I was just about to purchase an XP-80. I've read some recent posts that the XP has some defects. The post claimed the drum track occasionally loses timing and the polyphony doesn't seem to be a full 64.

I do large orchestral arrangements and really need the 64 polyphony.

Has anybody else had any problems with the XP?

Hi
I saw that you had seen my reply to sigma about the XP:s and that you were planning to buy a XP-80.
I´m not going to say that you shouldn´t buy one but I have to warn you about the problems I´ve had myself. (Specially when you say that you´re going to make large arrangements and need the 64 voices)
For some time ago I was recording some songs and I found out that the workstation had some problems with the seq./voice polyphony, and I wasn´t exceeding the limit of the 64-voice and seq.polyphony.
The problems showed up in different ways: the timing went up and down; an instrument disappeard or sounded strange for a short while; an instrument (BD, SD, HH) got completly out of time for itself.
Here is an example I have on a disk: I use only three tracks whit bass, drums and piano. I play four notes (keys) at the same time on the piano track (I use no sustain pedal and I don`t play any new [4]notes before the previous ones have fade out). Everything is very simple and I don´t play fast att all. The max. used polyphony comes up to about 25-voices. This in not exactly a large arrangement!! but when you listen to it the drums are already out of time. If you put on more tracks the drums gets worse and worse.
I was wondering why this only happened to to the drums so I did an experiment and let the drum track play another instrument (instead of the rythm set) but the problem was still there.
This means that the problem is maybe happening on the other tracks but it´s just more difficult to hear it on other instruments.
This is very complicated and it takes time to find out so I don´t think that so many people knows about it, and besides everyone isn´t using the sequencer.

I prefer workstations to computers but I don´t understand how they will compete in the future. A seq. program on a computer has an unlimited number of tracks and sequencer voices. The Roland workstations has only 16 tracks and it doesn´t work properly!!

Maybe the new Korg-Triton is something to wait for?!
I hope that you find something to make your music on whitout having any troubles.
Have a good time =)

PS: I´ll write some more information about those problems some other time

I have the same problems with my XP-80. I think the problem is with the sequencer's ability to play many simulataneous notes/sounds. I can, usually, have only about three simultaneous tracks as well. At first I speculated that the problem was polyphony. I thought that, since each patch is composed of multiple tones, perhaps each tone counted as one voice of polyphony. So, I counted the possible tones on the song I had recorded into the sequencer...they added up to only 34! So, I recorded parts into the sequencer using external MIDI gear and the problem persisted (I was using only 4-note polyphony on the XP-80 and no more than 3 note polyphony on each of the external instruments, so I had to conclude that the problem was with the sequencer becoming overloaded). From my experience, this is a problem with virtually any hardware sequencer (I don't know anything about such timing issues with software sequencing programs); my Korg M1 had the same problem, although I could seem to record more simultaneous notes on the M1's sequencer (but not many more!). I have never tried Yamaha hardware sequencers. I suppose what I am trying to say is that you may encounter this problem with any workstation keyboard. The onboard sequencer is great (very convenient; no other gear has to be turned on if you are using headphones) as a "scratchpad" for recording spontaneous playing and developing musical ideas, though.

I guess a way to successfully use the XP-80's sequencer would be to sequence only two tracks at a time and record them into a multitrack HDR (like VS880) and keep recording tracks upon tracks, using the HDR. (You can sync the XP-80 sequencer to the VS-880.)

I really like my XP-80, but am disappointed with the sequencer. Sound editing is pretty easy for not having knobs and the interface is, to me, extremely logical and easy to use. By the way, I very infrequently use the XP-80's factory sounds; I get virtually all of my sounds out of this synth from the expansion boards I have installed in it (I love their patches!). If it's the sounds you want, the JV-1080 and -2080, as well as the new -1010, have the same as the -80, but cost a lot less.

I also plan to buy a XP-60 primarily for the sequencer. After reading this I am concerned.

So today in a local musicstore I was lucky enough to test a XP-80 (and XP-60).
So I tried to crash the sequencer in the limited time and experience I have about xp.
I recorded 7-8 tracks. 3-4 tracks with 4 tones each. 4 tracks with 1 tone each (simultaneously). I did hear no slowdown, though a guy in the back of the store was playing really loud guitar with heavy distortion (hate when that happens...).

So either the slowdown was really small, or I erased by accident some tracks when recording (though I did hear at least 3 distinct sounds at playback). Or the sounds all had just 1 part.

I guess you have to play with it to see if you can live with it (in a more thorough way than I did perhaps).
I used to own a S-330 and W-30 which also showed very minor slowdown at limits but then again they only have 16 voices.

Thanks for all the advice. I love to hear more about this.

Lucien

Reading what Lucien did, I sequenced a few experimental multitrack four-measure loops on my XP-80. Here are some things I noticed, but don't know well I trust my ears: 1. It seemed that I could sequence many tracks without timing glitches if no slow/long attack time sounds were used. 2. When sequencing pads and strings sounds with long attack times, the timing of the drum track was thrown off; I thought that maybe it was just my ears deceiving me, but adjusting the attack time of the pads/strings to as rapid/short as possible did not improve the timing of the drum track. 3. Timing was generally better when entire loop was recorded in step time. 4. Sound dropouts were more common when the XP-80 was used to control other instruments (sounds from the tracks of other instruments dropped out). 5. When timing is off, it is off very slightly, but slightly enough to be bothersome. I don't know why these things happened and would love to hear from anyone that knows how to get around these apparent sequencer shortcomings (perhaps it is just the nature of hardware sequencers?). Perhaps there is something that I am not doing correctly? No matter what the sequencer can or cannot handle, playing and recording on the XP-80 is quite enjoyable!

The sounds I tried all were piano type sounds, so perhaps therefore the problem didn't show or not enough to hear.

I found the following on the net about XP-80 sequencer fixes. Does anyone have an XP-80 with ROM version > 1.04 and still have these problems ? I still like the XP so I hope this gets fixed.

XP-80: ROM 1.10, CPU 1.04 are the latest versions reported.

CPU 1.02 fixes a problem with the PR-C:098 patch. Sometimes when played 'popping' and 'clicking' can be heard.

ROM 1.10 fixes a bug when assigning GM patches 041-056 to Performance
PR-A:01 Part 1 no sound is heard. When loading and converting a SMF format 1, track 15 may be merged into track 16 and track 15 will be empty. In track info screen, track 16 _may_ show the information for a previously loaded song. When using soft thru and there are multiple note on messages, the XP _may_ only transmit one note off message. While recording the timing may have occasionally drifted.

ROM 1.08 fixes a bug when recording with the Quantitize function ON,
very rarely the XP may miss recording.

ROM 1.07 fixes a slave mistep bug, an external sync to clock bug and
a chain play bug.

ROM 1.05 correctes a timing problem with the sequencer syncing to patches (drum loops) from the Dance exp. board.
ROM 1.04 apparently fixed some sequencer timing problems when large amounts of MIDI data was being received.

To find out your current ROM/CPU version:
Press [SEQUENCER] button to enter Sequencer mode. Then press [SHIFT]+[ENTER]+[EXIT] simultaneously.

MIDI is a serial protocol, meaning nothing is sent through MIDI at the same time as anything else. Regardless of the sequencer you are using, sending 64 note-on events takes enough time that the delay is long enough to be perceived. If there is pitch bend, aftertouch, or too much other stuff being sent, that makes the problem even worse. Even one delayed note is enough to throw off your perception of the rhythm for several measures. The problem could easily be someone's sequence rather than their equipment; I wouldn't take seriously claims of hardware problems unless someone showed that the same sequence plays back fine with another sequencer or another synth of the same polyphony. One great thing about the XP-80's sequencer is that it makes it very easy to thin or erase aftertouch and other controller messages to avoid just this problem. You will have to look very carefully at your 64 part sequences to get them to play back right on any MIDI equipment.

I can't reproduce the problem on my XP-80, which is only a few months old. It looks like Roland has updated the ROM to take care of some specific timing issues they noticed.

I have a problem with my XP80 that I have not seen posted. While playing the sequencer the sounds played by the sequencer become very distorted and shrill. Almost like draging a phono needle across a record, but 10 times louder. I can stop the sequencer and the individual sound that I am playing is clear. This is very disturbing and disruptive to my gig to say the least. I can turn the unit off then back on and its fine. I have an idea that it may be the disk drive being dirty but I'm not sure. This happens very intermitently, sometimes months go by without incident. Anyone have a similar problem?
By the way ROLAND the 3.5 floppy is ancient history. How about a SCSI interface kit or even a zip drive. That would put the xp80 in the 21 st century where it belongs.

[This message has been edited by Randy (edited 04-20-99).]

I'm on a JV/XP mailing list and own an XP80 (ROM 1.07)myself. Personally, I have never encountered a prob with the sequencer and some of my arrangements(incl. orchestral) get pretty complicated. I do know that when I
record and I switch track/parts while recording, that will cause the drum track to be thrown off time. I need to stop recording, switch track/part and then start again. This is not inconvenient.

It has been speculated on the mailinglist that too many midi messages being sent at once will clog and cause probs. This would be a prob with most onboard sequencers. A way to get around it is to go into micro edit mode and adjust values so they aren't all played at the same time. There are also alot of internal keyboard settings that could cause probs. All you need to do is find them and reset the values.

I'm far from an authority on this stuff but I feel comfortable saying this is not a large issue. Thousands of users on the mailing list would have addressed this if it were a big problem and the XP80 wouldn't be one of the best selling keyboards. Besides, what else are you gonna buy? An EX5?

Here's another email on the subject I had saved . . .


>>Hi there folks!
I recently bought an XP-80 and I've been using the internal sequencer to
record songs. Recently I've noticed a problem with the sequencer timing
that's really frustrating me. Specifically, I've been loop-recording a rhythm
track and then copying it for the duration of the song and then recording
parts over it. I thought I was doing this correctly, but then I'll play it
back and in certain parts of the song it sounds like there's a slight delay or
pause before the next measure starts and the downbeat comes in. This seems to
happen at the beginning of the phrase that was looped (although it might
happen elsewhere; I'm not sure). But it doesn't happen every time the phrase
repeats; just on occasion. It makes it sound like I'm working with a bad
drummer. Quantizing doesn't seem to help. Has anyone else had this problem?
It's driving me crazy!

Any help with this would be massively appreciated. ( I'm a newbie, so be
kind!)<<

D Barnes wrote:
This is a problem all of we newbies tend to run into and even the experts have
to keep in consideration. It seems that even though the XP's have 64 voices, it
doesn't mean you can have 64 events occurring simultaneously when sequencing.
My understanding (and this may need to be clarified) is that MIDI can only
process 10 simultaneous events at one point time (tic). So if you have the bass
drum, hi-hat, snare, cymbal, hitting on a beat and at the same time (on the same
tic) as well as another instrument using let's say 3 tones, and another using 3
tones, thats 10 events right there. Any additional events happening at that
time (tic) will get pushed ahead and sound late. This is usually caused by
over quantizing at 100% resolution so that every note is assigned to an exact
part of a beat (or you have perfect timing :-)).

The solution is two-fold. To prevent the problem, don't use the quantize
function when recording or if you must, set it on a lower resolution, say 85% or
90% or even less so that events don't stack up on the same tic.

To fix a sequence that already has this problem you can either go into
Microscope Edit and use the Move function to move the offending notes a few tics
forward or backward (if it is only happening in a couple of spots) or use Track
Edit and use the Time Shift function (I think it is called, I'm not at my XP
right now) and shift a track or two farward or backward a couple of tics.

This is how I do it, but I'm sure there are alternative methods you will likly
see posted :-).

Good Luck

Dan Barnes