acces ROM waveforms when installed on E2 synth ? #77
Replies: 9 comments 38 replies
-
|
Yes, this will be possible in the future. When the PCM data is loaded from the flash, it's decompressed and sent to DSP RAM one word at a time. I can hijack this to copy PCM to CPU RAM, then dump it over sysex as a single blob. If I can get a debugger attached to the DSP then I may be able to fix the issue with sample saving and have the factory samples loading like normal. If you're short of samples until then, check out the carved beats set by @BKLronin. |
Beta Was this translation helpful? Give feedback.
-
|
What a great collection of samples! Thanks for making them available @BKLronin ! |
Beta Was this translation helpful? Give feedback.
-
|
Thank you very much. They were exclusively recorded and synthesised for hacktribe. |
Beta Was this translation helpful? Give feedback.
-
|
@bangcorrupt I've just reread this and thought about the implications: would this mean that its in theory possible that you would be able to in the future modify the firmware to send/receive sample data via Midi sysex? |
Beta Was this translation helpful? Give feedback.
-
|
Hey I would like to bump this as I am stuck on this exact step. e2sysex.py: def __init__(self, input_port='electribe2 sampler 0', output_port='electribe2 sampler 1', midi=None):
logging.debug('Initialise SysEx')
if midi is not None:
self.midi = midi
self.inport = self.midi.inport
self.outport = self.midi.outport
else:
self.inport = mido.open_input(input_port)
self.outport = mido.open_output(output_port)when running the script I am stuck at the same spot as @F-Thom (maybe you got more progress by now?). output of the script: That is where everything blocks. I cant seem to get a response from the other SysEx functions you mentioned, the e.get_current_pattern() and so on - it just runs indefinitley. Using Windows 10 and VSCode for that matter. Hopefully you can help and I am happy to provide any further information if necessary! Thanks for the effort in the whole project, looking forward to testing it with the PCM Samples. |
Beta Was this translation helpful? Give feedback.
-
|
I don't know why this isn't working for people. I have just tested and it worked for me, taking a little over a minute and giving me a 16MB wav file of the PCM data. I did get a 'Resource temporarily unavailable' message, possibly due to flooding the input with a large sysex message. Maybe this is handled differently on Windows and causes an error. I will try to test with Windows, but it might not be soon. |
Beta Was this translation helpful? Give feedback.
-
|
Can you check you have 'python-rtmidi' installed, not 'rtmidi-python' (these are different packages)? Could you also check which backend mido defaults to on your system? In a python console: |
Beta Was this translation helpful? Give feedback.
-
First of all, thank you very much for coming back to us and taking time to check this topic again. |
Beta Was this translation helpful? Give feedback.
-
|
Ok I finally got time to check on that again. The script however does not display the time stats that your log stated, it just stops at Writing pcm-dump.wav. The result of the mido.backend is:
Thanks again for your work, and I will now continue the workflow. If anyone has any questions maybe I can help a bit on that part! |
Beta Was this translation helpful? Give feedback.
-
|
I joined up here to ask this! I'm an E2 Synth user and I mostly use it for making ambient sequences and textures - I literally never use the 90% of the stock sounds (OSC types 001- 325 never get a look in) but I do get a lot of use out of 326-428 (the Synth and Instrument categories) My ideal would be to keep those, trash the rest, get those lovely new synth oscs that you've added, and maybe eventually add a few samples. Is this ever likely to be possible? Thanks for all your inspiring work |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, sent you an email |
Beta Was this translation helpful? Give feedback.
-
|
The firmware you sent me matches the one I tested and when I run it I get the samples dumped correctly. The audio files you sent are the correct length but look like unitialised RAM, matching the data in the raw binary files. This is what I would expect to see if the script was run against the main Hacktribe branch, or if the read address is somehow getting mangled during transmission. Can you confirm that you only have one Hacktribe connected via USB when you run the script? |
Beta Was this translation helpful? Give feedback.
-
|
As it appeared, my SD card already contained a folder for Sampler version with e2sample.all so E2 never decompressed PCMs to memory and I was getting empty memory dump. I removed the SD card before dumping PCM samples and everything went fine, I got the proper WAV. Thank you very much. Two following questions:
|
Beta Was this translation helpful? Give feedback.
-
I'm working on format definitions similar to those in e2_formats.py, along with a script to read the full PCM section from flash. These will allow parsing the binary to a dictionary, including a list of offsets for each sample. I think there is at least few bytes of zeroes between each sample, so splitting by silence may work, but if you can wait a while I should be able to script it.
I don't know the compression algorithm used, but reversing it could be possible. If I can't get the remaining features from the synth working in Hacktribe I might look at loading custom samples to the synth. In the long run I want to move all the content to the SD card in a single project file, maybe using the flash for alternate firmware apps. Thank you again for your continued support 😃 ! |
Beta Was this translation helpful? Give feedback.
-
|
Hello i would like to ask if there is more information on the structure of the wav or a script that splits the audio into the individual samples already? I have loaded it into Audacity and tried to label the samples by silence but that did not work well, there seem to be a minimum of a few audio points (samples) between those samples that i see. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, iirc the old "turn synth into sampler" header change hack allowed us to get the samples too. What happens if I run the python patch script but just use it to change header on src file (without applying the hack patch) would commenting a few lines of code (and calculating the new hash) do the trick or will it brick? If we can do this then we'd just export samples to SD, move them to 501 an on and use them. |
Beta Was this translation helpful? Give feedback.
-
|
It will prompt to install the synth firmware. Also, sampler firmware won't read the multi-sampled synth PCM correctly. |
Beta Was this translation helpful? Give feedback.
-
|
@bangcorrupt , Could you please tell more about multi-sampling support on E2/E2s? |
Beta Was this translation helpful? Give feedback.
-
|
They both support mullti-sampling but the sampler doesn't implement it fully. The sampler flash PCM section is the same format as the synth, but every sample is in a group of one. On playback the sampler reads and uses the empty key and velocity group data, I was able to map samples by editing values in CPU RAM. I didn't look into it much further, but I think when the sampler loads PCM from flash it assumes each sample is its own group which leads to bugs with instrument indexing and slice data saving. I think some day we should at least be able to load a map file into RAM after booting, hopefully it can be integrated into the existing sample loading. |
Beta Was this translation helpful? Give feedback.
-
|
Yes, I only have one E2… I don’t have Linux, but I could install it on a virtual machine… or on WindowsAlso I could try with SysEx Librarian, it’s an app that sends SysEx from files and records the response. I was able to send “dump pattern” with it, but i would need a SysEx file. I can also try with Apple native API CoreAudio, I don’t have experience with it but it would be interesting to try. Still I’ll need to read your code and replicate it with Swift and form SysEx messages same way as you do (first time I’m dealing with this stuff)…Before I try the options above, are there any specific requirements when dumping PCM? I mean, maybe i should have removed the SE card before switching E2 on?Sent from a mobile deviceOn 2 Jan 2023, at 17:52, bangcorrupt ***@***.***> wrote:
The firmware you sent me matches the one I tested and when I run it I get the samples dumped correctly. The audio files you sent are the correct length but look like unitialised RAM, matching the data in the raw binary files. This is what I would expect to see if the script was run against the main Hacktribe branch, or if the read address is somehow getting mangled during transmission.
Can you confirm that you only have one Hacktribe connected via USB when you run the script?
Are you able to test using a Linux system? Maybe mido interacts with the MIDI drivers differently on a Mac.
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
That could be it, didn't think of it. If it loads an e2sample.all from the SD card it won't decompress the PCM from flash, so it will never copy it to CPU RAM. When it shows the sampler boot screen, do you see the loading bar move across the bottom? The first half moves in steps, this is reading the compressed PCM from flash. The second half moves smoothly, this is decompressing chunks and sending them to the DSP RAM. If it shows a percentage loading, then it is reading from the SD Card. |
Beta Was this translation helpful? Give feedback.
-
|
I’m not at home, will check when I go back, but as far as I remember indeed there was a progress indicator loading, with percentages Sent from a mobile deviceOn 2 Jan 2023, at 18:16, bangcorrupt ***@***.***> wrote:
remove the SD card before switching E2 on
That could be it, didn't think of it. If it loads an e2sample.all from the SD card it won't decompress the PCM from flash, so it will never copy it to CPU RAM. When it shows the sampler boot screen, do you see the loading bar move across the bottom? The first half moves in steps, this is reading the compressed PCM from flash. The second half moves smoothly, this is decompressing chunks and sending them to the DSP RAM. If it shows a percentage loafing, the it is reading from the SD Card.
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Andrew, is it possible you can share your wav file with the rest of us? Thanks a lot, glad you got it working! |
Beta Was this translation helpful? Give feedback.
-
|
I'm sorry, I don't really want to get into copyright troubles, that's why I didn't ask @bangcorrupt to just send me his extracted file. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks! It was more of a lazy request trying to avoid risk bricking my electribe xD I didn't consider the copyright issue, though. Happy new year! |
Beta Was this translation helpful? Give feedback.
-
|
DId you ever make a step by step guide for this? I'm having issues extracting the synth PCM file |
Beta Was this translation helpful? Give feedback.
-
|
I've tried for two days to get this to work and no matter what I get the soft-bricked screen that's says SOFTWARE UPDATE and nothing else. someone email me the the PCM file please. |
Beta Was this translation helpful? Give feedback.
-
|
Not sure if this us relevant or of any use... This is a software for making multisample samples for other earlier korg devices, wonder if its the same? |
Beta Was this translation helpful? Give feedback.
-
would it be possible at some point to acces ROM waveforms on E2 synth ?
at least drums/percussions in E2 synth are perfectly usable.
also, E2 synth owners have no option to export factory samples, so after Hacktribe installation we get only VA/FM oscillators, and that's all. we have to ask for factory waveforms (from E2S) elsewhere if we need them.
Beta Was this translation helpful? Give feedback.
All reactions