shiihs

Hi Akshansh, nice to meet you. Since I don't have a Mac I cannot make builds for Mac unfortunately. However, if you are tech savvy you can run the tool directly from source code which you can find on github. If you have no idea what I'm talking about, I'm afraid I cannot help you since I don't have a Mac myself and I don't know how everything works there. https://github.com/shimpe/ac7renamer which depends on https://github.com/shimpe/ac7parser @Chandler HollowayUnfortunately I could not find a way to quickly fix the variation swapping. Internally the variations are tagged with all kinds of numerical IDs, and I suspect that older keyboards can only work with IDs that are "small enough", and the four variation boards use IDs that have larger numerical values than the older keyboards can work with. The only way I can think of is to completely rebuild the .ac7 file from scratch, with freshly assigned IDs to everything. However, that is quite an undertaking and won't be available any time soon (especially because I'm also not spending any time on it at the moment). On the positive side, I have managed to extract several variations from a rhythmic pattern from my WK-7600 and then map each of the instruments onto another synthesizer. So this means that in principle the parser for .ac7 files is quite sound. This is how I created the rhythmic track in my "Gayatri Mantra" video: .

@Chandler Holloway I've implemented my theory about track locking and I would ask if you can try to import the attached .AC7 file, to see if it unlocks the tracks for editing without weird side effects. I've already been able to unlock tracks on my WK-7600 with this same approach. Turns out I was looking in the wrong direction at first. Instead of removing a "locking marker", I had to insert an "allow recording" marker instead. This explains why the mysterious .exe file from @jose antonio tafalla blas makes the .ac7 files longer (and my approach now also does). If it works, I can create a new build of ReStyle that unlocks all tracks for editing while saving. I still haven't been able to see why some tracks are marked as "disabled" on your system - disabled doesn't seem to exist on my WK-7600. So I'm a bit curious what happens to those tracks after the "unlocking" took place. Unlocked Straight 8-Beat CT-X3000 Fixed.AC7

Thanks @jose antonio tafalla blas. With this I have enough information for now. I think I have an idea now how the locking inside the .ac7 files works. I need to think a bit how to best proceed.

My guess is they copy more than just rhythms between models, but the .AC7 files contain a lot of extra data on the ctx3000 (you can see from the file size already: the locked ctx-3000 file is 24408 bytes, and the unlocked mz-x file is only 17103 bytes. For that reason it would be very useful if you could produce a locked file with the mz-x by importing the locked ctx-3000 file into mz-x and immediately exporting it again without removing the locks (if that is possible).

Thanks, but the problem here is that the CT-X3000 .ac7 file contains many pieces of information that were not yet supported on the MZ-X. Could you therefore import the CT-X3000 ac7, and then export it again from the MZ-X while keeping locking in place? That should make the comparison a lot easier because then I have both a locked and unlocked file saved by the same machine (containing the same types of information).

please also add the original locked one so i can compare what's different between the two versions (btw, your unlocked file has 100% correct format)

I believe you, but if I understand correctly, @casio_style used another (Russian?) sw tool to unlock the styles, and the resulting style seems to be not 100% correct anymore. It may be that the code inside the Casios have some tolerance for certain errors. So if you could upload the same .ac7 locked and unlocked using the MZ-X it'd be interesting, as that would probably behave more as we expect. @casio_style I cannot do much with an .exe as I'm not a windows user. Also reverse engineering .exe files is not really feasible, except in very specific cases.

Interesting... The "unblock.AC7" file seems to be corrupt. Internally the file size is listed as 10965 bytes (which corresponds exactly to the size of the locked rhythm), but the unblock.AC7 file now is 11133 bytes long. Can you re-import the unblocked file without getting errors?

Thanks, i will take a look!

Maybe if someone on the forum could post two versions of the same .ac7: one locked and one unlocked, that would be very instructive.

Hi @Chandler Holloway, Thanks for your interest and sorry for my lack of responsiveness. I've been busy doing other things (related to my day job and to composing music mainly) so I haven't had much time/energy left for investigation. I still intend to give it another go when inspiration for new music dries up, so thanks for the new data you've provided. Nowadays I'm less using my Casio and more some other instruments I have (although, as it happens, I've used it yesterday after a few weeks of not even looking at it - maybe the universe is trying to tell me something ). Actually, I thought about this from day 1. I had noticed when editing the rhythms that tracks from the presets could not be modified, and I was curious if I could find something in the .ac7 file that indicates that a track is locked. I have a vague idea of where to look. I believe the .ac7 file format contains certain "markers" to indicate that a track is locked for editing, and so (at least in theory) it should be a matter of removing the marker to unlock the track. It is sometimes claimed that these locked tracks contain "richer" information that one could not create on the instrument itself, and there's some truth in that (e.g. on my Casio, the auto-arranger for some rhythms produces a completely different sequence depending on whether you start playing in a minor key or a major key - I didn't find yet how that can be made on the instrument itself, but in the .AC7 file there are provisions for different intros based on the type of key.). From what I remember, the preset tracks often contain "expression pedal" information that can be used to modulate some DSP effect or other parameter, and also pitch bend information, which automatically makes things sound better (provided they are used wisely). I think with more investigation it should eventually be possible to dump the contents of these preset tracks or even unlock/edit them (although I'm not sure Casio would be very happy with such feature - I still wonder why they are so protective of these file formats. Are they afraid that people will no longer buy newer instruments if they can do everything on the cheaper ones? No idea.). I strongly suspect Casio internally has a tool to convert between midi and .ac7, and they simply record and enhance their tracks in a DAW before converting it to .ac7. To be honest, I still suspect PP stands for phrase pad: first of all because it's a new feature in the more recent Casios, and these PP sections only appear in the .ac7's from such instruments, and secondly all other sections in the .ac7 file also start with a 2 letter acronym that relates to the information that follows in the file. The locked rhythms already exist in older instruments where no PP section appears in the .ac7 file. At the moment I cannot accurately parse the PP section (and I don't have such a recent Casio, so it's not really possible for me to do experiments - of which many would be needed to get a clear view on what goes where in the file format) . Lack of accurate understanding the internals obviously opens the door to introducing corruption when manipulating the sections in the file. I hope that the new files you've provided last time (and also the ones you've provided now) will shine some more light on the pending questions.

@Chandler Holloway Good luck I'm trying to make sense of your observations but t's not trivial. One potentially relevant difference that I see between manually swapped and restyle swapped variations is in a binary section which I really don't understand (but as it starts with the characters "PP", and based on browsing the CT-X5000 manual, my guess is it refers to phrase pad data - something that didn't exist yet in the time of the WK7600 model, and so impossible for me to systematically investigate). Having said this, every combination where you had trouble after swapping shows a difference in this phrase pad section between the manually swapped and restyle-swapped files. In particular, after swapping Var(2)->Var(4), the PP section is suddenly followed by all 0's, and you describe how Var(2) after the swap suddenly sounds blank. In a naive first wild guess, I wonder if these numbers could be some kind of mixer settings. Could you check with the restyle swapped version if there's s mixer volume or related setting somewhere (perhaps something phrase pad specific?) which has been turned to 0 by the swap and which can be re-enabled to hear the correct variation? If so, at least we have a new lead of where to search further.

I'm currently located in "Vlaams-Brabant" (about 30km from Leuven) but I spent all of my youth in "West-Vlaanderen". Don't worry about test timing: as you have noticed this is a somewhat slow moving project (but the important thing is it doesn't stall, at least until it's usable).

hi @vbdx66 I've incorporated the feedback you and @Chandler Holloway gave wrt to the naming and ordering of elements in the UI. I've also swapped the meaning of var 4 and fill-in 4 which hopefully makes using those while reordering behave as expected (unfortunately I have no way to test variations 3 and 4 myself, so I 'm relying on both your feedbacks to find out what is wrong.) The link points to a single executable file which you can start from where you save it.

For the adventurous among you, here's a new version: https://drive.google.com/open?id=1Sn6RDgWCYJFPYm-zF_zH0jKlRIr-AuFN

@Chandler Holloway Sorry for my very late response. I've had some unexpected delays caused by "real life". Since I don't have an instrument with so many variations unfortunately I cannot test it myself so I have to rely on your description of what happens. Do I understand correctly from your analysis that my guess of what constitutes variation 4 and fill 4 needs to be swapped? If so, that should be a very easy fix. If not, I need to reread what you wrote in more detail Next time I will make a single-file build again. Also thanks for the UI feedback. The order in which the variations and fill-ins are listed in the UI currently conforms to the order in which they are present in the ac7 file (this made some things easier in the program). There's no strict need for that though. I'll try making it a bit less confusing. Thanks again for taking your time to test out the tool.

hi @vbdx66, nice to meet you. Excuse me for the slow response. I have a performance coming up (tomorrow) for which I still need to fix some details and I'm generally low on free time at the moment. For installation it should suffice to download the complete folder located at https://drive.google.com/open?id=10ddxz5WB7DA8S8BPi5HR_yM0pN5g-z5B and putting all the files in a folder of your choice. There should be one .exe file inside which you can run to start the tool. No programming required I think the interface should be more or less self-explaining, but if you run into trouble with installation or running, let me know and I'll try to help you figure things out next week. As for swapping variations between two different rhythms... I'd need to think about the implications of starting to mix up things (if you take a complete variation I don't expect big problems, but if you take only certain instruments e.g., then you have to also check that the number of bars between the two rhythms is compatible etc...). My ultimate dream is indeed to have a full-fledged ac7 editor which allows editing "everything", but outside holidays it's difficult to find enough time to dedicate to the project. Also the complete lack of documentation about the file format is not exactly helping me to achieve that goal faster . But I haven't forgotten about the project yet.

In an ideal world, it should work for all valid .ac7 files, no matter where they are created or used. Of course it remains to be seen if that is attainable, but so far there haven't been any major showstoppers.

I hope google drive allows for downloading a complete folder at once (as opposed to downloading each file separately, which would take "forever"). Or feel free to let me know if you prefer the one file again. It shouldn't be overly difficult to trace back what I did to make those. Once we have something that works well enough, I can investigate making a windows installer so people can easily deploy it to their system if they so desire.

Oh... I didn't make a onefile build this time. You'd need to download the complete "ReStyle" folder (all the files) and put them together in a folder of your choice. After the change to a different computer, I kind of forgot how I made these builds before 😞. Was it just single file or a folder of files?

Did you turn up the volume button? Also check that you didn't plug something into the headphone jack by accident? (Yes, I know these suggestions are kind of "basic" but perhaps it's best to exclude the obvious first.)

@Chandler Holloway I believe I've found and corrected the problem. Could you please retry with the new version? Also... I've had to switch to a new computer as the previous was broken so I hope I didn't mess up something while reinstalling the new pc. Please let me know if you have trouble running the new version. Thanks! I'll just post a link to a windows binary build for now, as I think the linux builds are not very popular, and people using linux probably are not afraid of running from source code directly. (If someone needs it, please shout and I'll provide a binary build for linux as before). https://drive.google.com/open?id=10ddxz5WB7DA8S8BPi5HR_yM0pN5g-z5B

@Chandler Holloway Thanks a million for testing. I'm currently fighting with some deadlines for upcoming performances - it's the most wonderful time of the year after all - but when I find more time (hopefully around the beginning of next year) I'll try to address the problems you found. It seems like I've messed up something indeed. Thanks a lot for providing the before and after files, I expect they will be very useful for diagnosing and fixing the problem.

It looks quite fancy already. From what I've read, different IDES versions are really tailored to specific models of CASIO keyboards, so that may be an indication that the internal structures between these models are quite different. Your tool being written in C#, I'm not sure how easy it will be to test on Linux. I know there's mono (plagued with bugs from what i read, and .NET core (but most likely you will rely on extensions that are not supported on Linux). Just one more remark: please start a new topic for further discussion about your tool. Let's keep this topic on-topic.

Sounds pretty impressive to me!