Here's the scenario: I create patch user05, and call MyCoolPatch. I write it to user memory. Then I create performance user02 and call it MyCoolPerf, and write that as well. MyCoolPatch is voice 4 of MyCoolPerf. Then I write a song using MyCoolPerf, and save it to disk (song and sounds).

Then I decide to restore the XP-60 to factory defaults, destroying my patch and perf. I load the song. What sounds does it use? Have I lost my patch and perf?

How does this work? The manual is not really clear on this. Can someone explain it to me?

Thanks!

See, I was never really clear on this issue. I may not fully answer question, but I will start a dialog.

See, I use to have an XP50. I have all of these patch banks from various sources, plus I create some of my own user voices when I edit. On the XP50, there was no [Save Song+Sound] option. You always had to save the song and the sound separately.

Fast Foward a year and a half later. I buy the XP60. I see this [Save Song+Sound] option. I say to myself "Yeah right". See, my pre-assumption is that you can use this option if you consistently use the same patch banks together for every song you make.

Suppose I loaded in the "Dance Kit" and used that for my song. Then I reset the XP to the Factory settings (This means I have the original patch bank). Then I load in my Saved Song+Sound and assume that the song will play properly. To my knowledge, this would select the right patch numbers, but not the right sounds because they don't exist in the default bank.

Okay, maybe I should shut the hell up now. See, I never actually tried this because I'm too chicken to do that. We'll wait until someone clears this up for us.

Lastly - Did you remember to use the [Patch+Utility+Write] followed by the [Performance+Utility+Write] functions to save your Patch to the user area and then assign the Performance to realize that you changed a patch in the performance? You must do both.

The Infamous EPU.

Yes, that much I did.

The reason I ask this question is that I've had songs where patches suddenly disappeared, then magically reappeared weeks later.

If no one else has a definitive answer, and if I have time tonight, I'll do some experiments to see if we can figure it out.

I think when you save in SONG+SOUND mode, performance data and song data are saved, but no patch data is saved. The song/sequencing data, the tracks/parts of the performance, and performance data (effects, sync, etc.) are all saved. Assigning a patch or drum kit to a performance part only instructs the performance to play the sound at the memory location instructed.

If you create a patch (USER005:MyCoolPatch) and a performance (USER002:MyCoolPerformance) and assign USER005 patch to part 4 of the performance, then save it to disk, then initialize (restoring factory sounds to USER memory), here's what will happen when you try to load your song/performance: The performance "MyCoolPerformance" will load into temporary memory (has an asterisk by it like a patch that has been edited but not saved to USER memory) as will all of your song data. However, when part 4 of the performance plays, it will play whatever is currently in USER005 patch, not the patch you had created. The same holds true for USER drum kits. Does that make sense?

Anyway, if you want to save USER patches to go with your songs, you'll have to save all user memory as a sound file, then load it later when you load the USER performance/song.

I actually just got a reply from Roland about this very subject, because I'm afraid of losing all my custom patches as well. What they told me was to simply use the "save sound" feature in the disk menu. Apparently this will save your actuall sounds, because "save song & sound" only saves the midi info......or something. I hope someone is willing to test this, because I'm too chicken to do it as well

------------------
Edges of Seven

Saving sounds separately (onto disk) is what I was talking about in the last paragraph of my last post on this thread. Don't worry about saving your custom (USER) patches to disk and then writing over the USER patch memory...you can load your patches back into USER memory from the disk. When you save in sound mode, the entire USER patch, rhythm, and performance memory is saved and will be recalled when you later reload the USER sounds from the disk. Although all 128 patches, 32 performances, and 2 drum kits are saved each time, very little memory is required, so you can fit a bunch of sound files onto one disk.

So you are saying that really ALL of the important information is stored on my 3.15" disk? (Or more directly: can I go back to "factory presets", thus wiping all of my creations out, and then simply load the .sound file into my XP again and thus reincarnate all of my patches & performances, or will either one be incomplete/lost if I try this?)

I must admit I've never tried the "back 2 factory presets" procedure beacuse I thought space was convenient, but now my USER bank is somewhat full...

Thanx in advance!
XPert

I can tell you from experience that saving all the USER data to disk works great! You can totally initialize the user factory presets and then load up your previously saved SOUND file. When you load your stuff everything is just like you saved it, including which patch/performance/rhythm set you were on 'when' you saved it! It really is amazing how little memory it takes up on the disk too.

So when working on a song, I usually do one of two things:

#1: Use the method described above and Load the SOUND file that corresponds with the SONG or SONG+SOUND that you want to play. This is good for when you have several patch or performance changes in a song, so you have access to all 128 User patches/32 performances/and 2 rhythm kits.

#2: This method is good to use when saving songs as a .SMF0 or a .SMF1 (this is the usual method for saving songs to MIDI). When saving songs as a Standard Midi File, there is only one way to 'save' the sounds you used to make sure that they are the same each time you load your song. That is by using Sysex (I can hear the cringes now).

When I first started programming with sysex I thought it was really confusing, but it's not bad at all depending on how much you want to do. The JV's and XP's make saving sound data for each song a cinch! Just make sure you have about 4 measures of blank space at the beginning of your song. Just go to UTILITY-> DATA TRANSFER-> then you get to decide just how much you want to save from the user memory...Usually I have a Performance that I've made changes to so that little star is by the name. I then choose PERFORM TEMP: +PATCH By doing this, you save only the temporary performance plus all the patches 'currently' being used at the time you performed the transfer.

The great thing about this method of saving is that the sounds load right up with the song! It's great if you only have 16 parts to use with no patch changes. The sounds load at the beginning of the song and does so even if a patch is changed later. The downside to this method is that if you need to use more patches, they either better be preset patches (so you know what you're getting) or you must always keep your USER memory the same (so you know where you're getting the sounds and which ones...I don't like this idea).

This post is getting long so I'll finish up in another one.

Jake

Please pardon my previous post. It was long and had many grammatical errors. However, I assure you that I'm trying to be helpful in any way I can. This one will probably have more errors, so enjoy!

Using method #2 is also good if you are running low on floppy disk space. The .SVQ files take up more space than MIDI files (I prefer 'SMF1' because it keeps the tracks separate whereas 'SMF0' puts all the tracks together. That sucks and in my opinion is unnecessary!) So anyway, to get the best of both worlds, save your USER data as in method #1 then load it up before you play the song. Together with that, save your temporary patches in the first 4 measures of the song as a SMF1 file. This might seem like overkill to some, but I like to be as safe and efficient as I can be.

But if speed is more important to you than disk space, saving all your USER memory as a SOUND file will do the job. And if you like the song to start right up, loading the SOUND file and then simply playing back the SONG should do the trick! Just be sure you have a specific SOUND file to load for each song you have. This is really the best way to work. So have fun, relax, and don't worry...your sounds can all be restored from the disk.

Jake

And to avoid any confusion, I suggest that whenever you see SONG in my posts, think of that as SONG+SOUND. I think that's the only way to insure everything will come out right. It takes up a lot of memory though, which is why I prefer using SYSEX at the beginning of the song. The SYSEX routine can be used with .SVQ songs or Standard Midi Files, it doesn't matter. Isn't that great!?

On a side note, I've been posting in the EXPANSION BOARD post lately and I need some help. If you haven't checked out my posts, please do so if you have the time. And if you checked them out but didn't add anything well, screw you! I'm kidding, just help out if you want to. hehe

Jake
aka: MappyPlus