HD access noise for SIDE2?

[tuf]

I know this is probably a dumb question, but as an MIO user I'm definitely missing the HD access noise when using the SIDE2 and wondering if there's a utility or something that will provide something similar. BTW I'm not talking about the noise an actual HD makes - I'm talking about the SIO-ish noise the MIO would make when you access the HD.

 

Thanks!

 

[flashjazzcat]

I didn't realize the MIO simulated SIO noise when accessing SCSI drives. In any case, the SIDE driver would need to be modified to hit Pokey to approximate SIO noise during PIO transfers, and this would cripple the read/write speeds (HDD sector transfers are accomplished using tight, often unrolled loops).

[+Larry]

Nope, you get silence with many of these newer devices. The exception is the MyIDE-II cart which has a pleasant "buzz" during HD transfers. My Black Box has it and the new rom for the MIO also has transfer sound. I always liked the audio feedback from SIO and hard drives.

 

-Larry

[flashjazzcat]

I wouldn't want to halve (or worse) the PIO transfer speed simply in order make it sound like a serial drive. The SIO noise is a natural side-effect of Pokey handling the serial transfers. All the SIO driver has to do is set the volume and the rest happens automatically. With parallel transfers, making noise is rather a waste of CPU cycles.

[+CharlieChaplin]

Hmmm,

 

add a PC beeper (approx. 5mm), which gets power when you access the HDD (CF/SD or whatever) by reading or writing. It will not give you SIO sounds but some nice ear-killing beeps... and in a few months you may report back, that you are deaf like me. Do not forget to add a headphone jack and use headphones / earphones as often as possible, so you can get deaf much faster... ;-)

[+mytek]

I wouldn't want to halve (or worse) the PIO transfer speed simply in order make it sound like a serial drive. The SIO noise is a natural side-effect of Pokey handling the serial transfers. All the SIO driver has to do is set the volume and the rest happens automatically. With parallel transfers, making noise is rather a waste of CPU cycles.

I agree with keeping it as it is. Why waste computing power and increase access time just to make silly sounds. What would be the purpose?

 

- Michael

[+Larry]

I wouldn't want to halve (or worse) the PIO transfer speed simply in order make it sound like a serial drive. The SIO noise is a natural side-effect of Pokey handling the serial transfers. All the SIO driver has to do is set the volume and the rest happens automatically. With parallel transfers, making noise is rather a waste of CPU cycles.

 

Not even close to that kind of transfer reduction. @ Michael -- the point is that some folks like it.

 

-Larry

[flashjazzcat]

I had intended to run RWTEST on an emulated MYIDE-II in order to sample the effect and evaluate transfer speeds for myself, but gave up after an hour of trying to set up a blank VHD with a FAT and MYBIOS partitions. Out of curiosity, what is the RWTEST result on a QD partition under SDX?

 

EDIT: I found a video of MYBIOS disk transfers and I guess a similar noise can be generated by writing to $D01F once per sector transfer, so if authenticity is not a concern, I guess it can be accomplished cheaply or at zero cost, thus:

 

Edited September 10, 2016 by flashjazzcat

[Roydea6]

 

U1M equipped 800XL profile 2

SpartaDos X Disabled

PBI Bios Disabled.

 

Top run is with out Screen Blanking

Bottom run is with Screen Blanking.

 

MyIDE 2 Cartridge with SDX449b MyIDE 512K from FTe & DLT SpartaDos Website.

 

 

[flashjazzcat]

Thanks Roy. This would indeed confirm a zero-overhead implementation.

[+Larry]

Hi Jon-

 

What do you mean by "authenticity" -- the type of sound? If so, yes, it is a "buzz," not a "thunk, thunk, thunk."

 

The Black Box has a "purer" R/W sound, but I have no idea how Puff did it.

 

-Larry

[flashjazzcat]

Exactly - the type of sound. I remember some versions of Candle's SIDE loader provided a rather detailed simulation of sector churning sounds, but - apparently - with quite a high impact on performance. The quick test I did in the video, however, seems to correlate to the MYBIOS sound and is achieved by simply stuffing the last byte read from a sector into the CONSOL register.

 

I'm still reluctant to put this into tommorow's major U1MB/Incognito/SIDE firmware update (ROMs are sitting on my HDD here, but I gave myself a day's cooling off after the final bug fix), not least because it would require extra menu items in both the main BIOS and the loader for which there is simply not room. But if the lack of simulated SIO sounds is the only complaint, I will be happy enough, since the implication is that it's becoming harder to find things to complain about.

[+Larry]

Here's an "apples and oranges" comparison using the Black Box. I ran my HD backup program which reads the HD and writes to an APE image for backup. With the emulated SIO sound enabled on the BB, the time required for a 12000-sector backup was 15:33. With the emulated sound off, the time was 15:32. I would suspect that Puff also used CONSOL. (I remember reading about doing that in an early issue of Analog.) But since the BB sound is much more SIO-like than the MyIDE, that takes me back to "don't know how he did it."

 

-Larry

[tuf]

That'd be fabulous! Maybe change the Covox option to "Sound" and have Off/Covox/HD/Both or something like that to save a few bytes

[tuf]

Here's an "apples and oranges" comparison using the Black Box. I ran my HD backup program which reads the HD and writes to an APE image for backup. With the emulated SIO sound enabled on the BB, the time required for a 12000-sector backup was 15:33. With the emulated sound off, the time was 15:32. I would suspect that Puff also used CONSOL. (I remember reading about doing that in an early issue of Analog.) But since the BB sound is much more SIO-like than the MyIDE, that takes me back to "don't know how he did it."

 

-Larry

 

The MIO sound is very much like SIO as well, just a much higher pitch. Its not complex, it's just a pure pitch it sounds like.

[Kyle22]

It's still 3 bytes and (IIRC) 4 cycles per sector. Also, that buzz sounds a bit silly, and would get annoying quickly. I vote for leaving the I/O code as is.

[Mathy]

Hello guys

 

The nice thing about the SIO-sound is that you can hear if things are going wrong.

 

It's a pity the cartridge port doesn't have a Sound-In pin. Otherwise one might be able to somehow "copy" the data stream onto that pin.

 

Sincerely

 

Mathy

Edited September 10, 2016 by Mathy

[+Larry]

It's still 3 bytes and (IIRC) 4 cycles per sector. Also, that buzz sounds a bit silly, and would get annoying quickly. I vote for leaving the I/O code as is.

 

That's absolutely fine with me, but all the devices mentioned (except MyIDE-II) have Sound ON/OFF, so the user can choose which he wants. And that might be a nice addition for Mr-Atari's new Beta 5.x

 

But was not my point anyway -- just that it is not a performance killer. And If it is a PITA for Jon to change, certainly makes sense to leave it out. I always kinda missed it on the IDE+2, but I got over it.

 

@Mathy -- good point. When something is going wrong, you can hear it very quickly.

 

-Larry

[lemiel]

On/Off is a must if You, for example, play samples directly from hdd.

[flashjazzcat]

The Black Box has a "purer" R/W sound, but I have no idea how Puff did it.

The BB firmware I have here includes a countdown loop (using the Y register) which repeatedly stores Y in CONSOL and WSYNC. This probably generates a more interesting tone but also burns scanlines, so is not an acceptable technique here, not only because of wasted CPU cycles (which could be an optional sacrifice), but because I avoid WSYNC stores in the mainline code because they can mess up the DLIs in the loader menu.

 

Other than that, I tend to agree with Kyle: the perfunctory farts generated by the quick method aren't that great. While serial NAKs or timeouts generate characteristic, easily identifiable burps which immediately let our ears know something has gone wrong, a NAK or timeout on a PIO device would just result in normal cessation of the farts unless we went to the lengths of simulating some special error noise. Of course unexpected silence in itself might provide a cue, but IO noise is sufficiently fiddly and prone to unwelcome side-effects (such as the aforementioned sample playback) that I think I'll avoid it for the moment. One universal setting might have been quickly doable, but since the loader can exist stand-alone (on the SIDE cart) and has its own IO code separate from the PBI BIOS, it's not even enough for the loader to simply inherit the noise setting used by the PBI BIOS (since the U1MB NVRAM is not present in all setups). And the anomaly of silent XEX loads and noisy ATR and partition handling does not seem an appealing proposition.

Edited September 11, 2016 by flashjazzcat

[+warerat]

The way I added I/O noise to the MIO firmware is pretty basic and doesn't really incur much of a penalty. If you turn it on you'll notice the beeps happen only for read/write operations and sound the same.

 

All I do is stuff values into AUDF1/AUDC1 before the main I/O loop happens and turn it off when it's complete. So essentially the duration of the beep directly correlates to the time it takes to select the drive, send the command frame, and pull all the data off during the data phase.

 

Does it sound authentic? No, but it provides some auditory feedback and doesn't waste any cycles during the critical I/O region.

[tuf]

The way I added I/O noise to the MIO firmware is pretty basic and doesn't really incur much of a penalty. If you turn it on you'll notice the beeps happen only for read/write operations and sound the same.

 

All I do is stuff values into AUDF1/AUDC1 before the main I/O loop happens and turn it off when it's complete. So essentially the duration of the beep directly correlates to the time it takes to select the drive, send the command frame, and pull all the data off during the data phase.

 

Does it sound authentic? No, but it provides some auditory feedback and doesn't waste any cycles during the critical I/O region.

 

And it sounds great!

[Rybags]

Ideal would probably be to start a pure tone + noise to simulete IO sound at the start of a record, turn it off at the end, which would remove the speed penalty.

 

But IMO one of the big advantages of the IDE type devices is that Pokey is left alone which means 4 channel music even with digital effects can be done in the background.

By having the firmware take this advantage away would be detrimental.

The best method would be to have it as an option but where to put the flag - best would be not in Atari memory, but on the device itself if possible.

Edited September 13, 2016 by Rybags

[Marius]

A few years ago I asked Mr. Atari for HD I/O noise like the blackbox in MyBIOS. He added something that was even cooler than the blackbox sound. FJC posted a movie of that sound in this thread, which is rather accurate. The fun of this sound is, that is represents some kind of speed. While the BlackBox I/O sound is a real tone, it also sounds slow (even on a very fast storage device).

 

I like I/O sound, so if it can be done (also on IDE+ btw. in case Konrad reads this ) I would like it.

[tuf]

On 9/10/2016 at 5:28 AM, tuf said:

I know this is probably a dumb question, but as an MIO user I'm definitely missing the HD access noise when using the SIDE2 and wondering if there's a utility or something that will provide something similar. BTW I'm not talking about the noise an actual HD makes - I'm talking about the SIO-ish noise the MIO would make when you access the HD.

 

Thanks!

 

I am very happy to report that after installing v3 of FJC's amaze-balls firmware, I finally have the hard disk sound I've been looking for.  Thank you so much Mr. Flashjazzcat!