I was wondering over and over again what you all meant by this MIDI timing bug. Since a friend of mine has a JV2080 and I have a fully expanded XP60 and have never encountered the problems.

What I do know is that there is a software bug in the Super JV series, primarily the XP60/80. The problem is that ALL messages are sent using both the MIDI CHANNEL in question + CHANNEL 16.

To better explain this, if you press a key down hard enough to emit CHANNEL AFTERTOUCH messages on CHANNEL 2, the result will be CHANNEL AFTERTOUCH messages on both CHANNEL 2 and CHANNEL 16. Get it?

To see this in action, if you have an XP60/80, record something on any sequncer track. Make sure this something has lots of CHANNEL AFTERTOUCH, SUSTAIN (64) and BREATH (2) messages. Hell, use any message if you want!

When you're done, use the MICRO EDIT to see what's on track two. What you should see is everything you recorded + all extra control messages on whatever track used plus CHANNEL 16 information! Yes, this sucks!

If you have a JV2080 or JV1080, recording into a computer sequencer or equivalent should yeild te same results.

Big deal you say? Well, if you have 16 tracks going, you'll find that all 16 channels have information from CHANNEL 16 on them (generated by the current channel). If you used TRACK 16 to record, you'll get double information on CHANNEL 16.

Anyway, this all results in sloppiness and gives the impression that their is terrible MIDI timing which is NOT the case. The way to get rid of this is to just use the ERASE function to strip the current TRACK of all CHANNEL 16 information. This wont affect the stuff recorded on TRACK 16, since you are only getting rid of any CHANNEL 12 information that might have been recorded on another track.

Knowing all of this, I've NEVER had any MIDI timing problems using my XP60, that goes ditto for my freinds JV2080. I just thought I should post this info as I knew it for a while and believed this could be a very helpful solution to your problems.

The Infamous EPU.

No one cares about this? I thought that this was one of the most controversial topics?!

The Infamous EPU.

I care, or should I say I used to!!

Unfortunately I havent got my 2080 anymore EPU because I was sick and tired of its crappy misbehavings, in its place is the 5080, and so far It hasnt done anything wrong.

If i remember rightly, someone said ages ago about filtering out unused controller info, which did help, but, as ive said before the 2080 did differ from the rest because of its 3 EFX. There is no getting away from it. It is a hardware procesing problem, and not a software one, otherwise Roland could have come clean and said to everyone that they would update the eprom chips. Ive never used a XP60/80 etc so im not in a position to say how different they were.

Anyhow, Ive got a Nova, and that seems to be able to take huge amounts of controller info and doesnt budge on the timing.

The fact is that Roland, as a company produce some unbelievable products, but have been unable to get a processor fast enough to handle huge amounts of notes. (At the moment my 5080 has been up to about 80 notes and is fine, It remains to be seen whether it has the same problems)

HAPPY NEW YEAR TO EVERYONE!!!

No, the it has nothing to do with the Hardware from the standpoint that you are talking about. The EFX theory has no merit since it has nothing to do with polyphony.

For example, the SC880 has 5 FX processors and what happens is a small delay when switching to and from patches with differnt effect settings. An FX delay would occur, but that would hav enothing to do with processing a large amount of notes, it doesn't make any sense what-so-ever.

I'm almost certain that my theory is correct. The whole "Roland wouldn't release this info" even holds water in my scenario.

It's the simple fact that all of the Super JVs have this Ch X + Ch 16 problem. Everytime you send out control messages on Ch X, Ch 16 will output the same control message.

So get this straight - Filtering out Control messages will not correct this problem. Getting rid of all of the Ch 16 messages on Track X will.

The Infamous EPU.

epu

the efx processing has everything to do with polyphony. you dont think a synth like that has separate processors for effects do you, because i know that aint true. it is 1 central 32bit 66mhz processor doing the efx as well. it is a computer. its not like it has 64 oscillators in there, its software. and once a certain no. of mips has been exceeded, performance goes down the pan. i'm not disagreeing that what you say is true epu, you've got the gear, i havent anymore,so i cant test what you say, i believe its a combination of all these points made.

Anyone else got a take on this?

This whole thread has always puzzled me, since in my XP60 I have experienced neither timing errors nor controller data doubling on channel 16. It's not a global problem inherent in every JV/XP unit.

It is a global Super JV/XP problem. The reason I say this is becuase my older XP50 had this same problem. I sold that and got my XP60 in March of 2000.

Remember, the only way to test this is to record something on any track (for this example, use track 1) and while recording insert aftertouch messages by pressing hard on the keys.

When you finish recording, go to MICRO for the micro edit display and press TRACK 1 so it light up. You should now see channel aftertouch messages for both channel 1 and 16 on that track (look at the left hand side of the screen for the channel number).

I think it would be really odd if I had both the XP50, XP60 and JV2080 not to know this. Totty, also while the Super JV OS is written in software, there is a separate effects processor, the software controls this processor, that isn't in the software (at least for these models).

The 3 FX theory only seems to make senxe becuase the JV2080 is the only one that has it. The FX communication is somewhat laggish becuase there is only one address bus for all three processors, but those FX chips are separate from each other and the problems would result in audible effect problems.

The Infamous EPU.

if it's inherent to all superjv's, why don't you people address to the xp-30 as well? Is the problem solved due to new OS? speaking about software: can one upgrade it? and if so, how can one upgrade it?

I've never heard of the problem occuring in the XP30, however, I wouldn't be surprised if it did.

The XP50 could be upgraded as earlier serial numbers has FLASH ROM that could be updated by floppy. Later serials cannot. That's ditto for the XP60 and XP80 even though the Roland denmark site (www.roland.dk) has upgrades for these very instruments on their downloads page.

To see if it happenes on your XP30, open up a sequnecer program, record on track 1 (or any track) and go in to the sequencers micro edit window. If you see Channel 16 information on track 1, then you have the problem.

So if you have 16 track all with mistaken info from channel 16, you will hear lots of slop. Just use your micro erase to get rid of the tracks channel 16 information.

The Infamous EPU.

As I said epu, i'll take your word for it about the channel 16 info. However, before I got my 5080 I phoned Roland UK, and spoke to a guy there about whether the 5080 had the same problems as the 2080, regarding timing. His words to me were that the timing was better than the 2080 firstly and obviously because the processor is much faster, and secondly that the 5080 has seperate processors for it's MFX. As you say, this is only relevant for the 2080 as it is the only one of the JV/XP series to have 3 EFX.

I have heard it said before that the 2080's timing is the worst JV in the series (that came from someone doing a direct comparison of a 1080 + 2080, however what you say sounds like its true on the whole range, but made worse because of the 3 EFX on the 2080.

Wow, this whole lot is beginning to make sense!!

What I find most disturbing about the whole business is that us lot are discussing a big problem like this here, whilst Roland have not been honest with their customers.

I for one was very reluctant to fork out more wonga to the same company again for the 5080. However, I am here to say that the 5080, from my experience so far is excellent, kicking every other rompler out of it. (with the exeption of the JD990)

I hate to rain on your parade, but my XP60 does not exhibit this anomaly, i.e., it does not double any controller data on any channel or track (including aftertouch), nor have I ever experienced timing glitches or problems, and many of my sequences are nothing if not dense. Sorry.

Yes, I'm familiar with the sequencer, microscope mode, etc.; I've been using these sequencers since 1993 (also have a JV1000). I can't get my XP60 to duplicate the controller doubling.

I'm certainly not saying your units don't act the way you describe; I'm just saying mine have not exhibited any of these problems (I also had the XP50 before the XP60).

That's really weird DN. It happened on both of my units including the JV2080 and these units were made within years of each other and have no relation (other than the series).

Are you sure you're checking this right? I've had some other XP/JV owners try this around the way and they got the same results. Maybe this is a US problem only?

Also I thknk the EFX problem is null because even when I bypass the FX processors, I still get the timing problem, it completely disappears when I erase CH16 controller data however.

The Infamous EPU.

"Are you sure you're checking this right?"

Well, I know I've never had timing problems, so that's not a matter of checking. I do stagger some events by 1, 2, 3, etc. ticks/clocks if a passage is particularly dense, but I've never done it because I've been prompted by timing delays. I just do it; always have (even on the JV1000, which has some massively dense sequences).

As for the controller doubling, I just went in and recorded about six measures on track one, channel one, just random doodling, pressing down for aftertouch, using pitchbend and modulation, leaning into aftertouch again to get plenty of events. Checked in the microscope: nothing but channel 1 data, everything I had played; nothing from channel 16.

"Maybe this is a US problem only?"

I'm in the US.

So now what? That's why I said that this thread has always puzzled me.

Hey DN. I understand where you are coming from. What I now have o figure out is when this happens. I recorded a couple of tracks weeks ago. When I went in and looked at all of my tracks, they all had doubled controller information.

Today I recorded a couple of tracks and saw no doubled information. Now I'm confused. Maybe this thing happens on certain occasions? Maybe after a sequence of events? I'm really psyched!

I know this happened on my XP50 and has happened a lot on the XP60. I have to look out.

The Infamous EPU.

I have no clue; because it is intermittent, though, some parameter *has* to be engaging when it happens. Try going into System and turning your control channel (which I believe defaults to 16) "off", then see what happens. I'll futz around ("futz"?) with my control channel settings and see if it has any effect.

okay, I checked the doubling on an XP30 using Cakewalk 8. Absolutely NO info on channel 16. So perhaps the problem is corrected in the xp30. However, the first thing, as Dnarkosis said, this problem reminded me of was the controller channel parameter. I have noticed the parameter being switched back to it's original value (16) automatically. Very strange.... so maybe these problems are interrelated.
maybe it's a sequencer problem. Perhaps your sequencer is independantly doubling the info because both sequencer and JV/XP unit are trying to direct the info to a channel.

furthermore I experience very frequent timing problems especially when two or more program changes are sent on different channels and when there are lots of controller data events like pitch bends sent.

Rememeber when using Cakewalk that you have to check every track for channel 16 info, especially since it won't show up on channel or rtack 16 since all tracks can contain any channel.

I don't know what's up, but I'm pretty sure that it's got to be something like this. I've never had any problems with timing on my XP60 and that was my original point, so I never understood this arguement about MIDI timing, am I one of the few that never experienced it?

The Channel 16 thing was to provide a solution to some accidental MIDI slop that may occur giving the impression of bad timing, but it isn't. And last night I recorded a track 1 and there was channel 16 info on track 1 in the micro window, and this time the control channel parameter was set to OFF.

Maybe it's the Y3K bug - Lol!!

The Infamous EPU.

[This message has been edited by epu (edited 01-03-2001).]

The MIDI timing issue was definitely there (at least on the JV2080). I owned a JV-2080 for a year. Shortly after getting it home, I noticed it and it would drive me bonkers at times.

Over time, I got used to it and took steps to compensate for it so it would be noticeable in my tracks. At the time, I sold another module to get it so it's not like I could return it to the store (or I would have). I now have the XV-5080 and it is better timing wise. There are some OS issues still on it but it is definitely better in terms of MIDI timing.

Fernando

Thanks Fernando,

I'm glad someone else out there agrees! The 2080 was a pain in the ass. At the time I loved it and hated it at the same time. Im very pleased with the 5080 though-despite the sample load and save stuff having a few problems, but Im not using that part of it-yet.....

I can't understand how a JV-2080 can suffer from the problem described by EPU, namely doubling the MIDI information on channel 16 (in particular for aftertouch). The 2080 is a module: it doesn't produce aftertouch at its MIDI out (but is able to receive aftertouch info at its MIDI in). If EPU noticed some weired MIDI information at the "output" of a JV-2080, it might be because this information was already present in the input MIDI stream...

Hi Stephane,

I was not saying that the JV-2080 suffers from the same problem as the keyboards. I was really saying that it suffers from midi timing problems. These are real and I don't necessarily know what causes them. My assumption is that the processor for the 2080 is just overloaded with trying to support the 64 notes of polyphony, 3 insert fx, 2 master fx, plus the synth engine that it can't keep up and we get the midi slop in the meantime.

Simply playing it you won't really notice it. However, when you are using it in a sequence that requires some tight timing, you definitely notice it.

Fernando