PG Music Home Forums PG Music Product Forums RealBand Glitch and erroneous chord change occurring between bar 240 and bar 241 [Resolved]

I guess that's the only thing I can do with RB besides giving Tech Support another phone call, which I might do. Thanks for your help. Hopefully, the next release will be available before October.

I want you to try this:
1.open RB
2.set song to 330 measures
3. Put a chord at 238 then at 245.
4. Generate any Rt
5. Listen from bar 235 past 245.
Do you get an erroneous chord change at241?

After setting the time signature to 3/4, I entered a G chord into bar 238 and a D chord into bar 245. I didn't enter any other chords anywhere else. Then I selected "Select and generate a RealTrack" from the menu and picked #856 (12 string acoustic guitar). When RB was finished generating the track, I clicked on bar 235 in the Chords window and clicked on the Play button. I could definitely hear a chord change at bar 241---most likely to the C chord.

Ok thanks. If you open biab you'll see that the first chord is always C. That seems to be what happens at the 240-241 transition. I'd bite the bullet for now by putting the chord in 241 and see if the workflow is ok. You'll have to wait for the new V4. Soon.

I can do that, and I have done that because it's the most obvious workaround. But sometimes a brief glitch will appear at the 240/241 bar boundary, which can be a challenge to get rid of. Fortunately, there's no pressure for me to create a demo of this song for my producer like there was with the most recent song I gave him a couple of months ago. But I was using BiaB at the time and didn't even know that RB could handle songs with more than 255 bars. Plus, I had figured out my own workaround to BiaB's 255 bar limitation. However, as I found out during the last few weeks of discussions here and on the BiaB forum, my workaround isn't the best or most efficient. Nevertheless, it got me through the crunch I was in at the time until I was able to explore other solutions through these two forum discussions that I started. In the process, I discovered the existence of this bug in RB at the 240/241 bar boundary, and I hope that Tech Support will invest the time and effort that's needed to fix it.

RB uses a BiaB executable to generate. It may be that is where the root cause lies.
Just a thought.
Many workarounds have been mentioned. Workflow, workaround whatever.
You can get it done in RB ..
/I probably should have avoided this thread, but I guess my suggestion is just 'get r done' however it works best for you

I think you and Silvertones are both right about that, but I also think that it's possible for RB to somehow override BiaB's default C chord. Tech Support just has to find a way to implement it.

Also, this discussion is essentially over until Tech Support does come through with a fix, so your post and your suggestion to "'get r done' however it works best" are fine. In the meantime, I'm hoping that the upcoming release of RB 2019 Version 4 will include a fix for this issue.

"I'm hoping that the upcoming release of RB 2019 Version 4 will include a fix for this issue."

I think that's the ultimate resolution you'll be satisfied with.

Glad to let the drama of it all rest while we wait for next version to release.

So now may be a good time to raise this question. I didn't want to bring it up before since I hate to nitpic and it is really none of my business. But, why does your signature specs indicate you are using "MuseScore 2.1"?

That's a fair question, MusicStudent. The reason I'm using version 2.1 instead of 3.1 is that I don't want to risk having to tweak the layout of my current song files like I did the last time I updated MuseScore. The amount of time I may end up losing by making such tweaks isn't worth the extra features I'd be gaining with the newer version. Plus, I'm already on overload from dealing with numerous other stressful issues in my life (which I won't go into here), so I'd rather stick with "the known" than venture into "the unknown" until I'm in a better position to deal with the potential time loss factor. Of course, it's possible that version 3.1 won't change (mess up) the layout of my current song files. But I'm skeptical that it won't because this has happened to me several times before after I updated MuseScore.

Thanks for the answer, that makes sense. Good luck with your music.

Now, that's an interesting statement, Charlie, for two reasons. One reason it's interesting is your use of the words "ultimate" and "satisfied." While I would be "satisfied" if this occurs, I have to say that the "ultimate" resolution I'd be "satisfied" with is if the 255 bar limit in BiaB was eliminated completely so that I wouldn't have to deal with one or more of the numerous workarounds to this limit for any of my songs that exceed it. I also wouldn't have to consider the option (along with the pros and cons) of using RB instead of BiaB for such songs of mine that are still being developed or that I may write in the future. But as you and I both know, the chances of this limit being eliminated in BiaB any time soon (or even at some point in the not too distant future) are next to none.

The other reason your statement is interesting is that it could imply that the resolution of this issue in RB would mostly benefit me. In other words, your statement has the appearance of shifting the focus away from a legitimate issue in RB that does need to be resolved and placing that focus on me personally. Now, I'm not saying that this is the actual intention of your statement because I don't know and can't know for sure that it is. But this is a definite possibility, so I think it would be good for me to dispel such a notion in case you or anyone else has embraced it. I'll begin by explaining why I believe that such a notion is a possibility.

From my perspective, this entire discussion of the existence of two problems at the 240/241 bar boundary in RB has been like a wild roller coaster ride to the moon and back because of how many other topics (some closely related and some not so closely related) have also been discussed in this thread. After all, as you pointed out in a previous post, this thread (which curently has over 190 posts) is by far the longest and most involved thread of any thread in the last seven years past year on this entire forum. The gist of these two problems are as follows:

1) a glich that intermittently occurs at the 240/241 boundary as a result of an attempt to fix the other consistently occurring problem (by inserting a chord into bar 241), which is

2) an erroneous chord change to C when no chord exists in bar 241 and when the most recent chord in the song was not a C chord.

Considering the fact that (please correct me if I'm wrong about this) I am the first person to discover these two problems and the fact that I discovered these two problems as a result of using RB to open an .XML file of one of my songs that contains more than 255 bars, I need to make it clear that these problems are not related in any way to my .XML files (contrary to Silvertone's assertion in his post on 06/19/19 07:19 AM PT). In other words, contrary to the perception that some readers of and participants in this discussion may have had about these two problems (including myself as well initially), they are not due to RB's alleged improper handling of .XML files or to MuseScore's improper creation of .XML files. Instead, these two problems also occur when chords are manually entered into the Chord window after which any one of the "Generate Tracks" commands is executed in RB.

Considering the fact that a substantial amount of time was spent on various workflow processes in this discussion, I also need to make it clear that these two problems have nothing to do with my workflow. In other words, because these problems are not related in any way to my .XML files, and because my standard use of both BiaB and RB is to begin (as my first step) by opening a .XML file that I exported from MuseScore (rather than by manually entering chords into the Chords window of either program, as most users of these programs normally do), these two problems also have nothing to do with my workflow (contrary to your assertion in your post on 06/20/19 at 08:07 AM PT and contrary to MusicStudent's subsequent post).

Considering the fact that, even though this discussion has been ongoing since 6/7/10 and has 190+ posts so far, as recently as 06/20/19 06:07 AM PT (per a PM he sent me two days ago and which he reiterated in his public post on 06/20/19 12:35 PM PT), Silvertones thought that the newest RB 2019 Version 4 release candidate had fixed these issues and that the .SEQ file he had created from that version (to which he provided a link in his post on 06/20/19 12:35 PM PT) was proof of that. He also thought that these issues were directly related to RB's improper handling of .XML files. But when I opened that .SEQ file in RB (after I updated my RB V2 to V3), I was able to demonstrate that the issue had not been fixed in the latest release candidate version either. It was only after this that I was able to convince Silvertones that the issue had nothing to do with RB's handling of .XML files when I followed his instructions to test a new file in RB that I created by generating a single RealTrack after I had manually entered a chord into bar 238 and a chord into bar 245.

Finally, considering the fact that everyone who has used RB to automatically generate tracks for songs with more than 240 bars has ended up with an erroneous chord change at bar 241 of their songs (if there wasn't any chord already entered into bar 241 and if the most recent chord in their song was not a C chord), there shouldn't be any reason for anyone to assume that I would be the primary (if not the sole) beneficiary of a fix to these issues. Furthermore, I find it very surprising that no one else who has used RB for this purpose discovered these issues before I discovered them about two weeks ago. Surely, if soneone had discovered them, either these issues would have been fixed long ago or I would have been informed early on in this discussion that these are known issues that Tech Support hasn't fixed yet.

muzikluver,

I believe you posted an accurate summary of the main points in this thread so far. The discussion has touched upon many other topics outside the points you've raised.

But, I'd like to point out that it is pretty normal in this forum for a thread to venture way beyond the initial subject of a thread. But the discussion has been worth the effort because the discussion has helped to clearly define the problems and eliminate causes external to the program.

Sure, Jim, absolutely! I agree with you completely. And please don't get the impression that I was complaining or whining or criticizing this discussion in any way because I wasn't. I just mentioned these things for the purpose of summarizing the evolutionary path that this discussion has taken and to zero in on my main takeaways from all the various topics that have been touched upon in the process. Good point. Thanks for pointing this out!

<< Now, that's an interesting statement, Charlie, for two reasons. One reason it's interesting is your use of the words "ultimate" and "satisfied." While I would be "satisfied" if this occurs, I have to say that the "ultimate" resolution I'd be "satisfied" with is if the 255 bar limit in BiaB was eliminated completely so that I wouldn't have to deal with one or more of the numerous workarounds to this limit for any of my songs that exceed it. I also wouldn't have to consider the option (along with the pros and cons) of using RB instead of BiaB for such songs of mine that are still being developed or that I may write in the future. But as you and I both know, the chances of this limit being eliminated in BiaB any time soon (or even at some point in the not too distant future) are next to none. >>

I think a Fix of BIAB and RB's 255 Bar limit would be your and anyone's ultimate satisfaction because a fix is necessary in order to complete the expected and ordinary task of creating a chord chart, choosing a style and pushing the play button and BIAB/RB generates its soundtrack of the song. Ultimately, this should happen whether the song is 25 bars or 255 bars or 555 bars. My comment was spoken to you but meant for everybody. If you read the statement I highlighted and responded to, it referred to version 4 (which I don't have) may 'fix' the issue. Those highlighted words I replied to are your words not mine.

Personally, I don't see the difference between "I have to say that the "ultimate" resolution I'd be "satisfied" with is if the 255 bar limit in BiaB was eliminated completely so that I wouldn't have to deal with one or more of the numerous workarounds to this limit for any of my songs that exceed it." you say today and " I'm hoping that the upcoming release of RB 2019 Version 4 will include a fix for this issue." you said on 6/21 other than the number of words you used.



<< The other reason your statement is interesting is that it could imply that the resolution of this issue in RB would mostly benefit me. >>

Not at all. Everyone that will ever create a project that taxes this limitation will benefit. This limitation has been known for years and requests have been made over those years for a fix.

<< From my perspective, this entire discussion of the existence of two problems at the 240/241 bar boundary in RB <Snip> The gist of these two problems are as follows:

1) a glitch that intermittently occurs at the 240/241 boundary as a result of an attempt to fix the other consistently occurring problem (by inserting a chord into bar 241), which is

2) an erroneous chord change to C when no chord exists in bar 241 and when the most recent chord in the song was not a C chord. >>

Yes. I agree with your statement completely. I thought it was an interesting and compelling problem and enjoyed all of the thought and comment the various posters have made to be pertinent and instructive.

<< Considering the fact that (please correct me if I'm wrong about this) I am the first person to discover these two problems <snip> they are not due to RB's alleged improper handling of .XML files or to MuseScore's improper creation of .XML files. >>

Correction regarding the 255 limit, that's been known for years and was already an old issue when I joined the forum years ago. Regarding the chord glitch, I don't know about that. I've never seen it mentioned before but I also have never previously participated in a discussion about these two glitches so it could have been mentioned and I missed it. But as far as I know, you're the first to mention the chord issue. It seemed to me in his response, and this is just speculation, but Joe with PGMusic seemed to be aware that approaching the 255 bar limit caused RB to begin to act in sporadic and unstable audio renders. Whether that came from PGMusic programmers testing or clients reports I don't know. My speculation could be completely wrong too.

<< Considering the fact that a substantial amount of time was spent on various workflow processes in this discussion, I also need to make it clear that these two problems have nothing to do with my workflow. In other words, because these problems are not related in any way to my .XML files, and because my standard use of both BiaB and RB is to begin (as my first step) by opening a .XML file that I exported from MuseScore (rather than by manually entering chords into the Chords window of either program, as most users of these programs normally do), these two problems also have nothing to do with my workflow (contrary to your assertion in your post on 06/20/19 at 08:07 AM PT and contrary to MusicStudent's subsequent post). >>

I agree with this for the most part too. As I noted above, your workflow is fine except that BIAB/RB doesn't react correctly to generate your audio without error. As I also noted earlier, for me, those glitch issues are no more an issue than turning down the gain knob to keep a channel's input signal from clipping. I don't encounter the glitches you discovered by using a method where I generate or record audio in sections small enough so the number of bars is never an issue. That's how my workflow works anyway so I'd never encounter the two limitations as you did even If I had a project in RB that exceeded the 255 bar limit. Others also work in a similar manner and that's why they say that RB is not subject to the 255 Bar limit the same as BIAB. My workflow will allow RB to have an unlimited amount of bars subject only to what my computer can tolerate. If RB in my computer can accept 1,500 bars, I can do that and there will absolutely be no audio glitches, guaranteed. I drop pre-recorded audio into RB or record short lengths of audio into RB rather and then if I apply any regeneration done by BIAB/RB, it's done in sections rather than one large single render. I do this utilizing multi track techniques called overdubbing and Punch in/out.

As it turns out, it does have to do with your workflow because apparently BIAB and RB utilize the same architecture to generate audio tracks, and because with your workflow, you attempt to generate more than 255 bars in a single render. It's something BIAB and RB should do, but neither does. If the two glitches are eventually fixed in both BIAB and RB, then you and everyone else will be able to generate tracks the same as those that now generate songs with a smaller number of measures.

If you happen to get the mixing book I referenced, it is all about developing a workflow and it will work with any DAW including BIAB/RB. It will also bypass the 255 bar limit. If you adopt its principals and techniques it will be a change in workflow and a workaround as far as the glitches in BIAB and RB go. By using the multi track recording techniques of overdubbing and Punch in/out and working in sections of your project rather than a single full length render, the issues presented in a full render attempt will never be an issue.

<< Considering the fact that, even though this discussion has been ongoing since 6/7/10 and has 190+ posts so far, as recently as 06/20/19 06:07 AM PT (per a PM he sent me two days ago and which he reiterated in his public post on 06/20/19 12:35 PM PT), Silvertones thought that the newest RB 2019 Version 4 release candidate had fixed these issues and that the .SEQ file he had created from that version (to which he provided a link in his post on 06/20/19 12:35 PM PT) was proof of that. <snip> >>

That's what I had read at the time I made my post. Who knows if or when a fix may come?

<< Finally, considering the fact that everyone who has used RB to automatically generate tracks for songs with more than 240 bars has ended up with an erroneous chord change at bar 241 of their songs (if there wasn't any chord already entered into bar 241 and if the most recent chord in their song was not a C chord), there shouldn't be any reason for anyone to assume that I would be the primary (if not the sole) beneficiary of a fix to these issues. Furthermore, I find it very surprising that no one else who has used RB for this purpose discovered these issues before I discovered them about two weeks ago. Surely, if someone had discovered them, either these issues would have been fixed long ago or I would have been informed early on in this discussion that these are known issues that Tech Support hasn't fixed yet. >>

I never assumed you to be primary anything other than you were primarily the person I was having the conversation with. As noted, the 255 bar issue is a long known issue, most of the others over time have abandoned generation lengths that get near or exceed the limitation.

"So, I can't really use this song as a basis for answering your other questions." Yes. Your answer works perfectly.


"So, my plan is to use BiaB and/or RB to create a decent instrumental demo..."

Even without a fix for the glitches. BIAB/RB will work perfectly for you. Just not the present way you use them. You can still create your song using MuseScore to make and use an XML file. You can still provide him with a MIDI file, the lead sheet and lyric sheet and a BIAB/RB rendered MP3 file along with the MP4. You can also give him full length individual MP3 tracks of any BIAB/RB generated track such as the bass, fingerpicked guitar and strumming guitar as you did on this recent project. You can do this with any BIAB/RB instrument and including as many tracks as you want.


Regarding your producer requesting changes and edits, you can not only send him a new midi file, but also all of the other tracks your furnished him updated with the edits. You can literally do these type edits in minutes not hours using overdubbing and Punch In/Out techniques and arranging your song in logical sections. Were I doing it, I would generate an audio file for whichever specific section that contains the edit in BIAB and export that audio file and then copy/paste the edited audio to replace the original audio section in RB. The whole file does not have to be regenerated. Either RB or BIAB will crossfade the entry and ending bars into the adjoining original bars seamlessly.

Agreed. Thanks for the clarification.


The difference is that the elimination of the 255 bar limit in BiaB would make it possible to use BiaB to generate a complete arrangement with one click of the mouse for songs that exceed that limit in the exact same way that BiaB can currently do this for songs that don't exceed that limit. This currently cannot be done, and that is why everyone (including me) has to use a workaround in BiaB (which will require a different workflow) or use RB and be forced to take extra steps in order to deal with an erroneous erroneous chord change in bar 241 (per item #2 in my earlier post). This was the point I was making with my statement.


You misunderstood my statement. I was referring to the two problems that exist in RB at the 240/241 bar boundary, not to the 255 bar limit in BiaB. While the 255 bar limitation has been known for years, these two problems were only recently (about two weeks ago) discovered by me as a result of using RB to generate an arrangement for one of my songs that has more than 255 bars and that meets the requirements of item #2 in my earlier post.


As I said above, I know that the 255 bar limit has been known about for years, so I wasn't claiming to have discovered it. However, it's possible that someone else had discovered the chord glitch before. Also, due to the intermittent nature of this glitch and the likelihood that a track regeneration procedure may make the glitch go away, it's also possible that anyone who did discover this glitch simply never bothered to submit a post about it to this forum. So, I'm willing to relinquish my claim to be the first one to have discovered it and to say instead that I may be the first.


You make a good point here that I believe explains why RB only generates tracks in 240 bar chunks instead of in 255 bar chunks. This was one of the mysteries about the occurrence of the erroneous chord change and glitch at the 240/241 bar boundary that we had discussed during the early stages of this thread.


For you to clearly understand my claim that "these two problems in RB have nothing to do with my workflow," I need to differentiate between the results I've gotten from BiaB vs. the results I've gotten from RB with my workflow and point out the premise upon which I made this claim. Before doing so, however, I want to exclude the occasional glitch issue from the picture because we haven't done enough testing to properly characterize this issue and because its occurrence seems to be a byproduct or side effect of any effort to compensate for the erroneous chord change issue.

Every time I have generated a complete arrangement in BiaB for one of my songs that have less than 255 bars, I have never encountered an erroneous chord change anywhere in those arrangements. So, when I decided to try using RB to generate an arrangement for one of my songs that has more than 255 bars, I made that decision based on jazzmammal's claim that the issue of the 255 bar limit has been resolved in RB. I also made that decision with the intent of using RB in the exact same way and with the exact same purpose for which I was already using BiaB. As I've mentioned several times before in this thread, the primary purpose for which I've been using BiaB since I purchased it in December 2018 is twofold:

1) to make it easier for me to work out the kinks in my songs that I'm currently writing, and

2) to quickly create a decent, basic instrumental demo of those songs after I've finished writing them so that I can share my demos of those songs with other people for various reasons

However, the first time I used RB to generate a complete arrangement for one of my songs that has more than 255 bars, I encountered an erroneous chord change that occurred in bar 241 because of how RB implements BiaB's track generation algorithm. If my song already had a C chord in bar 241 or if the most recent chord in my song prior to bar 241 was a C chord, I wouldn't have encountered an erroneous chord change in bar 241. So, while it's true that RB has resolved the 255 bar limit that exists in BiaB, it's implementation of BiaB's track generation algorithm is faulty because of its erroneous change to a C chord in bar 241 in songs that don't meet the above criteria. This has nothing to do with the fact that I chose to generate an arrangement for the entire song as I normally do in my standard workflow procedure that I've been using for the two-fold purpose I mentioned above. The reason I say this is that the same thing would have happened if I had only selected bars 235-245 and then executed RB's Generate command. And it wouldn't have made any differenct if I had generated one track or if I had generated all the tracks. I know for a fact that this is true because I've done both of these things and seen the same thing happen both times.


I am planning to get that book in the near future. But as I explained above, the erroneous chord change issue in bar 241 will occur even if I "adopt its principals and techniques" and change my workflow accordingly. This why I keep saying that this issue has nothing to do with my workflow.


Because of the point I made above about the occurrence of an erroneous chord change at bar 241 (if the above chord criteria is met) regardless of workflow and regardless of whether an entire track is generated or only a section of that track that includes the 240/241 bar boundary and regardless of whether all tracks are generated or only a section of those tracks are generated that includes the 240/241 bar boundary, I think it's wrong for us to think that a fix to this issue won't come and that it won't come soon.


Although it wasn't certain to me that you were making such an assumption or that anyone else was making such an assumption, I appreciate the clarification nonetheless. Regarding other users, as long as they are using BiaB only, they won't encounter these two issues anyway. But if they use RB to do any track generation that crosses the 240/241 bar boundary for a song that meets the chord criteria I described above, they will still encounter an erroneous chord change in bar 241.

Here's a question for you.
Does it always occur at bar 240? Have you tried any other length?

It always occurs at bar 241 if the chord criteria are met that I described in my previous post. Yes, I did regenerate a 10-bar section from bar 235 to bar 245 that involved all the tracks as well as one that involved just one track, and the erroneous chord change occurred both times.

Try a 520 bar song. Does it do it in 3 chunks? Not at my computer.