Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Oliroby]

I'd like to add serial ports to a PowerBook G4 Ti OS 9.2, and have seen the Keyspan USA-28X USB to serial adaptor.

Would this work with the G4 PowerBook Ti, and which serial MIDI interface's would work with this setup?

Only I've seen it mentioned that any serial MIDI interface used with the Keyspan needs to be a powered one?

So a simple MIDI Macman would not work?

[GaryN]

Don't even f**k around with unpowered MIDI interfaces. They're primitive and will very likely give you nothing but aggravation.
Note there were both powered and unpowered versions of the Macman for that reason.
Why? Neither the Mac serial Mini-DIN 8-pin or a 5-pin MIDI connector can carry power - so the 5vdc on the USB port is NOT passed thru.
Old MIDI interfaces that worked without external power did so by "robbing" supply voltage from the data lines which was never a good idea.
They got by on tiny amounts of supply current by having a bare minimum of components. They were prone to data overruns and dropped bits as a result.

[Oliroby]

Kinda got two open forum topics to solve one issue....

That is that the Akai S950 refuses to work with anything but a serial MIDI interface, and that serial MIDI interface can only be one of the "simple" MIDI units like the Macman. A Unitor 8 or AMT8 serial versions will not work. I have this info from the Akai samplers FB group, with a video demonstrating this.

From my other thread I can see that a Wallstreet, although having a serial port, the USB PCMCIA card may not work with a USB MIDI interface. I defiantly need a USB port, as all my other samplers are MIDI'd into an Edirol UM880. This works great with Recycle on my G4 PowerBook Ti.

Which means that the Wallstreet would not be an option for both USB and Serial MIDI.....

So using my G4 PowerBook Ti I'd have to use a Keyspan USA28X with a simple MIDI interface. As you've explained the  early Macman has no power, so that is ruled out. Leaving the Macman with the one MIDI in three MIDI outs, needing a powered 9V supply as a "possible" option that would allow me to use the S950 with Alchemy 3.0.

I'll investigate this further....

[robespierre]

MIDI is MIDI. An instrument does not know what type of MIDI interface is connected to it.
There are certain features in the MIDI protocol that cause problems for some instruments, like Active Sense or Timecode. The way to handle this is to disable those features.

[GaryN]

That is that the Akai S950 refuses to work with anything but a serial MIDI interface, and that serial MIDI interface can only be one of the "simple" MIDI units like the Macman. A Unitor 8 or AMT8 serial versions will not work. I have this info from the Akai samplers FB group, with a video demonstrating this.

One thought: I've found that in general, 30 to 40% of info you get from FB users is accurate… on a good day.
There is NO reason I know of that would require a serial interface to address a S950.  It has simple MIDI jacks looking for the same. There may be other reasons somebody else "had to" use a serial and they think it's required because it happened to work for them.

That said, I don't own an Akai sampler anymore (too aggravating) so I might be blowing smoke here. I'd like to see the video you mentioned.

[Oliroby]

It seems the Akai S950 will only work with a serial MIDI interface in the context of a successful MIDI sample dump (not note on/off) The S950 uses its own sample dump protocol, different to standard MIDI sample dump.

The following is info as I have it from various sources....

Alchemy is the best bet for sample dump on a Mac with the S950. Two way sample dump send and receive works fine if using the Apple MIDI manager with Alchemy, but this can only be with a Mac system 6 or 7. The reason for this is the S950 requires the MIDI sample dump sent to it at a certain speed which the Apple MIDI manger allows.

But if using OMS with Alchemy (Mac system 8 + 9) then yes the S950 will work, BUT only for one way transfer from the S950 to the Mac. As noted the FB post said this was only possible with a serial interface. I can partly confirm this as using my USB Edirol UM880, Alchemy lists the samples in the S950, but when I try to receive, the "blocks remaining to be transferred" dialog will appear, then show a block size number and then it takes around 10 mins for that number to decrease to the next block size number. So a sample of 73 samples would take many hours. I could not transfer a sample from Mac to the S950. Could only partially get it to work with USB MIDI, but the transfer would take hours. For some reason the S950 is not playing well with USB MIDI.

The guy on FB is using Alchemy, OMS, Mac 9.2, with an Altech MIDI interface, and shows that this works for S950 to Mac. So confirms serial will work, but as I found USB MIDI does not work.

The attached screen shots are showing Wallstreet, OS 9.2, Alchemy, Altech serial interface receiving a sample from the S950. (note the Altech interface is a non powered one)  He tried the same with a serial AMT8 and Unitor and transfer did not work. Same guy confirmed that two way transfer works ok but only with Apple MIDI manger on OS 6 - 7.

I'm only needing one way send from the S950 to the Mac, as I'm finding a workaround for the lack of waveform display on the S950. I like nice crisp cuts on breaks, which I can do with Recycle and Akai S3000. So if I can dump the sample from the S950 to Alchemy, then open that sample in Recycle, I'd be able to see the exact sample number on the Recycle split point to type into the S950 for a nice clean cut. As I could not get the S950 to play with the USB Edirol, and it seems that a serial interface will work, my bet is a Keyspan serial and a Midi Macman powered version should work with my G4 PowerBook Ti... fingers crossed.

[Oliroby]

Have been doing some research on the Keyspan USA28X and came across this in the manual......

"7.1.w - MIDI Information
The Keyspan USB Twin Serial Adapter supports the following MIDI clocking modes: x1, x16, x32, and x64. This means that if you have a MIDI device that uses any of these clocking modes, your MIDI device should work with the Keyspan USB Twin Serial Adapter.
However, due to the different implementations of the MIDI standard, some devices may not compatible with the USB Twin Serial Adapter. Please contact your MIDI device and/or software manufacturer for compatibility of their products with the USB Twin Serial Adapter or contact Keyspan for more information.
Currently, Keyspan only supports Mark of the Unicorn's FreeMIDI software."

I'm hoping that's an old manual and they did add OMS support?

Can anyone confirm?

Also interesting there's no mention that the serial MIDI interface needs power as the USB port will not send power to it.

[robespierre]

From the MOTU MIDI Timepiece AV manual:

Direct Connect is a special mode that allows non-standard MIDI data transmission between a MIDI device and the Macintosh. In this mode, all cable merging is disabled. You may need to use Direct Connect mode with hardware that does not conform to standard practices. For example, some samplers require Direct Connect mode in order to perform sample dump transfers to and from the Mac. If you have difficulty with sysex transfers, try Direct Connect mode.
...
Direct Connect mode can only be turned on and off with the front panel LCD.

There is also a setting (in software) for delay between SysEx packets. It's common for instruments to need time to process a packet before receiving the next.

[Oliroby]

From the MOTU MIDI Timepiece AV manual:

Direct Connect is a special mode that allows non-standard MIDI data transmission between a MIDI device and the Macintosh. In this mode, all cable merging is disabled. You may need to use Direct Connect mode with hardware that does not conform to standard practices. For example, some samplers require Direct Connect mode in order to perform sample dump transfers to and from the Mac. If you have difficulty with sysex transfers, try Direct Connect mode.
...
Direct Connect mode can only be turned on and off with the front panel LCD.

There is also a setting (in software) for delay between SysEx packets. It's common for instruments to need time to process a packet before receiving the next.

Interesting.....Could be the issue with the S950 and what has been tried so far.

But I do know that there was Atari ST software called "Zero X" which was very similar to "Recycle", and that came with an Akai S900/ S950 profile. There'd be no way to alter any such "direct connect" type settings with the software or the Atari ST ports. Unless the author knew about this, and such needs were already addressed within the S950 profile. I'll see if I can send the author of that software an email.

[GaryN]

Currently, Keyspan only supports Mark of the Unicorn's FreeMIDI software.[/I]"

I'm hoping that's an old manual and they did add OMS support?

I've used only OMS with Keyspan units for years. I suspect that sentence was written by a hired writer who had no idea what he was actually saying. OMS pre-dates FreeMIDI. MOTU developed FreeMIDI because in the beginning, OMS was proprietary to Opcode Systems only. Very shortly thereafter, Opcode let OMS go open-source as it was already becoming a standard and the industry needed one. FreeMIDI continued to exist although it does exactly the same thing as OMS because they could include little things to specifically address MOTU interfaces just like OMS addresses Opcode ones.

A very important point is made in this sentence:
"Direct Connect is a special mode that allows non-standard MIDI data transmission between a MIDI device and the Macintosh."
It describes the crux of everything you're dealing with. It's also related to the reason that, as I said before, I no longer use Akai samplers because they're too much aggravation. All of this is because they needed to have a way to force sample dump data thru MIDI lines that were NOT designed for that purpose. One of the big issues (maybe the biggest) doing so is reliably clocking all of that data thru at a rate that gets it there reliably and doesn't take hours to accomplish.

Anyway, this is the bed you're lying in and I think from here on out it's gonna be 99% trial-and-error.

A couple of notes:
* The Altech interface is internally identical to the Midiman. They were all made in Chatsworth, CA and sold with a dozen different names.
I can guess the reason they work is because they don't attempt to do anything at all to the data stream and a sample dump, with its density, provides enough "extra" current to run the chip inside. So, you get a sort of automatic "direct connect" mode. Repeat: that's only a guess.

* I have yet to see the FB video because it's a private group one must join first…

[Oliroby]

Currently, Keyspan only supports Mark of the Unicorn's FreeMIDI software.[/I]"

I'm hoping that's an old manual and they did add OMS support?

I've used only OMS with Keyspan units for years. I suspect that sentence was written by a hired writer who had no idea what he was actually saying. OMS pre-dates FreeMIDI. MOTU developed FreeMIDI because in the beginning, OMS was proprietary to Opcode Systems only. Very shortly thereafter, Opcode let OMS go open-source as it was already becoming a standard and the industry needed one. FreeMIDI continued to exist although it does exactly the same thing as OMS because they could include little things to specifically address MOTU interfaces just like OMS addresses Opcode ones.

A very important point is made in this sentence:
"Direct Connect is a special mode that allows non-standard MIDI data transmission between a MIDI device and the Macintosh."
It describes the crux of everything you're dealing with. It's also related to the reason that, as I said before, I no longer use Akai samplers because they're too much aggravation. All of this is because they needed to have a way to force sample dump data thru MIDI lines that were NOT designed for that purpose. One of the big issues (maybe the biggest) doing so is reliably clocking all of that data thru at a rate that gets it there reliably and doesn't take hours to accomplish.

Anyway, this is the bed you're lying in and I think from here on out it's gonna be 99% trial-and-error.

A couple of notes:
* The Altech interface is internally identical to the Midiman. They were all made in Chatsworth, CA and sold with a dozen different names.
I can guess the reason they work is because they don't attempt to do anything at all to the data stream and a sample dump, with its density, provides enough "extra" current to run the chip inside. So, you get a sort of automatic "direct connect" mode. Repeat: that's only a guess.

* I have yet to see the FB video because it's a private group one must join first…

Really good info ! Appreciated!! It seems my best bet would be to go with the G4 Ti, and now that OMS has been confirmed working with the Keyspan I'll look to pick one up as well as a powered Macman.

I've had a look at various pictures of the Altech version, but there is no 9V input on it. Strange it was left off, but the FB video shows it working with a G3 Wallstreet serial port, so that would be how it's getting its power. Getting a powered Macman will hopefully work with the Keyspan.

I'll report back once I've put this all together.

[DieHard]

I have been kinda sitting back and recovering from my gal bladder removal this week and letting Gary, who is definitely the MIDI master, give you his knowledge.

I remember, back in the day, I had a ton of clients, and myself that worshiped AKAI samplers.  Even after moving to soft samplers, it was clear that the many AKAI units I owned had a different sound and recorded with thicker low mids than the same samples when placed in HALion/Kontakt or the like... I do not want to trigger a huge debate on the issue, but the hardware filters and other settings on the AKAIs are in their own class; due to convenience and rapid workflow, the majority of us moved on to load all our samples in the box, or the mac, so to say

OK, I'll finally get to the point, at the time that myself and others relied on these hardware sampler boxes, SCSI became a must !  Is there a reason you don't want to go via SCSI ?
AFAIK you can get this:
https://www.ebay.com/itm/265414612114?chn=ps&norover=1&mkevt=1&mkrid=711-117182-37290-0&mkcid=2&itemid=265414612114&targetid=1263433204214&device=c&mktype=&googleloc=9031498&poi=&campaignid=14859008593&mkgroupid=130497710760&rlsatarget=pla-1263433204214&abcId=9300678&merchantid=115306020&gclid=Cj0KCQiAzMGNBhCyARIsANpUkzPGeEr9ETa5YrfyLrsoCPRdwgmEV_cvil8yq9Is0z1OqMhvxLmRUVsaAtx6EALw_wcB

SCSI opens the world of Sample transfers to be so much easier (maybe I am missing something here) then trying to dump it via a type propriety sysex data dump ?

I am not feeling great and I may be missing something here, but...

You can fly into and out of Recycle, you can save to CF cards, external HDs... yada, yada, yada... please let me know if this sounds like a winner

UPDATE: I am guessing the $69 one I linked may not work with the S950, probably going to need a hard to find ib105 as mentioned by Chris
http://macos9lives.com/smforum/index.php?topic=2668.0
 

[Oliroby]

Appreciate the info and I have made use of SCSI but only by way of SCSI2SD's HD's fitted in my Akai S3000 floppy bay, and external SCSI2SD drives with a Roland S760, Roland S750, and Emu E64.

But as I use a PowerBook Ti (only for sample editing, Recycle) using SCSI is again very hit and miss on this machine as it lacks a SCSI port. I have a Belkin USB to SCSI card, and that only works with the Emu E64 for sample transfer with Recycle. It refuses to work with the SCSI port on the back of my Akai's and Roland's. Also I sample direct on the machines, sometimes hitting the AD convertors hard. Sounds amazing on the S950, the Roland S750 and S760 are really deep and warm. The filters are amazing too, really have something special about them, especially on the S750. Amazing bass on those, very full sounding. So I'd not use SCSI for transferring ready rolled samples from the computer to the samplers, bypassing the AD in's and half the point for keeping these old machines going.

All my samplers are MIDI'd into my Edirol UM880, and I use a MacBook Pro and Logic X or Renoise for sequencing. When I come to sample edit with Recycle, I just swap out the MacBook Pro with the old G4 Ti and use the same UM880 with it. Simple swap.

I'm cool with working with Recycle and its MIDI sample dump profiles for the Akai S3000, S750, S760 and Emu E64..... its just trying to shoe horn the S950 into the setup that's proving a little tricky. I do have an SD card floppy emulator fitted in that machine....... if only Chicken System would finally add the long promised S950 compatibility to Translator 6/7, then I could just take the SD card, put it in my Mac and display the saved sample from it in Recycle or a regular audio waveform editor and find the split points..........maybe one day

[GaryN]

I've had a look at various pictures of the Altech version, but there is no 9V input on it. Strange it was left off, but the FB video shows it working with a G3 Wallstreet serial port, so that would be how it's getting its power. Getting a powered Macman will hopefully work with the Keyspan.

I think we're still not quite in sync here.
Look at the pic of the Macman. You'll see the 9v input says "optional".
Look at the Altech. It's identical to the Macman.
Look at the Mac serial port pinout chart. Note there is NO pin that provides power. It's data-only.

This interface was first sold without a power input. They worked pretty well in the early days when MIDI was new and you were amazed you could connect your computer to your new DX7 or Oberheim or… wait, that was it until the Prophet 5. Problems developed when idiots like us decided that no MIDI port should be allowed to exist unconnected and we started hooking up Roland TR drum machines and Yamaha FB-01's also. Pretty soon, we had missing notes and better stilll: STUCK NOTES that happened when a synth didn't get proper a note-off. Opcode SVP and Cubase et al had "panic functions" that sent note-offs on ALL MIDI notes with one keypress.. It was decided then it might be a good idea to provide a little extra power to keep the serial< >MIDI chip happy.

These were/are basic-to-the-bones units built around one single chip that is extremely power-efficient. MIDI ports however, require opto-isolated connections to prevent ground loops. Those chips suck a little more power too. so, the 9vdc input. Those boxes will work powered OR unpowered but they work better powered.

My point - if I actually have one - is that your enormous pile of "obsolete" hardware and software will likely always have one shortcoming or little aggravating quirk or another and all you can do is trial-and-error your way to the most stable working mess of spaghetti you can.
Personally, I salute you for your tenacity and determination with all of that stuff.

Nikola Tesla would be proud. 

[robespierre]

There'd be no way to alter any such "direct connect" type settings with the software or the Atari ST ports.

Note the last line of the manual quote.
"Direct Connect" is a feature on some multiport MIDI interfaces, to make them compatible with instruments that don't obey the conventions for SysEx packets.
Software isn't aware of the difference, except insofar as all communication with the MIDI interface goes through a single port, "without merging". Merging is the key, because when multiple MIDI streams are merged into a single computer port (serial or USB), the interface needs to find the beginning and end of each MIDI message so it can multiplex them together. When there is only a single MIDI stream, there is no merging and the boundaries between messages don't need to be detected.
Multiport interfaces rely heavily on interpretation of MIDI messages; they contain message processors that the basic interfaces do not have. In order to do things like sync a click track to MIDI timecode, or do CC filters or studio patch changes, they need to find and extract specific MIDI messages from the data stream. If the instrument's sample dump has a non-standard format, the message processor will get confused because it can't find the beginning and end of each SysEx packet.
An Atari ST has only basic MIDI ports with no message processing hardware, so software has complete control over the data format. It doesn't need "Direct Connect" settings because there is no merging hardware to be disabled. (I'm not sure if the Notator box, whatever it's called, has hardware merge.)