signaljacker's Forum Posts

  • Thank you, that is a good suggestion.

  • Thanks Ashley, I'm testing on an M4 Mac Mini (base model), I seem to be able to find precious little info about the GPU other than it's got 10 cores. Unfortunately I can't redesign the viewport size to 1080p in this particular project, but that essentially answers my question.

    I'm pretty happy with the performance of the game upscaled to 1080p, but my fear is that anyone using a 4k monitor who wants to get a fullscreen experience (which is what the game would load as by default) is going to get a bad experience and it will reflect badly on the game. Perhaps though I should trust that with more widespread adoption of these monitors there will also be more powerful GPUs driving them. I just wondered how others were dealing with this - 4k gaming isn't important to me, and wasn't even on my radar but I guess I just got a shock when I saw how much of a hit the performance does take, and the average consumer is probably going to flip out that a little pixel art game can bring their fancy computer to it's knees.

    My target platform is PC but I bought a mac to play around with and figured why not release on that too.

  • I've got a 480x270 pixel art game that is using some shaders/effects applied to the layout (necessary for a distinct style that my game requires). I've been happily developing with 1080p desktops in mind, but recently bought a 4k monitor and was kind of horrified at how badly the game ran when scaling up using the "high" fullscreen quality. Now, that might come as no surprise, but as this size monitor is becoming more and more ubiquitous and in the interest of future proofing my game as best as possible I'd like to see what sort of solutions there might be.

    My settings are:

    480x270 viewport

    letterbox integer scale

    fullscreen quality: high

    sampling: nearest

    no pixel rounding

    Shaders/effects used (not all at once necessarily, but these can be switched on and off by the user)

    BetterCRT

    Diskblur

    RGBAChannelSeparation

    The game runs great on the low quality setting, but the issue is that the shaders and text then look terrible when stretched out over a high res screen.

    On high quality the shaders are the performance culprit as switching them off results in decent enough performance. We could just blame the shaders here, but I'm wondering if there's an intermediate solution that can balance quality and performance.

    I know it's kind of weird, but for this particular project I need the best of both worlds - both crisp, perfect pixel art, but also some high quality effects such as blurs, CRT etc

    Digging around the forums I came across this post, which seems to be discussing a similar sort of thing - but the file being discussed is no longer there - it does look like Ashley provided a solution at the end though, so that gives me some hope...

    What would be great would be to upscale the game from 480x270 to 1080p using the high quality, and then upscale that again to 2k 4k etc using the low quality (if that makes sense). Is there a way to do this?

    How are you guys that are publishing games to desktop commercially dealing with future proofing your games?

  • Thank you very much, I really appreciate the detailed breakdown of where this has gone wrong and how it can be done properly. Your solution is so elegant in comparison to what I was trying to do! Thanks so much for sharing your knowledge!

  • Hi all, this is driving me nuts so just wondering if I can get a second pair of eyes on it to tell me where I am going wrong!

    Here's a capx

    What I'm trying to do:

    When spacebar is pressed the emitter sprite generates a projectile.

    This projectile has 4 line of sight behaviours attached to it - I want each of these to pick a random point on the screen at specific intervals.

    When they intersect with a solid they generate a sprite at the intersection point. In addition to this, to represent each line of sight path, 20 dot sprites are generated and these should be spaced evenly between the projectile and the intersection point and continue to adjust position accordingly until a different intersection point is chosen.I'm trying to do this by lerping between the projectile position and the intersection point.

    To keep track of which dots are linked to which line of sight behaviour they are given a variable on creation which is set to 0, 1, 2 or 3 and are picked accordingly.

    The problem is - none of this is working!

    I'm not sure where I am going wrong - possibly the way I am using the viewport to pick positions? The way I am picking things? The way I am using Line of Sight?

    Any help or insights is greatly appreciated! :)

    Tagged:

  • I agree, this is an absolute headache to do in Construct - and something that would be great if the engine did more of the heavy lifting with. It can be done with a combination of variables, local storage and key codes - setting it up is a total creativity killer. I've been learning a bit of Godot recently and I was flabbergasted at how easy it was to set up in comparison!

    Here's one example of how it can be done in C3 that I think helped me: https://www.youtube.com/watch?v=bkdOdAX2mrA

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads

  • Great solution - thank you so much!

  • Hi, I'm trying to set up a system whereby the player can step on a pressure plate and it will open a corresponding door. There is something wrong with my logic, and it's probably something simple, but I can't seem to get it. Only the first trigger seems to work correctly.

    dropbox.com/scl/fi/evyy46ausnqsuwwx4oyf8/multiple-doors.c3p

    Thanks for any help! :)

    Tagged:

  • A lot is going to depend on your style of animation but it's a good suggestion to use the timeline feature (which is basically a version of construct animate within C3 as I understand it). Breaking your art down into parts and animating it there could work, but may also be tedious and if you've already done the animations then it's also double handling (I haven't really explored the timeline features much) - there was some software called "Spriter" which had construct integration and could allow you to set up bones/rigs and animate and then import into construct - this would be a good solution however I'm not sure on the compatibility with C3 and when I was experimenting with it I found the integration with construct pretty fiddly.

    You could also experiment with using the timeline features to dynamically destroy and load in animations as needed - eg if you need a facial animation you could destroy the static face and load in an animated version and then when that's finished destroy that and load a static one back in. This might be a lot of work with events and there might also be visible visual delays as the animations load in if they are really big, but might be worth experimenting with.

    Would also experiment with your export settings, sounds like you're exporting at a very high res, could this be reduced drastically but still look good? Experiment with the scaling settings in the project, you might be surprised at how good smaller assets can still look when scaled up.

  • Hi Liberador I've done up an example of how you might implement something like a car engine changing pitch - obviously the quality of the source material and how well it loops etc will have a lot to do with success, but I think something like this would work fairly well - example can be downloaded here

  • Hi all, I've never really played with the double jump feature in the platform behaviour but have been having a quick look at it, and am I just missing something but there doesn't seem to be an easy way to set a different animation for the second jump in a double jump? This is pretty much a staple in any polished platformer with a double jump mechanic (eg the first jump is a regular jump, the second one is a spin etc). The only way I've found to do it is by tracking it with a variable and then using a vector y on the second jump to simulate the double jump (not actually using the one built into the platform behaviour). It feels like an "on double jump" would be super handy to have. Any tips here?

  • Hi all, I'm doing a pixel art game with a resolution of 480 x 270 and this scales nicely for my needs in fullscreen etc. My issue is that I have an animated company logo that is a 1080p video that I'd like to play when the game boots up. Now straight up I realise this probably isn't wise as there are still many laptops out there that don't support a 1080p resolution, so I will likely need to make the video at least a bit smaller - but I don't want to shrink it down to 480 x 270 because it loses far too much fidelity. I have seen many commercially published pixel art games do this where they have a crisp intro animation/logo at the beginning but then the game itself runs in low res. What I want to know, is if I am switching resolutions on the fly like this (forcing the user's computer to do this) are there any potential issues I should take into consideration and does anyone have any tips on best practices? End goal is just a PC/Steam release.

    I think the product has a bit of an identity problem, being that it is heavily marketed toward absolute beginners - who are probably also the audience least likely to commit to a subscription model. And while it might be beginner friendly initially, anyone who has worked on a substantial project will know that there is a hell of a lot of hard work hidden away there to get something really good done. I generally avoid any software subscription service and have only bent the rules with Construct because I was already familiar with the workflow before the subscription model was introduced - had I not been, I wouldn't have even booted it up to try out. I've been using construct since classic, own a C2 license and pay for a C3 sub, I don't like the subscription system as a consumer but I do understand why Scirra use it as a small business, and while it would be one of the things hindering really widespread adoption, it is obviously working well enough for them at the moment, so I guess we're stuck with it.

  • I think a lot of users probably don't ever check the beta releases - it feels risky to do so when working through a large project, so it's understandable that the vast majority of users will be exposed to a new feature for the first time on it's official release when there is a prompt to update. I actually do really appreciate the ability to download the betas though (and did in this particular case for the first time). Just chiming in too, I think this is good start and can appreciate that there are probably a lot of behind the scenes complexities in getting this to work really solidly. I haven't found a use for flowcharts yet, but if we could build state machines with them then they would become extremely useful. I look forward to seeing how the tool evolves.

  • Hi all, I haven't had time to check out the new flowcharts feature yet in much detail, but I remember some discussion somewhere about it possibly being useful to build a finite state machine to control animations. Has anyone explored this yet? Is the flowcharts system feature rich enough for this and does it seem like a viable solution? Looking at the examples briefly I can see its use for dialogue, but I like the idea of having a visual representation of a state machine to control my character animations rather than having to keep track of everything through actions.