Jump to content

White Water by Endprodukt


Recommended Posts

Just now, DJRobX said:

You don't HAVE to match on multiple layers to utilize this!

I'm not sure what this means - can you explain?

Just now, DJRobX said:

CPU performance is the only concern I see with it as entering LCM means you're doing all the same "global" hash calculations as before, but now adding a new round to check all the local masks against all the LCM frames.  

Earlier conversation in this thread lead me to believe this method was more efficient (faster) or less taxing on resources, so I hope you're wrong about this!

Link to comment
Share on other sites

I've been using this colorization to add support to Freezy's.   Very nice work, Endprodukt!!     The LCM scenes look great!   

I think this new mode is going to be a huge step forward for pin2dmd colorizations.   I can see it even being helpful to "Contain" basic color mask sequences as it gives you a local set of masks to work with once the scene has started.     That will allow you to make very scene-specific "keyframes" without worry of taking away from the precious global masks.   You don't HAVE to match on multiple layers to utilize this!    The only semi-downside is that you have to have a keyframe outside of it to get out of it, but most thorough colorizations will have this anyway.   Exciting stuff.  

On 4/15/2020 at 5:54 AM, Endprodukt said:

Checking that via live preview only ever works by first triggering the starting keyframe for the LCM. Without going to that one first, you'll end up with no colors. 

Also, the delay for LCM scenes is not adjustable. Although LCM is supposed to basically "overwrite" transitionframes (usually 24ms), it doesn't seem to work for me very well. Using single frames and setting the delay to 1500 has been solving many issues for me, although of course not all of them. 

Right, but you should be able to assign multiple keyframes to the one LCM scene to get back into it (and/or it can be a more generic keyframe).  Once it goes back in, LCM will color the correct frame.

From my understanding of the code, the delay value isn't applied here. The LCM colormask is "sticky" until there is a new detection.   So it should "last", similar to setting a high delay. 

CPU performance is the only concern I see with it as entering LCM means you're doing all the same "global" hash calculations as before, but now adding a new round to check all the local masks against all the LCM frames.    But I don't really know that CPU usage has been much of a problem, I imagine it's mostly of concern on a real pin that doesn't have help from a desktop class CPU.    Time will tell there, I guess.   

Link to comment
Share on other sites

3 hours ago, DJRobX said:

I've been using this colorization to add support to Freezy's.   Very nice work, Endprodukt!!     The LCM scenes look great!   

I think this new mode is going to be a huge step forward for pin2dmd colorizations.   I can see it even being helpful to "Contain" basic color mask sequences as it gives you a local set of masks to work with once the scene has started.     That will allow you to make very scene-specific "keyframes" without worry of taking away from the precious global masks.   You don't HAVE to match on multiple layers to utilize this!    The only semi-downside is that you have to have a keyframe outside of it to get out of it, but most thorough colorizations will have this anyway.   Exciting stuff.  

Right, but you should be able to assign multiple keyframes to the one LCM scene to get back into it (and/or it can be a more generic keyframe).  Once it goes back in, LCM will color the correct frame.

From my understanding of the code, the delay value isn't applied here. The LCM colormask is "sticky" until there is a new detection.   So it should "last", similar to setting a high delay. 

CPU performance is the only concern I see with it as entering LCM means you're doing all the same "global" hash calculations as before, but now adding a new round to check all the local masks against all the LCM frames.    But I don't really know that CPU usage has been much of a problem, I imagine it's mostly of concern on a real pin that doesn't have help from a desktop class CPU.    Time will tell there, I guess.   

Everything you say is true, I could set more keyframes of a more generic one, yes. 

Haven't had any CPU problems on the real machine yet. But I converted some of the LCM scenes (not all) back. They simply worked better in the traditional way.

While it is true that the delay function shouldn't be necessary...it just doesn't work as good in real life. Having the traditional single frame with a delay of 1500 seems to work cleaner as of now. 

 

Thanks for your efforts. 

Link to comment
Share on other sites

  • 1 month later...
  • 5 months later...
  • 2 weeks later...
  • 4 months later...
  • 4 months later...
  • 1 month later...
  • 2 months later...
  • 3 months later...
  • 1 month later...
  • 5 months later...
  • 3 weeks later...
  • 2 months later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
  • Create New...