@OneTimer1

Quote:

"Your problem is simple. The Amiga can’t do Midi.” However, there was one rather serious bug that many people reported — that sometimes when recording Midi input data, Music-X would lose data — occasionally a note or control signal would be “dropped” and not appear in the final recording […]the problem of why Music-X (and other Amiga music software) seemed to have a problem recording Midi data. […] It turned out that the Amiga’s four timer chips were interfering with the serial port. Both the timer chips and the serial hardware were interrupt-driven, and the timer interrupts were a higher priority than the serial interrupts. Worse, the Amiga’s serial chip only had a 1-byte buffer — which means that if you didn’t pick up the data before the next byte arrived, the data would be lost. […] mitigate the problem somewhat by having AmigaOS turn off timer chips that weren’t in use. Unfortunately, they couldn’t all be turned off — AmigaOS needed one timer, and Music-X needed another. Turning off two of the four timers greatly reduced the frequency of the bug, but didn’t eliminate it entirely.

https://dreamertalin.medium.com/music-x-b4abc68d6f78

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Music was what got me into the machine in the very first instance and I never had a problem with dropped midi data, even on an A600.

Having said that, I was using midi compatible trackers like OctaMED, so maybe the situation was different with notation based software.

_________________
Doing stupid things for fun...

@Karlos @Kolla


It seems as if it was only a problem in MIDI recording

@OneTimer1

I still used a controller keyboard to capture live performance without issue. Maybe it was less of an issue because I was running on an accelerated machine by then.

_________________
Doing stupid things for fun...

@Kronos

Quote:

OCS/AGA games on a naked A1200 felt like a decade old compared to PC hence few good ones were made and even less bought as most of the penny pinching MagicPack customers had no budget for SW.

It was only good for basic platformers. Of course with AGA it could do 16 colour background and foreground layers. Which doesn't sound so much but did look pretty. With improved sprites. 256 colour scrolling games possible. But at that point it was taxing for the CPU and the blitter didn't help.

Quote:

A "cost reduced Draco", so 68030/40 with ISA slost and off the shelf PC parts would have been compatible enough with productivity SW and it would have been close enough to a PC so they we might have gotten more game ports at a better quality then what we got for AGA.

They would have needed to market it as an Amiga Workstation. Say with RTG and audio cards. Plenty of Amiga RTG and audio cards at that point. Pros were using RTG and simple audio cards with only one stereo 16 bit channel. They didn't need multiple DMA driven hardware channels to produce pro audio. Just CPU power.

This would have outed the legacy gamer user base and real Amiga enthusiast. But a lot of Amiga people were already ditching AGA if they had it for PC video and audio cards. So not ditching it for real Amiga cards, but ditching it for PC specific PCI cards. The Amiga users themselves were writing on the wall how they wanted to replace their Amiga piece by piece with PC parts. At least those who used the Amiga for real use until the end. It would have been a total break, at least from the classic chipset, but for real use it would have been acceptable in the long run I think.

I used to run A1200 on a TV and really needed a monitor. So I ended up replacing it with a VGA monitor and patched the system with VGA hacks. I mostly got by but had to give up most games. Early start up was an issue. It was an early monitor so a quirk meant I could see half a 15Khz screen on VGA at half resolution. There was an Aminet project to scroll the screen but later VGA monitors refusing any lower signal killed that off. I must have replaced the A1200 with the half screen hack with an AmigaOne as I never did end up buying a proper Amiga monitor.

@kolla

According to that the 440 has no direct support so needs to trap and emulate. Rather like the 68K would need to in the past. Then the Sam 440 can only run e440 code and not any other PPC code, so has that limitation, but fine for compatible code.

The X5000 with e5500 core supports KVM HV. Being an E core it can only run e5500 compatible code. But having HV is useful.

The X1000 with PA6T supports KVM PR. Unfortunately this looks to be inferior. Despite the PA6T being superior to the e5500 in a few ways by being a BookS core with AltiVec and beating a G5 with little endian support. The X5000 has seen a lot of development lately and the X1000 looks to be almost left behind or in the shadow. Including in the Linux world. So the X1000 could be an untapped resource, by being capable of running little endian Linux, but lacks support and GCC cannot compile LE kernel for it.

https://forum.hyperion-entertainment.com/viewtopic.php?t=4225

@ppcamiga1

Dumb framebuffer is not an accurate way of describing it. You forgot the copper and sprites. Plus blitter for assistance. Framebuffer also implies packed pixels in most cases. The nearest AGA with 256 colours of 8 bitplanes comes to a linear framebuffer is a linear bitmap. And an interleaved bitmap where all lines appear in sequential order would be the closest to a linear framebuffer.

Even VGA, with its direct chunky pixels, isn't always linear. The framebuffer is split into four colour planes. So each RGB has a colour plane. There is the common 256 colour mode using a linear frame buffer. But there is also the other 256 colour mode, where pixels are split across four planes, which apparently Doom uses to write pixel columns quickly. So the rendering speed was not all based on simply chunky pixels but hardware tricks for pixel parallelism.

Having brought up Doom. I ran Doom on my PiStorm Lite A1200 setup using standard AGA 256 colour mode. Even with the overhead of chunky to planar, with no native chunky mode, the speed was amazingly transparent. It easily met the the Doom FPS standard of 35. It ran so fast it looked like a Pentium with a fast VGA card. It was amazingly smooth and the way it was meant to be. And that was using just a standard RPi3 and not RPi4 that's become the minimum people want now.

Last edited by Hypex on 17-Jan-2024 at 03:00 AM.

@Hypex

Why do you bring up little-endian?
Most obvious is to run OS4 on Qemu/KVM on old G4 mac mini.
(I wasn’t even considering exotic AmigaOne systems, I don’t think of them as “actual PowerPC systems”, but rather as dev boards for embedded.)

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@Hypex

aga is dumb framebuffer nothing more.
aga sprites are still 16 colors therefore useless in 256 color games.
aga copper is as slow as ocs.
copper chunky take all aga bandwith and work in low resolutions therefore useless.
no reasons to keep aga.
it is nothing special and may be replaced by anything.

@ppcamiga1

You are like a broken record! We know that you love ECS and no further and that you cheated on 68k and the ECS custom chip buy forcing yourself onto your new love the PPC CPU with all it's similar Endians but with some more power. She did not appreciate your advances and is now feared obsolete/MIA!

IT'S ABOUT THE SOFTWARE STUPID AND AGA SOFTWARE IS MORE USEFUL AND FUN THAN PPC STUFF!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@Kolla

OT but worth to be mentioned:

Midi seems to work over FS-UAE on Raspberry 4

Tutorial:
https://fs-uae.net/2020/04/13/using-fs-uae-on-linux-with-real-midi-hardware

Maybe it will be interesting for some users.

some people don't accept reality.
aga in 256 colors is just dumb framebuffer nothing more.
no reasons to keep it.

@ppcamiga1

Don't you get iy yet? Nobody else cares what you think.

_________________
Doing stupid things for fun...

@ppcamiga1

Quote:

ppcamiga1 wrote: some people don't accept reality. aga in 256 colors is just dumb framebuffer nothing more. no reasons to keep it.

That's why you should get a pistorm and use RTG, on a plus side you also get a mouse and keyboard interface for rpi.

Last edited by pixie on 18-Jan-2024 at 12:08 PM.

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga

Does anyone else remember Mega_RJ saying he was leaving but would be back in his final form that no-one will suspect? (or words to that effect)

This wasn't long before ppcamiga1 slithered out of a hole.

Couple that with it being impossible to be as stupid as he behaves.........

I cannot believe this is not just a character someone is playing for sh@t stirring purposes. Anyone who is that challenged in real life would likely be supervised when using a computer, and anyone supervising would discourage the sorts of posts he excretes.

Then there's the zero effort he makes vs. the effort he used to put in. The one dimensional nature of the character must make it boring to play by now, hence the low effort and repetition.

Not sure which I'd prefer for the person behind it though. It'd be a little sad to have such little going on in life that a person even entertains the idea of playing this character, but no less sad to actually *be* that character.

@fishy_fis

Quote:

Then there's the zero effort he makes vs. the effort he used to put in. The one dimensional nature of the character must make it boring to play by now, hence the low effort and repetition.

Yet he keeps quite a lot of people busy for a looong time. Maybe he *is* a loony, but hey, he probably and obviously knows how to bait a handfull of characters here, the usual suspects, and keep them hooked.

One can simple ignore such a character, whereby he becomes just a nuisance in a thread and it is a proven strategy to keep one from posting over time, even a loony or mentally challenged person.

Just ignore this post...

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@kolla

Quote:

Or maybe it was research and education - even universities here in Norway had tons of Apple machines and they were used for just about anything.

I'd agree. Apple had a big steak in the education market. Even in Australia schools were filled with Apple micros. Later with Macs but they were more expensive so not as cost effective. But it did help when the iPhones came out as it helped to embed the Apple brand in peoples minds from an early age,

@Hammer

Quote:

In 1992, Cubase was released for Windows 3.0. In 1993, Cubase Score for Windows 3.11.

With PCI MIDI cards the ST would have been gone after that.

Quote:

Shapeshifter with RTG runs well on PiStorm-Emu68 accelerated Amiga.

Didn't think of doing that. Good idea. Faster than a real Mac until a MacStorm comes out,

Quote:

I have a fast switcher between RTG (HDMI-to-VGA) and Amiga RGB (VGA).

That brings me back to the 90s. Needing a switch box. It tends to be popular to convert RTG to VGA. Though I wonder if it would lose quality that way. Of course even that way needs a monitor to support 15Khz unless a messy scan doubler is involved.

@OneTimer1

Quote:

The Amiga could actually do MIDI, the ACIA in Paula could run the right bit rate, you just need a few parts converting the signals of the RS232 into MIDI compatible signals.

It could but it wasn't exact. Like most bit rates it was approximated by the hardware timings and couldn't run at the exact MIDI baud rate. The ST is known for it's timings. On top of this there is no DMA to assist here. So any CPU tax will slow it down.

Quote:

You can easily solder such a simple interface by yourself:

Ha. That's actually a pretty involved circuit. The average user needing to do that would explain how the ST got ahead with it built in. Of course there were boxed MIDI interfaces. No idea if music shops stocked them back them.

@kolla

Quote:

Why do you bring up little-endian?

A few reasons. I found recently the G5 lacked endian support which because a point of contention. I don't know how it lacked as PPC has three hardware tiers for LE; LE mode in MSR register, MMU pages and LE instructions. The G5 was an IBM so perhaps Motorola performed a Motorolaism and removed the instructions. But it caused a problem with VirtualPC which needed a special G5 build that was technically inferior and slower. Unlike x86, PPC doesn't have a bswap. It only has swaps at the memory access level where it's relevant, but take that out and fancy rotates won't save you. I've laboured my point.

The other reasons would be that Linux has become a backup to OS4. Realistically it would be more capable for daily tasks. But now Linux on AmigaOne hardware has faltered and browsing is suffering due the same endian issues as for OS4. The difference is that it needs way less work to recompile it on Linux since depends can be met. PPC has a breath of fresh support with PPC 64 LE. With Debian and Ubuntu now supporting PPC again with ports. But, despite any CPU support, there is no LE Linux for the AmigaOne machines. Which, like IBM did, putting the CPU in LE modes solves a lot of endian issues for common software designed to run only on LE as is the case now days. It's also why PPC graphic drivers have terrible support for anything above a HD6000. No 3d, no acceleration and only a dumb framebuffer available. No BE support in hardware any more so driver must run in LE and there is too much work fixing a driver to be portable.

Quote:

Most obvious is to run OS4 on Qemu/KVM on old G4 mac mini.

I don't know why the Mini is so popular. Moana ran on that as well. But I never had one, never knew anyone who had one, and only had iBooks myself. So a laptop is better target for me. Also since the lesser PineBook failed an iBook makes an obvious target for the long desired laptop.

Quote:

(I wasn’t even considering exotic AmigaOne systems, I don’t think of them as “actual PowerPC systems”, but rather as dev boards for embedded.)

Well, the AmigaOne XE and X1000 have desktop class PPC, the rest are embedded. All would be dev boards. I've always been interested in running OSX and MorphOS on my X1000 and if Qemu would help. But I'm not interested if there is a lot of work compiling it myself and the end result is emulated anyway.

Last edited by Hypex on 18-Jan-2024 at 02:57 PM.