home *** CD-ROM | disk | FTP | other *** search
- Path: news.eunet.fi!fipnet!kone!jsaarinen
- Newsgroups: comp.sys.amiga.programmer
- X-NewsReader: IntuiNews 1.2b (31.7.94)
- References: <38231935@kone.fipnet.fi> <4d15vb$48g@sunsystem5.informatik.tu-muenchen.de>
- From: "Jyrki Saarinen" <jsaarinen@kone.fipnet.fi>
- Date: Thu, 11 Jan 96 16:30:32 UT
- Comments: Illegal date header - new date added by quicknews
- X-Original-Date: Thu, 11 Jan 96 18:12:30
- MIME-Version: 1.0
- Content-Type: text/plain; charset=iso-8859-1
- Content-Transfer-Encoding: binary
- Distribution: world
- Subject: Re: C2P
- Message-ID: <38231988@kone.fipnet.fi>
-
-
- > |> It does two passes almost free even on a 28MHz 020/030. Completely
- > |> free on a 40MHz 030. The blitting time was quite small if
- > |> I remeber correctly, the whole time was 17ms, and the CPU
- > |> part is ~5ms, so ~12ms blitting.
- >
- > 5ms is the time for copying a 2x2 frame.
- > 10ms is the time for one blitter pass.
-
- Well, something around those figures.
-
- > The values seem to make sense, but isn't it rather 7ms for cpu doing
- > passes not completely free, and 10ms for one blitter pass ? ;)
-
- Nope, starting from a 40MHz 68030 it does those two passes free.
- Almost on a 28MHz Amiga, and in real-life situation (8bpl lores)
- I could propably claim that even a 28MHz 020/030 does those
- two passes free.
-
- > BTW how much is the 2pass routine on 020 ?
-
- I have not tested.. the old one was 20ms (7ms on 030/50),
- I think the new 2pass routine could be about 10-15ms on a
- 020/14. I have to test it..
-
- > My one pass routine needs 10ms for a 2x2 screen, so 2times more than
- > copying, I hope doing 2 passes is ~15ms and not ~20ms...
-
- Yep, Luddes one pass routine took also 10ms on 020/14.
-
- -- _
- a Stellar programmer _ //
- "Amiga - back for the future" \X/
-