[UPDATED] Introducing Amplify Bloom – Industry Level Post-Process Bloom

I must admit that I haven't taken a look recently.
I've chosen Amplify because the bloom (and the AO) effects are really beautiful, have enough configuration for most cases, are fast and are easy to configure.

Last time I looked at the Unity post-processing they weren't as good as the Amplify products.
But you're right, if I want something that won't break anytime soon I should stick to Unity.
It's just that sometimes what they provide isn't good enough, but they improve, so I'll take a look at what they can do now.

 

Any news on when the github version will be available? I've been using Amplify Bloom for years, and honestly kept trying to replicate the results of it with things like the unity PPS2 but never could get anything that looked as good and lacked flicker.

I'm running into some problems on vulkan with linux on certain older GPUs, and I'd love the ability to go in and debug things on my own. I definitely respect the reason why you guys had to deprecate this product, and I even more appreciate the fact that you're open-sourcing it.

 

Thank you for your support. We don't have a specific ETA yet but it should be available soon, we will post details here as soon as possible.

Thanks!

 

Concerning this issue, I figured out what the issue was. When activating tone mapping the actual histogram output is completely changed. Amplify bloom taking the data after that pass it does not find any value getting under the threshold.

You should use the "Debug" property to display the "Main Threshold" data and adjust it to a much lower value when using tonemapping. It allowed me to use both effect in conjunction and needless to say amplify bloom is way better than unity PPV2 bloom.
When Amplify bloom goes to open source I'd really like to add a PPV2 module with all the parameters so we can use one profile to handle every effect. (Adding a module for Hx Volumetric light is also something we'd like to do)

 

Unity's PPV2's bloom instabilities are killing my game, has this thing hit GitHub yet?

 

I'm afraid not, planned for future additions.

 

Any Update on the GitHub Release ?
Best

 

Not just yet, thank you for your interest.

 

Please, at the very least, share a direct link to the asset as it was. You can't just retire an asset this good without alternative for over a year!

 

I'm afraid we can't do that at the moment; we will share it as soon as possible.

 

If you bought a copy of this previously, you can go into your downloads and get the last version before deprecation from the asset store. I'm still actively using the latest version of their asset as it was. It's only if you never purchased it that you would not be able to get it.

 

If it was on the store, even with ZERO support I would buy it. I have a huge problem with the bloom in the post-processing stack causing all my lights and emissive materials to flicker in the distance. It's a total buzz kill.

 

Thanks folks, we definitely appreciate the interest. Hope we can offer something in the near future.

 

I do agree that I have not found a suitable replacement for Amplify Bloom, and I've bought most of the major bloom solutions on the store, as well using as the old and new post processing stacks. There's just not anything else that does as good a job, period.

That said, I think that some of the temporal AA solutions may help quite a bit, like CTAA by Livenda as one example. I bought v2 of that three or four years ago when it was really expensive and prior to it being on the asset store. At the time it was a bit janky (most specifically with motion vectors on moving objects), but did do a good job.

They've been really updating that asset a lot in the years since then, and it seems like the newer version that they have on the asset store is getting a lot of good attention. I have not been in a situation where I need to use TAA for the last few years, so I haven't picked it back up.

I understand that unity has TAA in its deferred pipeline now, too. It felt underwhelming the last time I looked at it, but I'm largely using the forward pipeline and have just not had my focus on the in-depth scenery like I was working on in 2016 and 2017. So grains of salt on my advice. But there do seem to be other options out there to try, beyond just amplify bloom.

 

It's true TAA helps with the flickering but CTAA would have to be A LOT better than unity's solution cause it still flickers a lot.

EDIT: Oh wow, just checked their video... thank you!

 

I am reluctant to fully endorse CTAA, as I have had very poor experiences with Livenda as a company, with several of their products. This was years ago, however. That said, my experiences with CTAA itself were as follows, with a far older version of it than they currently sell for about a 1/10th of what I bought my version for:

1. It was indeed vastly better than what unity has for TAA. Specifically, I had their version of TAA before unity had any TAA at all -- this was 2016 -- and when unity introduced TAA I was massively underwhelmed.

2. It was -- again, in 2016, and an older version -- a real pain in the rear to set up for moving objects, because the motion vectors and such had to be tracked by separate scripts on each object with a mesh. I absolutely loathed this, and never felt like it worked all that correctly. I felt like it was clunky and slow. My understanding is that they have completely reworked how this functions, but I have not looked into it in any detail.

3. For non-moving objects, aka things that were flickering because of either just random floating point issues, or edges, or the movement of the camera, this is where it was doing a very good job. But at the time -- 2016, I need to say one last time -- this was a very heavy process. I was able to have it running on upper midrange PCs of the time period, but I had serious reservations about its suitability for VR. And there was no question of mobile or some consoles. My understanding is that they have resolved all of these issues, too.

My bad experiences with Livenda are mostly from the 2015-2016 time period. They had created a number of cutting-edge technologies, some on the asset store and some off it, before anyone else had them in the unity ecosystem. Or they had better versions of them. However, they somehow seemed to kind of only 98% work, often got deprecated and then re-sold as a new package, were not responsive to support requests, and in general felt like they were heavier on the CPU and GPU than a polished final product might be, but then again this was the bleeding edge.

In the time since I've had any interaction with them, they seem to have upped their presence on the asset store, have seemed to be using unity's own pipeline expansions in a smart way, have lowered their prices, and I have no idea about support. They're still around, so I'm guessing they are doing something right.

 

Thanks for all the info. Just checked their reviews, it seems they still have support issues but some swear by it and most complaints are about the new render pipelines which we don't use. Also, they compare the performance cost to FXAA for CTAA nxt V2 and V3 is comming. I think this will work for us.

Thanks again!

 

My pleasure! And holy cow if their performance costs are along the lines of FXAA now. I may need to look into that myself sooner than later.

 

I know it's been asked many times now but are there any confirmed release dates on when the GitHub version will come out?

If not then are there any estimates on when it will come out?

 

Sorry, not at the moment.

 

Realistically, it may be a good idea to focus on a different solution. Amplify Bloom is amazing, but it's a codebase that is not being updated, and it also has quite a lot of draw calls (screen grab and write passes) that help it achieve its effect.

I agree that many of the other solutions on the market are not up to snuff in terms of the temporal anti-flicker, but at this point that is a problem beyond bloom that is being solved. You have TAA from unity (which is so-so), and there are some TAA solutions on the asset store. Beautify 2 has some bloom and some (IIRC) TAA.

I am a huge fan of the Amplify product lines, and if they had a new bloom product that they were going to maintain and that they were excited about, I'd be buying it right away. But there are also some really excellent other developers on the asset store, and I think that putting too much pressure on any one of them is likely not a good idea.

If your bloom flickers, then probably you also have some other flicker that would be solved by TAA or DLSS or whatever your preferred approach is. I don't think DLSS is really feasible right now for most purposes, but still.

As someone who has Amplify Bloom and has loved and used it for 6ish years now, I recognize that it's probably time to move on, and I'm looking at alternatives. For other folks who don't have it and are looking to get into this on github now, I'm not sure that's... where your focus should be. It's not up to me to tell me your business, but something that you can't have is always more exciting than something you do have.

 

Indeed; Amplify Bloom it's not ideal to wait for this solution for commercial projects. Definitely explore the many choices out there.

 

Many of the alternative choices for bloom out there are just not good enough compared to Amplify Bloom. In my opinion they're either too simplistic or have too little customizability options compared to Amplify Bloom. Believe me, I've seen all of them. Also I don't really work on any big scale projects, just small ones that I do in my spare time because I'm not much of a professional myself lol.

But anyways, thanks for the response!

 

Amplify Bloom needs to get back on track!! There is nothing comparable out there and I am working in the industry.

This is a waste...

you can trash all the other blooms out there! no matter which one you wanna show me... I also know them all.

 

Definitely a shame we don't have the resources to bring it back, it will be available via Github at some point.

At the end of the day, we lose money working on Bloom; as much as we love it, we had to pull the plug. (we kept it going at a loss for quite some time)

 

That's because of the fact we buy an asset only once, and then you have to update it for dozens versions of Unity.
I can understand that it's not good for business.

I wouldn't mind having to re-buy the asset about every 3-5 years (I have no idea what time frame would be realistic). The price is very reasonable and as others said, there's no equivalent to that outstanding bloom.
Maybe you could make a 'New Amplify Bloom' asset, working for built-in, URP and HDRP.
We would have to buy it so you wouldn't work for free.

 

Yeah, that's why most of these plugins end up being deprecated, and engine changes of course; it's just not viable, nor do enough users want to support it when a free alternative is provided. A new Amplify Bloom for URP/HDRP would be a big investment on our end, given our limited resources we've bet on other more future-proof solutions for future assets.

At the end of the day, the demand is just not there, this will probably work better as an open project on Github. Believe me when I say that we didn't make this decision lightly, we're very proud of Amplify Bloom it was a shame to pull the plug which we avoided doing for a long time.

 

Unity has out and out bought some asset store providers, like Cinemachine or ProBuilder, but being folded into unity's staff and structure is probably not on the agenda of every solid asset store developer.

Given the immense value brought to unity as an ecosystem, if there were some sort of subsidization/grant process from unity to some of the more popular assets that need ongoing development, it seems like that would benefit them while still allowing companies like Amplify to remain independent. Your choices shouldn't have to be "get acquired" or "do everything way beyond the point of profitability" or "constantly deprecate cool things."

I would argue that this robust system of independent asset store developers, of which I would put Amplify at the very top (but not remotely the only such company) is a gigantic advantage that unity has compared to unreal.

 

Indeed, we'd rather be independent as we have multiple teams working on different projects; some not specific to Unity assets.

Thanks, we appreciate the vote of confidence. I wouldn't be surprised if they came out with some sort of grant system similar to what Epic does, Unity understands the value of the Asset Store community. We'd definitely consider using something like that, our Unreal team has actually been awarded a grant to continue working on a Unreal Asset, it works.

 

Ah, I had not noticed that Epic was giving out grants already. I knew that that was happening in the Oculus space, but it makes sense that Epic is following suit. Hopefully that continues into the Unity space. Glad that things have worked well for your company with that process in the Unreal space.

 

More common to games than Assets but they are open to it. That's one of the reasons we like to stay independent, we can have different teams for each engine and explore both at the same time. I should point out that our Unreal team is focused on art mostly, our R&D gears are powered by Unity

 

Assets are such a broad term, and really does a disservice by not disambiguating. My background prior to going indie developer in 2009 was B2B business software, and we just called it "middleware." Unity itself is technically middleware, kinda-sorta, but it's also a platform kinda-sorta. What you guys make is middleware, like SpeedTree or Umbra or PhysX or many other companies.

OS manufacturers have long used middleware engagement as a sign of health in their platform, and when you have specific sub-products that are plugin-rich (like Blender, Sketchup, Photoshop, basically every music piece of music production software ever, Microsoft Office, and on and on), then the same sort of thing applies. If there's a healthy middleware market, then there's a much healthier final-product market, and companies are much more likely to migrate to a platform.

I used to have so many business clients who relied on Excel for huge amounts of things partly because there were so many plugins and templates and VBA scripts that did hyper-specialized calculations for them. Excel was already something they would have bought and used, but because of the middleware that was around, prying Excel out of their hands was almost impossible.

When it comes to all the other forms of assets -- art or foley sounds or whatever -- I think that also has its place, but it's a really different proposition from middleware. Game templates, which are great for teaching new developers but get a bad rap because of asset flippers are another thing.

I don't mean to be dragging this thread off topic, but we're looking at a really prime example of the system NOT working properly when it comes to Amplify Bloom. I pay for subscriptions or ongoing maintenance for all sorts of middleware, and a lot of it has much more niche applications than bloom -- and yes, also has free alternatives. Not being able to keep up with Unity's changing pipelines is absolutely a valid reason to stop developing a particular product line, but seeing the best bloom solution on the platform die on the vine because of the way the store ecoysystem is structured is frustrating and unnecessary.

I think a lot of the people in this thread imagine that once you guys pop the code on github, they'll just be able to use it and hooray it's going to do exactly what they want. The reality is more likely than once you guys release it, it's going to take some other graphics pipeline experts a bunch of time to figure out what you did and make it work in the new pipelines. If they work for a large company, they'll keep that internal. For everyone else, there's a good chance they won't be able to ever migrate it to the newer pipelines.

 

Indeed, the line between Asset/Middleware gets a bit blurred at times. I understand what you mean but considering the changing playing field that has been Unity for the last few years, I personally think we also have to be aware of shorter, or version/renderer locked, solutions.

Not all products will be maintained indefinitely, this is something that users need to be educated of and account for. It's also critical that the publisher communicates this to the users as clearly as possible; as long as that is done we should be alright. In our case, we make it clear that our shader editor offers long term support, while other packages such as bloom cannot.

I truly hope most users don't expect that Amplify Bloom will "magically" work with SRP's.

Amplify Bloom was never made to work with URP/HDRP, nor was it made to work with the Unity Post Processing Stack; built-in renderer only. The version that's going to be made available on GitHub in the future will not support the new SRP's, it's on anyone interested in expanding it to adapt and share it.

 

I was using this asset on Unity 2018.4 and it is just perfect, I love it. But it does not work well with Unity 2020.x.. Is there any workaround or suggested replacement?
I love the glare/flare effects applied to light as textures, I cannot find any other automatic replacement for it. It was amazing for vehicle and street lights.

 

Wow all these years later and still nothing tops Amplify bloom.I speak for all when I say we would buy this asset in an instant.

Default post processing bloom is truly abysmal!