About improving the WinUAE Floppy Drive Sound emulation...

Hi there,

Been a great fan of WinUAE and love the new floppy drive sound emulation.

But, although the drive sound emulation is very good already, it's missing one essential part of the drive sound: it doesn't emulate the change in pitch when the drive motor moves from the inside to the outside of the disk!

The best example of this being with the great Hybris.

When loading the game on a real Amiga and the drive motor moves over the disk, the loading sound goes like:

KKKRRrrrrrrrrreeeeewwwweeeepppp (pitch changes higher towards the end of each sweep over the disk)

while in WinUAE it just only does:

KKKRRrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr (NO pitch change in the sweep over the disk!)

Now I sincerely hope people recognize this omission in the emulation, because it ads a whole different dimension to the sound!

I always LOVED the Hybris track loading for its exorbitant drive masochism, so it would be great if the programmer of WinUAE could put WinUAE and a real Amiga next to eachother and compare the Hybris loading, and improve on the WinUAE emulation! I don't know how hard it would be to implement, but is sure would be great if it could be done!!

Well hopefully I made sense here and thanks for reading

[Chipmunk_UK_1]

Could somebody with a real miggy, and this game generate an mp3 of the disk sounds, and stick it in the zone?

I've made some minor changes to the disk click code, but I need something to check it against to see how far out I am :hoo

(I do have a real miggy, but no way to get the game onto a floppy as of yet... still need to get the right cable... damn I'm lazy! )

[Mojo]

As soon as I get my 2.5 - 3.5 adapter (within a week), i will do this, my miggy is uhh well not working atm

[RCK]

Nice geek idea, I love it
Thread moved to WinUAE suggestion

[teh]

Would any external floppy sounds be needed?

[hal]

i'll mp3 the original sounds used in the current version
the other pitch should be there though not used i think

[Toni Wilen]

Pitch probably can't change because distance of tracks stays the same in both ends of disks and stepping speed also stays the same (except if loader wants to be weird..)

Modifying pitch made sound very weird

IMHO the whole sound changes because of changing acoustics (resonances etc..) when the distance between floppy drive's rw-head and other parts of drive/case changes.

Emulating this may be very complex..

[Chipmunk_UK_1]

Yea, I did think that it wouldn't work very well, but wanted to try it anyway

The pitch doesnt need to change all that much, just enough to be noticeable...

But you're right, it doesn't sound right... the sound in reality is changing because the position of the head relative to the spiral gear (not sure what it's called, the design of this bit changes alot from drive to drive :}) is changing (coming closer to your ear, when you've got your head close to the drive to hear it ) and parts of the drive react to the force of the motor moving the heads differently depending on the positions of the various parts

I don't think we want to get into acoustical modelling to get the sound "just" right :eek imagine how much that would slow the emulation down

[hal]

besides it's a job for life

[Nexjen]

I actully didnt know this was a new Feature of WINUAE. I think its actully okay , I like it , its one step to the real thing.

Hi,

Thank you Toni and Chipmunk for looking into this issue Too bad the disk pitch change boils down to accoustic sound changes, which will indeed be difficult to emulate. I never realized it would be related to accoustic sound change, but now that you said so I believe that it's a plausible explanation.

Thanks again.

Cheers.

Originally Posted by Toni Wilen

Pitch probably can't change because distance of tracks stays the same in both ends of disks and stepping speed also stays the same (except if loader wants to be weird..)

Modifying pitch made sound very weird

IMHO the whole sound changes because of changing acoustics (resonances etc..) when the distance between floppy drive's rw-head and other parts of drive/case changes.

Emulating this may be very complex..

[robotriot]

I'm a fan myself of the floppy sound =) There's this part in Hardwired, where the starfield in the vectorbox zooms towards the screen, and the floppy drive sound seems to be synched to it, that's just a great moment (don't know if it's intentional or not, but it's cool anyway :). Are there any demos that utilize the floppy drive to make sound? I've heard of a program "playing" a song on the floppy drive, don't know if it's only a rumor though.

[Mangar]

I had a program That made the floppy drive play a sound. It sounded like a violin. I do know it was the cause of a lot of drives suddenly needing repair. Vibrating heads is no good.

[Chipmunk_UK_1]

Just to revisit this thread

Perhapse a closer approximation of the sound could be achieved by 84 click samples (one for each position of the head, probably formatting a disk in xcopy or somthing to get big sample with the clicks far enough to seperate out?), and then play each one when appropriate.

So instead of the same sample being played when the head moves, the appropriate one could be played (this could model the acoustics a little better)

Anyway, crazy thought of the day over for me

[derSammler]

Instead of using 84 click samples you could also just modify the sample depending on the head position.

[Chipmunk_UK_1]

Certainly that would be the better solution, but writing an acoustical model of a floppy drive is a little ott don't you think ?

I think the easier and faster solution, would be to have the 84+ samples and play the relevant one depending upon the head position.

I may give this a go myself when free time permits.

Some samples of the formatting process (All 84+ tracks please) would certainly help speed things along a bit .

[derSammler]

Depends on how the sound is changing... maybe a simple pitch bend is enough...?

[Chipmunk_UK_1]

I did play with a simple pitch bend, and the effect didn't work that well at all

(I stated this in post #8 of this thread in fact )

[mr_0rga5m]

omg ... /me spies NERDS!!

[Toni Wilen]

84+ samples and play the relevant one depending upon the head position.

I agree, this is probably the best and easiest solution. I'll also try sampling if I find working microphone and my sound card's mic-in isn't too noisy (never tested..)