|
The memory remains by Unchained
[nfo]
|
||||||||
|---|---|---|---|---|---|---|---|---|
|
|
|||||||
|
popularity : 52% |
|||||||
alltime top: #26742 |
|
|||||||
|
||||||||
added on the 2017-07-28 15:37:20 by Asato  |
||||||||
popularity helper
comments
very nice! i'm a sucker for skylines ^^
rulez added on the 2017-07-28 16:11:55 by sensenstahl 
+ very nice design for an 8k
+ I love background, moody raster skies and sphere mapping bouncing balls
+ for adlib music
- I am a bit puzzled with speed. I have the same issue with your previous intro. I haven't tested on 286 yet, but it seems that even in full dosbox cycles or a faster 386 or Pentium 3 I tried, the speed never gets a smooth 60 or 70fps. Are you locking the speed in purpose or maybe use a timer that is not accurate enough? Even if it's made for 286, I'd love to go to faster PCs to enjoy this in smooth frame rate.
+ I love background, moody raster skies and sphere mapping bouncing balls
+ for adlib music
- I am a bit puzzled with speed. I have the same issue with your previous intro. I haven't tested on 286 yet, but it seems that even in full dosbox cycles or a faster 386 or Pentium 3 I tried, the speed never gets a smooth 60 or 70fps. Are you locking the speed in purpose or maybe use a timer that is not accurate enough? Even if it's made for 286, I'd love to go to faster PCs to enjoy this in smooth frame rate.
yes, I lock the speed, waits at least 65 microseconds in order to not to run too fast on modern hardware, even on my i3 it still runs decent with real plain dos, I could remove it but only vsync wouldn't slow it down enough, so on faster machines I guess it would suck.. the first screen could be much more faster but I slowed it down to the same speed of the second screen which is very CPU-intensive (about 3600 pixels for the sphere that are computed each frame), and since its mode 13h the double buffer has to be copied too (tried modex but its equal or even slower in speed, 'cause plane masking requires too much cycles for the sphere) .. hm, but I'll think about how to realize better fps !
.. 65 ms ..
borked dl url thumb. too bad, the description sounds promising.
@TheMechanist: I see, thanks for the answer.
65ms that would make it 15fps.
Solution 1, optimize for the lowest target machine (my art :)
2, maybe use a timer for the movement
3, provide a question or autodetect, change movement increments depending on the speed (I know, can be tedious)
Though I want to make something for my 386dx with fastest card now, so I have to also see how to handle best for similar dx or sx with slow cards. I used to like to vsync and then add stable increments on movement (better for smooth scrolling, movement) at least in my old demos. I need to see how to read stable accurate timer and depending on the elapsed time decide whether to vsync and increment or base movement on time (but will make code more complicated). But if I target to get vsynced full frame rate for my 286 Amstrad laptop with very slow gfx card, I will be definitely leave it as is, anything would probably run faster than this.
65ms that would make it 15fps.
Solution 1, optimize for the lowest target machine (my art :)
2, maybe use a timer for the movement
3, provide a question or autodetect, change movement increments depending on the speed (I know, can be tedious)
Though I want to make something for my 386dx with fastest card now, so I have to also see how to handle best for similar dx or sx with slow cards. I used to like to vsync and then add stable increments on movement (better for smooth scrolling, movement) at least in my old demos. I need to see how to read stable accurate timer and depending on the elapsed time decide whether to vsync and increment or base movement on time (but will make code more complicated). But if I target to get vsynced full frame rate for my 286 Amstrad laptop with very slow gfx card, I will be definitely leave it as is, anything would probably run faster than this.
uups, never thought about it, that this decreases fps so much, just checked how fast/slow it runs on slowest target and increased wait time to see it at the same speed on 1.5 GHz machines :-) .. but he - from 16 MHz to 3 GHz same speed :-) .. btw. if you sync to music like in this with two parts, its getting even more complicated, should have used song playing timing .. hm, and hard to realize, if code size matters .. okay, challenge accepted for next prod ;-) ..
Thx @ Sensenstahl for fixing the dl link.
Pretty cool tiny prod, nice night transition too
Pretty cool tiny prod, nice night transition too
:)
updated video capture, @optimus: updated file (yandex dl url) with decreased speed lock-in in part 2, but just a little bit, if it's decreased too much, it's getting too hard to read the scroll text on the sphere even on 486DX40 ..
TheMechanist: I simply included your updated .zip into the 1st version. So no need to fiddle around with different hosts :)
Nice little intro. :)
@Sensenstahl .. thx !
=)
submit changes
if this prod is a fake, some info is false or the download link is broken,
do not post about it in the comments, it will get lost.
instead, click here !