin that has more than 5-10 things to tweak. Except when I know them by name and can filter the list with a few keystrokes, so for that and other reasons I think a normalized is also useful. It doesn't have to be one or the other.

Photographer here. We know what these functions are doing to the image (if not literally then intuitively) and want the total control that the sliders-next-to-input-fields paradigm provides. You might be surprised how much math actually goes into photography. Yes I do want to bump the contrast by .1 and drop the exposure by 1/3 of a stop. I've got my eye on the histogram and I know what I think looks good.

I guarantee that any photo editing software that deviates from this UI will get shat on by the professional photography community. I'd love to see your suggestions nonetheless.

The type of interface that I think you're suggesting is Instagram, one that takes control away from the user because the user doesn't actually know what they're doing. This editor actually has something like that with the filters you can select on the left. The problem is, actual photographers are frustrated by these interfaces (or treat them as a toy) because they would rather have all the functionality exposed to them so they can make their own adjustments. You could argue that "constraints breed creativity," but when I'm working through 1500 wedding photos I want my sliders exactly where I expect them to be, doing exactly what they've been doing with high precision for the past 20 years.

I have my pictures on my computer and I have native applications to edit them. For my needs the free ones (as in beer and as in speech) are good enough. I can Save as... the images I find on the web and edit them too (licenses that allow reuse, Google has a search for them).

So far the only feature that Polarr has and a native application doesn't have is that photo styles from the community thing (not a quote because the spinning pictures are covering the words on my browser most of the time). Unluckily the home page doesn't say much more than that. Can somebody explain it? Thanks.

The community feature isn't in its prime time yet, once we nail most of the issues out for the editing experience and attracted a sizable audience, the Lookbook will perhaps become the largest photo style guide, crowd-sourced.

On Ubuntu 14.10, with Chrome and WebGL enabled I get an unsupported browser message.

Here is my UA: >Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.104 Safari/537.36

Odd. Im on 14.04 with the same UA and it works fine. Do webGL based apps work on other websites? You might have to enable webgl on Chrome I think I remember doing that at some point... My guess is he is just checking for webgl/experimental-webgl and popping up unsupported browser vs checking user agents. Not an ideal error message though.

User-Agent:Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.104 Safari/537.36

Try https://www.polarr.co/editor?force_browser_support=1

Thought I fixed that but..

This link works, thanks!

This is an awesome idea.

I think what people might be missing is that even with the features you currently have, there's a market for this application already, though maybe not as a stand-alone lightroom replacement.

If you think about batching, my brain immediately starts gravitating towards real estate, headshots, HS seniors photos, ENG.

In short: drop the filters, put the "basic" controls on the left, focus on what you do best.

Kudos, folks, I'll be watching this one!

Make sure you signup so that you can receive our mailing list, we're building a basic "editor-less" mode for Polarr right now.

Will do - but I suppose a "filter-less" version might be what I was leaning toward, as a potential user.

Your pricing is incredibly confusing. What is "polar box" and do I need to pay for it? What are the advantages of paying for it, what are the paid-only features?

from the "box" name i assume it's some sort of storage features, but when i click on the "$4.99" link at the top of the page it just takes me to a list of general editor features that i can seem to get for free by using the editor.

there is a demo video which explains everything.

Looks great. Any chance this could be bundled as chrome native app ? Similar to https://chrome.google.com/webstore/detail/pixlr-touch-up/jkl...

For a photo editing app, I hate opening up the browser.

Also curious, If the app is using emscripten in some way for the photo editing part.

Yes, we do!

https://chrome.google.com/webstore/detail/polarr-photo-edito...

The app is not using emscripten

Aweseome. For my quick photo editing needs, I've been using the Pixrl Touch App chrome app. I'm pretty sure Polarr is going to replace that.

Found it is also featured in Chrome App Store (on the top). https://chrome.google.com/webstore/category/apps?hl=en

It looks good and impressive. I'm assuming the audience is for casual photo manipulators who like to post their filtered photos on Facebook?

Adobe currently has a Lightroom/Photoshop bundle for $9.99/month, and I'm not sure who would choose this online app vs. native apps. Maybe if the cost was $50/year it would be more competitive in the space... $4.99+/month just seems like a lot.

Hi $4.99 is for Box, which is a different product from the editor.

Very impressed by the performance. With a 10mb png I don't notice any lag.

Is it just me or is the mousewheel zoom backwards? Max zoom seems a bit weak too.

I notice you mention RAW support. Is that really your target market? It seems like people shooting RAW would probably be using a dedicated solution most of the time.... maybe not. As far as I'm aware there are multiple different raw formats from different manufactures and adding support for all of them would be a huge undertaking.

If I am to be honest, I don't like how the controls are. The text size could be 1-2 pts bigger. The whole popout thing seems pointless as the space isn't used when it isn't popped out. I think either fix it or make it user controlled and go all the way in with a tab.

The ?/? in zoom isn't very clear that that is editable. I think the ? should go in a text box.

I think it would be nice to see a custom scrollbar as well for the menu. It seems to distract from the look of the website which I like a lot.

Very cool overall. One thing you might consider for monetizing is https://www.theprintful.com/ or a similar website and let users get posters/prints and such directly from your site and all you have to worry about is integrating an API.

1) It's zoom forward on many mac computer touchpad, the max zoom isn't 100% but is currently limited to a good memory performance level, we will add "read" 100% inspection

2) We've seen people requesting RAW image editing, and technically it's interesting for us to find a way to support that, this could be an open source effort. On the C/C++ community, dcraw.c already support all brands of cameras for RAW.

3) Nice suggestion on the controls!

4) That was a dirty hack but good to hear that people're using it..

5) We will add that.

Thanks a ton for all the suggestions!

For people who get stuck on the homepage due to CSS issues

To go to the editor directly: https://www.polarr.co/editor

To view it using any browser: https://www.polarr.co/editor?force_browser_support=1

Slightly off topic. I have yet to be convinced that non-native applications is the way of the future? It seems like the current trend is to make everything run in the browser these days. I don't really have an argument against it but I just can't quite buy into it yet.

Probably cloud app is better for on-the-go solution and native app is more straightforward for batch processing? I've also just installed Polarr's Chrome App from https://chrome.google.com/webstore/detail/polarr-photo-edito...

From my developer point of view, it is mostly distribution. It is not necessarily non-native or native but whatever it takes to make updates, on-boarding, and selling/free trails easier would attract developers and perhaps going to save them time and energy so they can focus more on building. I think thus the end users might expect higher quality works from non-native apps. But the platform web apps is building on is still lacking in terms of the "app box" metaphor as well as access to local file systems, gpu and guranteed amount of memory etc, which makes web-wrapped apps (like Slack, hipchat? dunno) appealing.

Wow, this is awesome. I made a WebGL image editor as a side project (http://photogl.net) so I know how much work it takes. I'm impressed. I'll be really excited if/when you get 16bit RAW working.

http://www.photogl.net/ does not work

Huh, without the www it works. Gotta check that...

Thanks! Your work actually inspired the team a lot and made us think what we are doing are in fact possible. The saving/outputting part of photogl.net seems to be prone to crash. Are you planning to continue the project at anytime? If you're interested in chatting, please reach out to xiao@polarr.co!

This absolutely blew me away. Brilliant performance, excellent introduction tips, no sign up to get started, just drag a bunch of photos on top of it and you're tweaking away.

I'm definitely bookmarking this.

Just out of curiosity, how did you overcome slow HTML5 canvas' getImageData and pixel reading from an image affected by CORS (tainted canvas)?

i am interested in this as well.

the app is awesome btw!

with WebGL?

WebGL textures suffer from the same problem as shaders can be used to extract data. WebGL is hence using CORS as well. Do you say the performance penalty in WebGL is non-existing contrary to canvas?

what data do you want to extract?, all postprocessing effects are made with GLSL shaders, you don't process data with javascript algorithms.

Just the call on getImageData to extract texture and read all pixels from an image can take 15s if CORS is in the play...

you don't call getImageData to extract a texture data. You load the image, make a texture of it, and then apply all postprocessing effects with a WebGL shader.

OK, thanks for the hint! I'll try to play with CORSed image texture binding to see what the performance is.

Very impressive online tool. Looking forward to being able to use it on my cell one day.

I am getting motion sickness with all the swoopy swoop swoop. SO MUCH SWOOP!

it sure is. could use some design change...

Wow! I tried a few things and I love it. Really snappy.

Very cool. Is there a plan to allow third party integration like Nik Efex Pro?

Nice. The UI labels are a bit tiny though, hard to read.

Wonderful idea and Wonderful App!

I love this, looks amazing!

love it.

very neat! love it!

A wonderful idea and wonderful App!

IE 11 supports WebGL, any technical reason for browser sniffing?

I tried it with the flag, and it works fine in IE 11. (https://www.polarr.co/editor?force_browser_support=1)

The support is incomplete and we're still evaluating memory issues with IE as well as their toDataURL. We're two developers and as soon as we have more bandwidth we will start to support IE.

Reply on news.ycombinator.com