When using an Amiga mouse, the pointer moves smoothly on screen. What I mean is it doesn't move jumping 3 pixels a milisecond.

I have used many USB and PS2 PC mouses and all of them behaved the same with a single exception. The Logitech G7. When moving the pointer using G7, it always moves extremely smooth whatever the resolution is. Does anyone know how this is possible and is there any other mouse with the same behavior?

Last edited by Ancalimon on 22-Apr-2010 at 08:15 PM.
Last edited by Ancalimon on 22-Apr-2010 at 05:10 PM.

@Ancalimon

It's actually fairly simple. Let us start with a mouse with ball, to keep it simple. To get from this to a IR mouse doesn't take a lot of imagination...

The ball rotates against the X/Y rollers, both are read using two LEDs and two sensors.
The sensors tell you which direction and speed the mouse ball has.

The Amiga mouse transmits this information directly to the Amiga which counts up/down and the next time the screen is redrawn the mouse is placed in the new place.

A PC mouse doesn't. It increments or decrement counters and a certain number of times each second the counters are transmitted to the PC, where they are added or subtracted to the mouse pointer position, then cleared. And another certain number of times each second this data is used to draw a new pointer.

If the refresh rate of the mouse is not the same as that of the display, then this will be jerky. But if you increase the refresh rate of the mouse to something much higher than the refresh rate of the screen, the movement will be smooth.

So there are two acceptable solutions:
Sample counters at screen redraw (Amiga aka bus mouse)
Sample counters much more often than screen redraws, then sample the sampled data at screen redraw (good PC mice)

Unacceptable solution:
Sample at frequency different from and not significantly higher than display refresh (most PS/2 and USB mice)

You need to look for Bluetooth mice and USB gaming mice to find the highest sampling rates... Or find a way to make the mouse sample at screen refresh. Which actually wouldn't be a bad idea

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

Interesting stuff.

I find moving Mac to Windows uncomfortable at first as the Mac applies a different acceleration curve to mouse movements. The PC method makes for quick movement, but I find the Mac method to be more precise.

I'm trying to think about how it works in AROS but I haven't consciously thought "oh the mouse pointer isn't moving how I want".

@Ancalimon

I thought you were going to say "the silver coating on the buttons that stopped them from working after about a year"! EDIT: I miss my Naksha mouse though!

Chris

Last edited by clebin on 22-Apr-2010 at 07:17 PM.

@Ancalimon

Quote:

Ancalimon wrote: When using an Amiga mouse, the pointer moves smoothly on screen. What I mean is it doesn't move jumping 3 pixels a milisecond.

Ok both mice are totally different, but more important for a smooth mouse movement is they way how and when the display will be updated. If you update the mouse position in a exact timing raster when a new display is transfered to the monitor you get a smooth mouse display. If other more important operations like sound, video, ethernet or HD access have a higher priority you will get a lot of disturbing events. But think for yourself.

_________________
AmigaONE. Haha. Just because you can put label on it does not make it Amiga.

I borrowed this comments from here (#27 & #28):
http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=38873&forum=2&start=20&order=0

@clebin

The same happened to my Logitech G7. :)

@Arko

I guess the Xena can be used to stop these kind of things from happening?

@Ancalimon

Well, I guess it could, but finding a mouse to use with it (other than something like a classic Amiga mouse a la Catweasel) might be tricky.

Or, like has been said, a good USB mouse... USB has enough bandwidth and low enough latency to handle mouse inputs fast enough to give a smooth display. I personally have a couple of Logitech mice which are both very smooth on any machine I plug them into - my A1 included... I have used awful mice too on PCs which show the effects you describe.

_________________
RobTheNerd.com | InstallerGen | SMBMounter | Atoms-X

@Ancalimon

Not from a USB mouse. As soon as you listen to it over a USB, you are bound to the refresh rate that the mouse can transmit. That rate is also affected by the rate that the USB stack can re-read the mouse, and it isn't going to be locked to the screen refresh rate.

Having changed from a passive, read-only device with switches to an active device connected to a general purpose bus, we are stuck with the latency that such an indirect information channel gives us.

_________________
cheers
tony

Hyperion Support Forum: http://forum.hyperion-entertainment.biz/index.php

@tonyw

Duh, XENA can of course interface directly to a bus mouse and the graphics driver could have a VBLANK interrupt that requested the counter from XENA, thereby giving smooth operation from an original Commodore Amiga mouse (or any other type of mouse). This would take up a very small part of the processing time. Of course, it would eat 6-7 pins (XA, XB, YA, YB, BUTTON1, BUTTON2 and optionally BUTTON3).

Quite simple, really.
You don't need to emulate AGA or even OCS to get some of the functionality, just recreate the ideas

Of course, if you have a gamer mouse this should be LESS of a problem, but even they have the inherent jerkiness due to not being synced to display refresh. Wonder if the USB mice could be polled from the VBLANK interrupt? Or a timer interrupt preceding the VBLANK interrupt, so that new data would always be available on pointer redraw.

Where are the USB gurus when you need them?

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

In Windows they actually CALL it the "poll rate", so they might be actually doing timer interrupts not locked to VBLANK. If there is a way to lock the poll rate to VBLANK then this will give smoother operation than ANY other poll rate. Simply due to the fact that you don't need to resample to a different rate.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

http://www.computer-engineering.org/ps2mouse/

Quote:

Remote Mode In remote mode the mouse reads its inputs and updates its counters/flags at the current sample rate, but it does not automatically issue data packets when movement has occured. Instead, the host polls the mouse using the "Read Data" (0xEB) command. Upon receiving this command the mouse will issue a single movement data packet and reset its movement counters. The mouse enters remote mode upon receiving the "Set Remote Mode" (0xF0) command. Remote mode is rarely used.

I would like to test a driver that uses this remote mode synced to VBLANK. Is that possible?

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@Ancalimon

I can add that using Windows 2000 and a PS/2 mouse, it's actually possible to change
the polling rate in the Mouse settings, from ca 30 to 100. It made the cursor move a lot more smoothly. However, now with Vista and Windows7 and USB devices, I don't think it's possible
anymore.

regards,
Kotler

@Ancalimon

Quote:

Ancalimon wrote: @Arko I guess the Xena can be used to stop these kind of things from happening?

If Xena has the possibillity to disrupt HD access and to send data to the GFX card it might help
(that was ironic, Xena could adapt the olf mouse but could not do the display updates)

Last edited by Arko on 23-Apr-2010 at 08:13 AM.
Last edited by Arko on 23-Apr-2010 at 07:03 AM.

_________________
AmigaONE. Haha. Just because you can put label on it does not make it Amiga.

I borrowed this comments from here (#27 & #28):
http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=38873&forum=2&start=20&order=0

This mouse smooth movement thing I would like to see improved, perhaps with the help of xena.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

A use of Xena is to get the mouse pointer moving properly? Really?

Chris

Oh no!!! you will draw me crazy with a similar discussion! I've never worried about mouse movement on any computer I had until now... what will I do tomorrw, if I'll find jerky pointer movements on the screen?!?

@olegil

Sure you could poll the USB in sync with the VBLANK interrupts, but you'd need to reserve one bus specifically for the mouse - I don't think you could do it successfully with the generic USB with other devices hanging off it.

But if you want to make a direct connection to an "Amiga" or PS/2 mouse without the USB conversion, the problem is: where are you going to get a mouse that isn't USB?

_________________
cheers
tony

Hyperion Support Forum: http://forum.hyperion-entertainment.biz/index.php

@tonyw

But XENA can do USB in software... This is a non-issue. You need a minimal XORRO board with USB compliant IO, then hook the mouse in there.

Amiga mice can be purchased from Amiga vendors and enthusiasts. Don't really see the problem?

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

Quote:

by tonyw on 23-Apr-2010 12:12:48 @olegil Sure you could poll the USB in sync with the VBLANK interrupts, but you'd need to reserve one bus specifically for the mouse - I don't think you could do it successfully with the generic USB with other devices hanging off it. But if you want to make a direct connection to an "Amiga" or PS/2 mouse without the USB conversion, the problem is: where are you going to get a mouse that isn't USB?

You might consider using legacy connectors for compatibility with older devices. However, finding such mice could be challenging given their obsolescence and scarcity in modern markets. Alternatively, creating an adapter or emulator software may offer some solutions while respecting the original hardware limitations of Amiga systems.