marcnet

To bypass sonicstage and directly convert from tracks on the Hi-md unit -> .wav I would need: a) Knowledge on Sonys encryption system and how to extract the encryption codes and decrypt the music data myself (which I have no idea about - encryption systems are not really in my programming experiance) or The ability to enumare COM interfaces on Sony's DLLs (which I can do) and to use Sonys custom COM interfaces on Sony's COM objects to manipulate the hi-md unit to upload data without destroying the original (which is a lot more trickey) or c) Same as ( except only extract the encryption code for an individual track. The data in the large file in the hi-md folder on the device is in a simular format to that of .oma files but without the encryption/decryption codes. As for larger block sizes - im also working on this. Currently it seems that the larger the block size I use the more repeated audio appears at the end of each block and is not detected correctly. At the moment the default is 10 seconds, not 5. And im also working on AN EXTREAMLY SIMPLE GUI system. Basically if you dont supply any command options (ie: you simply run the program from explorer) then a "Open file" dialog will appear, followed by a "save file" dialog. Log files will be generated by default in this mode. All other options will be left as default too. All these features to look forward to in version 0.03 :smile: Oh, and thanks for the donation A440. :smile: I look forward to buying that new mansion now :rasp: And I didnt know that Sonicstage trashes audio uploads sometimes. Thats just rubbish programming from Sony. Ill edit the first post and copy & paste your warning

It might be easier for us to diagnose and help with your problem if you add the -d option when rendering to .wav. This option produces a log file. If you open the resultant log file in notepad and copy & paste the last 10-15 lines of that log file we might be able to be a little more helpful. :smile:

To MZ-NH1: You dont need the .cpp and .dsp file for anything unless you're a programmer :smile: . These files are the sourcecode and the Microsoft Visual Studio project file respectivly. So If you want to change,add, remove features from himdrenderer, or include a GUI then you'll need these files. If not then you don't :smile: Ive released version 0.02 which fixes a few bugs, including (hopefully) that one. If you are using version 0.02 (it says so at the top of the help text)then it looks like its back to the drawing board again. Ive updated my initial post on this thread to link to version 0.02: http://www.marcnetsystem.co.uk/himdrenderer002.zip As for the repeated audio at the end of files, im still in the process of fixing this. And a free audio editing program: http://audacity.sourceforge.net/

Yes I will add a gui when ive run out of functionality to add No I will not make it a commercial project since I earn enough money at my job (Guess what, im a software programmer) to eat, sleep, buy CD's or whatever. However, I woudnt mind the odd donation of cash via PayPal :grin: I could add OMA->mp3 conversion. I could add OMA -> OGG converion. I could add OMA -> AAC conversion. But all these are low priority and I might pick the wrong encoder, resulting in everyone just rendering to .wav anyway and using another seperate encoder.

Ive fixed some things and release Version 0.02: * OMA end of file detection has been fixed. * Fixed incorrect WAV file size being written to WAV header * Added -f option, or "max allowed difference between samples". Basically the repeated bits of audio are not bit-for-bit exact. This parameter tells the program how tollerant to be when looking for repeated audio. The greater the value the more likely non-repeated audio will also get the chop. The smaller the value the more likely there will be repeated bits of audio left in the final .wav. This parameter is optional so you can leave it if you have no idea what im talking about * Improoved calculation of where the program predicts where the next bit of repeated audio will be. If this dosnt work then specifying option -c will go back to the 0.01 dumber method. * The Block size paramter is slightly less experimental due to the above changes. Try specifying a value of 20 or event 30 seconds. I have, however, left the default at 10 seconds. * Less crashes in parameter reading and wave rendering in general. * Ive included .cpp and .dsp source files in the zip. http://www.marcnetsystem.co.uk/himd/

If you feel that donations of cash are worthy - go to http://www.marcnetsystem.co.uk/ and click the paypal donation thing.

Freedom as in "Sony are rubbish and appy DRM on people personal recordings" rather than freedom of "MarC's program is sooo, like, retrictive." I see now. Thanks for clearing that up. Made me feel better, anway :smile:

hmm... accuracy I give you... My method is not 100% accurate due to the rather hackey cumbersome way it uses DirectShow... but "freedom"? .. Could you explain more how my method is restrictive in this department? You never know, if its something obvious then I could improve on it :smile:

Now this IS an odd error. Try adding the -k switch to keep the temporary file. How big is the temporary file? If its zero then I can only suggest to reinstall the latest SonicStage 2.1 ------ In regards to mp3 encoders. LAME is the *only* one you should use. The rest are complete crap, in my opinion. Download: http://lame.bakerweb.biz/ A good front-end to LAME: http://www.dors.de/razorlame/index.php Main LAME page: http://lame.sourceforge.net/ ----- And ive edited my instructions to *completly* remove the word DOS, and ive also changed the gender description of linzq :smile:

When i meant DOS i didnt actually mean "quit out of windows into DOS". In fact running the program without windows running wont work at all as my program heavily uses DirectShow - which you cant get access to under DOS. Windows has a command-line fuctionality which emulates the behaviour of DOS. It is this command-line functionality that I intended people to use, and not pure raw DOS itself. However, I have changed my post to not contain the word DOS, but instead use the phrase "command line" instead. Hope this makes it clearer

Admin edit by kurisu: Whilst this method is still tried and true, and allows more flexibility with the conversion process [flac, ogg, etc], I will note that Sonicstage 3 now has a straightforward *.wav converter tool for Hi-MD. HOW TO UPLOAD YOUR RECORDED TRACKS FROM HI-MD AND CONVERT THEM TO .WAV FILES *FAST*. This method will only work with tracks recorded onto HI-MD formatted media. Tracks recorded in NetMD or standard MD mode cannot be uploaded using this method. Further more, if a oma/omg track will not play in SonicStage then it cannot be converted using my program. ------ Oh and... This method is not guarenteed to work 100% all the time. This method is still in APLHA stage - meaning that it is still experimental. I take no responsability for transfer, corruption, loss or addition of legal or illegal data. If it goes wrong then Im not taking the slack. Neither am I reponsable if you do naughty illegal things usingthis method. You have been warned. The chance still exists that SonicStage and your HiMD will trash tracks randomly during upload. If you are using -only- himdrenderer to convert recordings after uploading them, then you are taking the risk of permanently trashing a first-generation recording without having a backup of any kind. AND: Feel free to mirror my program. Just give me and my website credit for my hard work :smile: And finally just to repeat what I said in another thread Its hi-speed yes, but it may miss out or repeat bits of audio... I hope not, but just dont go relying that this will produce a 100% accurate .wav version of your .oma files. ---------------------------------------------- 1) You will need: * The latest version of Sonicstage [currently 3.3] installed and working. * A copy of my HIMDRENDERER program version 0.52. Download version 0.50 here. You will need the DLL files from this version: http://www.marcnetsystem.co.uk/himdrenderer050.zip Download version 0.52 here. This version fixes a few bugs that are present in version 0.50. Simply replace the HIMDRendererWin.exe with the one from the version 0.52 zip file. http://www.marcnetsystem.co.uk/himdrenderer052.zip ---------------------------------------------- 2) Select Input file. There are 2 ways of doing this: (You will only need only one of (a) or ( b ) ) a) Browse your computer for the input file. First press the browse button at the top right-hand corner of the HIMDRenderer application (as show in the image above). A window should then appear that allows you to browse your computer for Sony OMA/OMG files. Browse to the file you wish to convert, select it, and press the "Open" button at the bottom right of the file browsing window: Once selected and opened, the file and path should appear in the "Input File:" box. --- OR: --- b ) Find the input file via the builtin "Find..." button At the bottom left corner of the HIMDRender window is a "Find..." button. Simply select the path on your computer to search (the default top option is where SonicStage saves its audio files) and click "Find...". After a few seconds the search will be complete and all results will be displayed on the left hand side of HIMDRenderer: You can then simply select an audio file from the list of results and click "Use Selected" to fill in the contents of the "input file:" field automatically ---------------------------------------------- 3) Select the output file. HIMDRenderer will automatically choose a filename and path to use as the output file. This will be identical to the input file filename and path but with the appropriate 3 letter extension (.wav for PCM wav files, .mp3 for mp3, etc) appended. If this is not what is desired then simply click on the "Browse" button next to the "output file:" field (circled in red on the image below) and choose a location and filename of where HIMDRenderer should put the output file. ---------------------------------------------- 4) Options. There are 3 option sections: Processing options, Advanced options and Output options. Click on the apropriate option button to show options for that section. Options in the first two sections only need to be changed if the output file contains errors such as repeated or missing audio. The "output options" change the file format of the output file. If the "Dual channel mono" option is enabled then HIMDRenderer will pick either the left or right audio channel, depending which is louder, and duplicate that channels data on the quieter channel. The effect is a mono sounding output file. This is useful for recordings that only have audio on the left or right channel. If the "cut start and end silence" option is enabled then HIMDRenderer will attempt to cut silence at the start and end of each file. This is still experimental and it is recommended to leave this option off. The "Output Encoding" option selects whether to save the output file as a PCM .wav (lossless), a MP3 lossy compressed file, a FLAC lossless compressed file or an OGG Vorbis (.ogg) lossy compressed file. Lossy compression discards audio information that the compresser thinks you cannot hear. Lossless formats keep all audio data. The options following "Output Encoding" depend on the selected ouput format: PCM No options available for PCM .WAV files MP3 Stereo Mode: This selects wheter to use Full stereo (Best channel seperation), Joint Stereo (Good quality, better compression) or Dual Channel (left & right channels represent different audio streams) Bitrate control: Variable bitrate (2 methods ... the newer method is faster than old, but slightly worse quality), Average bitrate or constant bitrate (standard mp3) With variable bitrate you can select the minimum and maximum bitrates allowed. The "VBR quality" setting balences file size against audio quality. The lower this value the better quality but larger file size. The higher the smaller the file size but worse quality. Average bitrate and constant bitrate only allow a single bitrate to be chosen. The "Quality" (last option) option balences encoding speed with audio quality. The higher this value the faster the encode but the worse the sound quality. FLAC To be honsest, I have no idea what these options do. Im not really a FLAC expert... in fact ive never used it. Sorry. Can anyone help me here? OGG VBR Mode: Here you can select to encode by quality or bitrate. You can also choose to use an averageing bitrate method. With quality mode, the lower the value the worse the quality of the audio and the smaller the file size. OMA Just copies the OMA file. No conversion is done when using this output method. ---------------------------------------------- 5) Reneder. Once everything is set up press the "Go Render!" button at the bottom right of HIMDRenderer. Conversion should take a short while (A percentage indicator is given inside the "Go Render!" button) ---------------------------------------------- 6) Batch Mode If you want to convert multiple files in one go then press the "Batch Mode" bottom at the bottom of HIMDRenderer. The "Batch Mode" button will turn into a "File Mode" button. Clicking on "File Mode" returns to the initial single file state. In batch mode a file list (batch list) will be displayed on the left of HIMDRenderer. You can either add files using the "Find..." method. This will search for files as described in section ( 2b ). Selecting a file and clicking "Use Selected" will add the selected file to the batch list. Holding down the CTRL or SHIFT buttons and clicking items in the search result list allows multiple items to be selected. Clicking "Use Selected" will add all selected files to the batch list. You can also click "Add Files" to display a file selection window simular to the one in section (1a). The batch mode version of this file selection window allows multiple files to be selected by holding down the CTRL or SHIFT buttons and clicking the required files. The "Remove Files" button removes items from the batch list. Multiple items can be selected for removal via the CTRL OR SHIFT methods. Select a directory to store the converted files by clicking on the "Browse..." button next to the "Output directory:" label. Press the "Go Render!" button at the bottom right of HIMDRenderer to convert all files in the batch list. ---------------------------------------------- 7) Bugs and updates (NOTE: These features require an active internet connection) If you feel you are not running the latest version of HIMDReneder then press the "Latest version" button at the bottom right of HIMDRenderer. A message box will appear detailing the very latest available version of HIMDRenderer as well as where to get it. If HIMDRenderer could not process one of your files, or the output of a conversion produced audible errors then please do the following: * Click "Bug Report" at the bottom right of HIMDRenderer. You can go back to normal HIMDRenderer operation via the "Exit Report" at the bottom of HIMDRenderer. * Fill in your email address (I wont spam you, I just may want to contact you for more information) and any additional information you can think of about the error (Sonicstage version, Windows Version, Type of error,etc ....) *Click "Send Report" at the bottom right of HIMDRenderer. The log file(s) generated by HIMDRenderer will be sent automatically ------------------- Notes: Thats it. Hope I havn't missed anything. HIMDRenderer is freeware. However, if you feel my program is worth donating to then I accept donations via NoChex and Paypal. Go to to dontate Ive set up my own message board at: http://www.marcnetsystem.co.uk/himd/. You can always find the latest version of HIMDRenderer there. And if you want to chat then I can be found floating around the Undernet IRC network in the rooms #chatterz or #marcchat. While im here spamming my website, be sure to check out my music that I created at: http://www.marcnetsystem.co.uk/music/

TO: papagoose and linzq: Basically the OMG transform filter does not like receiving connections from the OMG source filter (or the other way around) and is crashing. What versions of Sonicstage do you use? Try re-installing the lastest Sonicstage (2.1), including the OpenMG secure modules.

http://forums.minidisc.org/viewtopic.php?t...t=6019&start=60 Ive released a first alpha version of my .oma -> .wav file converter. Give it a go. Just remember that it might do unexpected thing ....

They've probably already done the core part of their .wav converter, its just not been released yet because 1) Sony are still designing a bulky ugly GUI frontend, just like they did with OpenMG, SonicStage and Simple Burner 2) Sony are still trying to decide if allowing to convert our OWN audio files from .oma to .wav is going to, in some way, effect their record sales.

Well I did say it was an alpha... Anyway... It sounds like the oma-> temp.wav is not detecting the end of the file for some reason and is repeatly re-rendering the last 15 seconds. You might want to check this Ill have a look tonight, after work.

Linqz - I kept getting this error when I was debugging the program with MSDEV. Basically sony's DLL's have protection in them to stop any programs loading that have debugging capability. But I doubt this is the problem (unless you got Softice or anyother debugger than runs behind Windows running) Try using the GraphEdit method described on page 1 of this thread and see if that crashes.

What you need to do is: Copy the himdrenderer program to the root of your c: drive. Go to your windows start menu and select "Run" In the "Open" text box type: c:himdrenderer -i <path to .oma file> Sorry there's no gui or simple key-press interface, but I develop all my programs, initilally, with a command line only style.

Im kinda nervous about this as its a very first release and eveything: http://marcnetsystem.co.uk/himd/ But there it is... a very alpha version of the program. Its hi-speed yes, but it may miss out or repeat bits of audio... I hope not, but just dont go relying that this will produce a 100% accurate .wav version of your .oma files. In other words: DONT DELETE YOUR .OMA FILES, EVEN IF YOU ARE REALLY SURE THAT MY PROGRAM IS ACTUALLY REALLY PROBABLY SORT OF WORKING Feel free to post replies on my message board, or on this one. If you want to chat then I can be found floating around the Undernet IRC server in the rooms #chatterz or #marcchat Enjoy... The battle may be won but yet there still may be casualties

Ive got a full capture program which now cuts out repeated bits of audio. I just got to polish off now and ill probably release it tomorrow. Its probably going to be kept as a command line app for now - so get out your "How to use DOS" books MDFreak was right about repeated audio not being bit for bit exact.... and ive also taken this into account. Now if the 16 bit values being compared vary by 2, then they are still considered the same... I may tweak this value depending on how my "polishing and testing" goes. Tomorrow will either be a barrier broken for the HI-MD folk, or it will make me look a bit silly if it falls over and dies.

Ive so far written a program to extract the whole .wav using 5 second chunks. Ive done this by telling the filter graph to play from 0-5, 5-10, 10-15, 15-20 etc However, the seeking granularity of the IMediaPosition interface is rubbish. It works in seconds (not milliseconds or anything else useful) ----- Result so far: Either there are bits of PCM missing in the .wav (if I seek to where I told the filter graph to stop) or there are repeated bits (if I seek to 1 second before where I told the filter graph to stop). ----- Next: Im going to find some way else of controlling the filter graph (I think ive actually found something with something like 100 nanoseconds of granularity) Or if that dosnt work: Im going to implement a second pass that cuts out repeated bits of audio. I know where they are and how often they occur - 5 seconds. I just dont know the length of the "repeat". And yes, I know 5 seconds is a bit short and I could use 1 minute instead - but the inaccuracy of the IMediaPosition will still produce errors even with 1 minute blocks. Its just easier to see/repair/test for errors if Im doing 5 second blocks rather than 1 minute blocks. But im getting there: Onwards we go further..... (with crappy resuming accuracy and without debugging capability :/)

Hello. I wish to report myself as being a bit of a moron.... Way back when I was hacking net-md dll's (in particular the Netmdapi.dll - I didnt get very far as all data was passed to this dll in encrypted form) I tried attaching a debugger to the OpenMG software. Strangly, It crashed. I investaged furthur and found the one of NetMD modules was dynamically loading (LoadLibrary()) one of the OS dlls and calling the IsDebuggerPresent() function. Guess how Im running my filter-graph software. Thats right - via msdev debug & run feature.... doh! Thats why I get exceptions when I create a connection between Filters. Onwards we go..... (without debugging capability :/)

Ive tried both methods in this order 1) I created the filter (855FBD04-8AD5-40B2-AA34-A6581E59831C), queried it for the IID_FileSourceFilter interface, used that interface to load the .oma file. I then enumerated the pins for an output pin and tried to connect it to the input pin of the transform filter. Crash. 2) I then tried the AddSourceFilter function of the graph builder object as I desribed before Here is a link to the code ive got so far. Its a bit messy... http://marcc.gotadsl.co.uk/files/main.cpp

I have a different theory: When the OMG source filter reads the last byte of OMA data, it tells the filter graph (or whatever) that it has finished spitting out data and that the filter graph should stop. Which it does. Meanwhile, the file writer fitler is still writing wav data. The filter stops (as told by the OMG source filter). Wav writing stops. Im also having c/c++ coding issues: I create a GraphBuilder object via CoCreateInstance I use the GraphBuilder function AddSourceFilter to add a ".oma" file to the graph. This function succeeds. I enumerate the pins on the source filter (as returned by AddSourceFilter ) and get an output one I tell the graph to render the pin via the Render funtion. An exception happens.... An exception also happens If I manually connect the output pin to an input pin of an OMG TRANSFORM filter that has also been created (CoCreateInstance) and added to the graph (AddFilter) . Anyone got any ideas?

I tried TMPGEnc as soon as I saw this post... I used the standard 5 minute track I used with GraphEdit to try to prove everyone wrong that the 10.3meg limit isnt there ... Which I failed doing . as the 10.3mb limit IS there. Anyway.. TMPGEnc as the same sort of limit. It produces a larger .wav file (i think its the correct size, actually) but after about 1 minute in the .wav file it is all BLANK. Development continues... the search for the ultimate .oma -> .wav continues to elude us all....

Ill have a go tonight ... My DirectShow knowledge is a bit rusty so it might take the weekend for me to get something done and in a releasable state. Last night I did get as far as creating the OMG source filter object and giving it a file... All functions (methods, procedures..... whatever) returned S_OK (which means "success") , so I think im getting somewhere......