Yes, sounds like standard format for PC - as you say, if it wasn't you'd get error message or nothing visible. No need for disk analyser?
So, no point in sending me an image, thanks. Might be interesting to compare the Style files you've got with the Style files that come with 'Band in a Box' (I have a VERY old version of that) but when I'd ever get around to exploring that is anyone's guess!!
This is a little complicated, pretty much Yes, BUT....
I suspect that the situation is similar to that for Yamaha disks for their 'Disk Orchestra; files, and their ESEQ files, that I have some experience with. I've not had any of the Roland ones to play with.
Disks have two aspects to the 'format', the low level system which defines the number and details of the tracks and the sectors and the size of the sectors on the disk, and the high level system which looks after the way that files are arranged on the disk. The low level system seems to be the same as that used by the PC (the computer) and will be 80 tracks of 8 sectors for the DD disks, and 18 sectors for the HD disks. However, the high level system is different, so the files are not arranged in the same way, so if you put a RA-95 disk into a PC and try to use MSDOS or Windows to access the disk you will see nothing, or maybe garbage. However, if you use a suitable piece of software that bypasses the DOS filesystem and accesses the disk directly, then this software WILL be able to access the Roland files and read, play, or save them.
I'm sure that Roland have such software available. I have such software for the Yamaha system, which can list what files are on a disk, and can convert a Yamaha disk to a MSDOS copy that CAN be used normally on a PC, and such a disk can be converted back to the Yamaha format.
In addition to the above, I understand that there are some differences between the file systems. Some of the file types used by Roland are peculiar to Roland systems and may need further special software to manipulate them, but I cannot say much more without seeing the files. As well as this, there seem to be some complications regarding the structure of supposedly midi (.mid) files, such that a file saved on a Roland system MAY play back OK on a PC sequencer, but a 'standard' midi file (/SMF or .MID) saved on a PC system may NOT work on a Roland device (such as the RA-95) because it uses options not supported by the Roland system (for example the number of tracks in the midi file).
I would be VERY interested to get hold of a Roland disk to have a look at the data on it. The Yamaha software I have may not work, but I have other general disk/sector editing software that will look at the disk sector-by-sector and tell me what's on the disk, and how it's organised. I used such software on Yamaha disks before I got hold of the official Yamaha sofware to do the job properly, and back then was able to extract music files, and convert the data from the Yamaha (not quite) mifi file to useable midi files.
More details if you want?
Oh, another complication. The sort of software that will do this sort of thing on a PC may NOT run on later versions of Windows from XP onwards, as such Windows systems do NOT allow the required low-level access to hardware (the floppy disk controller). You will need an older PC, preferably using MSDOS Boot. Then again, such older systems are more likely to have the floppy disk drive anyway!!
Yes, I saw the piece from Roland including the text you quote. But I saw other things as well, which suggest that it's not as simple as you suggest, also reference to the Roland disks having similarities with the Yamaha disks, and I do know specifically about the Yamaha disks.
But, I could be wrong about the Roland disks. That's why I'd be so interesting in seeing one to check.
I be glad to mail you a diskimage if you want, because i bought it together with some floppies. I can't say if it is midi or styles on disks though "well until i get it".
I have two XP computers of one also holding win98. But only one have a floppystation so lets hope it work, 20 years since i used any floppies but they will come handy for use with my Korg 364 also.
It is the first arranger i ever had i guess the synthengine is some sort of SoundCanvas clone, could not find any info on sample rom size and sample resolution. Could be 44.1 khz 16 bit?
Thanks for your note. Yes, I'd love to get an image (what sort - I'm familiar with the .IMD type, but I have converted from a TD0 type as well).
If this is a factor, XP is not DOS enough, but the Win98 should be, as that's WinDoze running on top of DOS, whereas XP is loading Win and then running a pseudo DOS from within XP. But this might not be a factor.
I'll look up about the RA-95, maybe I can find some further info? Sounds interesting anyway.
I found the midi implementation chart, but i guess you already got it otherwise i can send you, quite alot of midi functionality.
I am most familiar doing images with Win Image, i think it use IMG but i am not sure it support IMD?
IMD was that teledisk a dos program?
I can see if i get hold of it.
I was thinking to make my sequensing program to use the banks of RA-95, but so far i have not found any instrument file for it *.ins, i do not think it is compatible with any of the soundcanvas but i only have my Sound Canvas 7 to compare with and that is basicly just a single soundbank and a drumbank. While this seems to hold similar instruments into separate banks. If you have or see any RA-95 ins file attach it in a message.
It would be fun to make/implement some sort of intelligent AI style program myself but i am not sure i am up to it.
But of course one could start with a chord file and work from there.
I guess a pure style arranger application not necessarily need some AI just scales, "and that is alot of music theory", but i feel some sort of chord program could be a start and within reach.
My programming been on hold for some months, but while i am exploring the RA-95 i will try pick it up.
Yes, I've seen the full manual, and the midi supplements - there are a LOT of 'implementation charts and I've not worked out which ones apply to what. Each possible mode of the RA-95 seems to have a different MIC
Not familiar with a Win image system, but the IMD is used by a system by M Dunford. I think the TeleDisk is something different, but that produces the .td0 images. Utilities within the IMD package (which I have) will convert .td0 to .IMD, or plain binary, and from the binary image you can get to most other things.
I've not found any tech details of the sound ROMs in the RA-95, but it is Sound Canvas, and it's GM/GS. The GS means that some of the sounds have alternatives in separate banks, that can be switched in to replace normal sounds, but keeping to the GM overall system. There do not seem to be any totally separate banks.
Not seen any ref to an .INS file. As I understand it, the .INS file is just a text file containing the Program numbers, and a full text name for the sound, it's just there for info/display on the screen. You could find an .INS file from somewhere else and edit it, or manually create a new one from the details in the manual. I don't think the RA-95 supports such a file, but your own system could?
I'll look up about Win Image - it may not be a problem??
Aha - just done that. Google tells me that WinImage produces a 'raw' image file, which sounds the same as what I was calling a 'binary' image. So, the file contains the raw/binary data from all the sectors, with ZERO formatting information. Many of the the more complex image systems are designed to produce images of disks that are copy-protected in some way, i.e. there are 'tricks' with the sector numbering, dummy tracks, fake damaged sectors, etc, etc, and the image needs to contain extra data to allow things like that to be saved, and a 'correct' (or correctly incorrect) disk to be recreated. If the disk is straight normal/correct, then no need! So a WinImage file will be fine.
Instrument files is used to list names on computer enable it to set bank and program on device using computer, at least that how i use them.
I think already early GM programs like Steinberg 24 had the ability to set patch from computer.
The bank standard was later with GS?
Of course it could be read in manually but i guess the instrument files so small it does not really matter".
Honestly i can't see why DAWS and sequensers do not read in the available ins file to start with but maybe the format proprietary to Cakewalk?
It would been nice if the new MID standard had the ability to just dump a instrument file from any device, but i guess those patch and banknames not really part of standard.
I've found an .ins file that might be some use to you. It was with a virtual keyboard system, and it includes a stack of stuff from GM, GS patch names with extra bank stuff and on to XG stuff. The original was called gmgsxg.ins. Is this any help?
The idea with GS is that if you don't do anything with Bank Select, but use ONLY the normal program Change, as would be in any GM midi file, then the music would play normally with the basic GM sounds. If you are using a GS instrument (sound module) and you add in a Bank Select before the Program Change (PC), then you get the GS optional instrument instead. If you play that midi file using a non-GS instrument, then the bank select has no effect and you get the normal GM sound. If you use the RA-95 to play back, and you want one of the optional extra GS sounds, then you still have the relevant PC, but it is immed preceeded by the bank select command as per the table In the case of GS the Bank Select is implemented with just 1 of the Bank select codes, Checked back with the original PDF of the table, as I was having trouble reading the bit you attached, and the Bank Select needs to be CC o (MSB) with the data shown in the table, i.e 0 is the GM sound, the other numbers will swap to the alternate numbers. This does NOT need the LSB, (CC 32) which is always zero. Then followed by the PC command, which in effect also sets the Channel. Everything is 'actioned' only when the PC is actioned.
Regarding your note, yes, it may seem odd, but the Bank options are supposed to be subsidiary to the main sound, and in effect the overall bank is irrelevant. This is maybe another result of the need to maintain GM/GS compatibility?
is there a header or other way to easily inspect if a *.mid file is GM or GS or must you check every message to see if there is some CC for bankchange?
At some point there must been pure GM players "128 patches/sounds no banks", can those players cope with the extended GS standard?
Or is mid files always GM and XG files GS?
But i guess the drums "Channel 10" was in some way an early attempt on separation into banks.
Can i ask you is midi programming your profession or just a hobby?
What kind of projects/applications are you working on?
There is no marker in a midi file to say if it's GM, GS, XG or whatever (I have some files for MT-32 that sound VERY nice when played on my MT-32 device, but NOT so what played with GM setup). The more friendly folks who create midi files put one or more text messages in the midi file to say it should be played with GS or whatever, again, I have some very nuce XG files that warn so. Otherwise, it's not easy to determine. One of the little progs that I've written over the years checks thru a midi file and reports/displays some details, and this lists the Program numbers used and displays some text messages, and this would be a help, but I'd need to add some more 'features' to the prog to specifically check for GS or XG.
As I noted in a previous post, GS should sound OK if played as GM, maybe just not as nice as it sould be. But an XG file could come out totally wrong.
Drums on Ch 10 is really just a 'convention'. Drums do not HAVE to be on Ch 10, you can use any ch. Someone asked about this not so long ago, basically wanting to have more that one drum channel, with different kits. Look for the answers on that thread.
The midi programming is just a hobby, since the early 1990s, although I've been interested in electronic music/synths/etc just about forever. I was working professionally as a programmer for a number of years, employed, then self-employed, but this was business systems (accounting, stock control, invoicing, etc) initially using dBase, then Clipper, then C (with a dBase/Foxpro library). I'm now supposed to be retired, but actually I'm still 'supporting' a couple of customers who are still using my software, and in no rush to change.
I have a number of projects that were started, are now working to the point of being useful, but still need a lot more work to be 'finished' but probably never will be! That's how it goes!
Hope you get your RA-95 soon, we'll see what happens then??
Hello tonight i finally took some time to start scan thru the floppies, and did find original Roland style disks they appear to be normal floppy disks 80 track 1.44.
I have not diskanalyser on my Windows 10 machine but i guess if format was not standard it would not read.
Thanks for the note, but if the disks are just normal PC/DOS type disks then they're not really of much interest.
There's nothing I can do with the style data/files. Other - maybe - try to compare the data in them with the style disks from the 'Band-in-a-Box' system that I have a VERY old copy of, and who knows when I might have the time to do that. If any of the disks you got had any standard midi files on, they might be more interesting, but then just the files and not the disk image.