Mod Archive Forums

Music Production => NitroTracker => Tracking => NitroTracker Bug Reports => Topic started by: darkbluedrew on July 13, 2008, 05:19:41

Title: 0.4 corrupting samples?
Post by: darkbluedrew on July 13, 2008, 05:19:41
Been using 0.3 for a while, and had not come across this problem until switching to 0.4. It seems that editing samples, then saving the song, corrupts the sample that is embedded in the .xm file. All the tracking info stays, but some of the samples turn to white noise and/or buzzing sound. Using 0.4 on Supercard MiniSD on DS Lite.

To replicate, open notominous-a19.xm, click the third sample (ahh.wav). Go into sample editor and delete the last bit of the sample on the right (sample will still sound correct at this point). Save the song as a different file name, and reopen. I get 1 as white noise, 2 is normal, 3 is a harmonic buzz, 4 is normal, 5 is white noise, 6 is normal, and 7 is white noise. I did it with this file 3 times, and sometimes the instruments are white noise, sometimes they are more harmonic, sometimes the first half is buzz and the second half is kind of buzzy but sounds like the original sample.

I chkdsk'd my sd card to see it it was corrupted, but found nothing.
Title: Re: 0.4 corrupting samples?
Post by: usrfriendly on July 13, 2008, 10:40:18
it might be a write issue with nitrotracker on your card.  apparently, you didn't upload the file, so I cant check for you.  Did you test it on another program (a tracker or mod player)?  have you had bad saves with any other files/programs on your supercard?  and is this the only song it happens to?  test the last one by making a song using the same samples.
Title: Re: 0.4 corrupting samples?
Post by: 0xtob on July 13, 2008, 13:50:06
Now that's a good bug report with detailed reproduction steps! The problem looks like there's a bug in the DLDI (no handling of unaligned writes). You could try the Supercard DLDI patches instead:

http://dldi.drunkencoders.com/index.php?title=SuperCard_(SD_Card) (http://dldi.drunkencoders.com/index.php?title=SuperCard_(SD_Card))
Title: Re: 0.4 corrupting samples?
Post by: darkbluedrew on July 13, 2008, 14:56:13
0xtob - I had originally patched the file with scsd.dldi, but I redownloaded the file and patched again (used the unpatched file out of the rar each time). Still does the same thing. I also tried scsd_moon.dldi, but it also did the same thing.

pogue - I wrote reproducing instructions with an .xm that came with the software so that no one had to download anything. This is not the only xm it does this with, it happens with any xm that I edit the samples - even recorded samples. I've never had a bad save or file on this card with any other software (but nitrotracker is the only reason I will open my ds lately). One time, nitrotracker 0.3 corrupted one of my .xm's, but if I remember correctly, it was my fault because I pulled out the SD card when it was running or something.

[edit] I also redownloaded Dlditool-win32-gui just to make sure it wasn't the patcher. It wasn't.
Title: Re: 0.4 corrupting samples?
Post by: 0xtob on July 13, 2008, 23:39:43
Hmm, this is odd. Have you tried sclt.dldi as well?
Title: Re: 0.4 corrupting samples?
Post by: darkbluedrew on July 14, 2008, 00:03:22
I don't have an SC Lite, but I tried the patch anyway and it didn't work. Attached is an example of the problem per the method I originally posted.
Title: Re: 0.4 corrupting samples?
Post by: 0xtob on July 14, 2008, 02:29:37
Sorry, I misread MiniSD for MicroSD and thought it was an SC Lite. The file you posted looks exactly like the alignment problem. Does the problem occur only when you change samples or do you also get noisy samples when you just load a song and save it again under a different name?
Title: Re: 0.4 corrupting samples?
Post by: darkbluedrew on July 14, 2008, 05:31:20
Only when I edit the samples - I made some tracks with looped sine waves and they saved fine. This was not a problem in 0.3 as I successfully recorded and edited a bunch of samples and they saved ok (same sd card/supercard/ds).
Title: Re: 0.4 corrupting samples?
Post by: People on July 14, 2008, 10:40:40
I experience the same problems as darkbluedrew with recordings from the mic and loaded samples. After saving whatever I do in "NT 0.4" to an ".xm" file, I get discouraged because everytime I load anything created in "NT 0.4" either some recordings or samples or both get messed up. You never fully hear and get back what you previously saved. I never had this problem with "NT 0.3". I am using "NT 0.4" on the NAND memory on AK RPG.
Title: Re: 0.4 corrupting samples?
Post by: darkbluedrew on August 03, 2008, 14:22:25
After much trial and error, I figured out that by removing "dldi.scp" from the root of my sc card, I am able to use the full potential of NT! I am very impressed by this version. Thanks!
Title: Re: 0.4 corrupting samples?
Post by: 0xtob on August 03, 2008, 23:04:17
darkbluedrew: Glad you got it to work! It seems like your card was auto-patching NitroTracker with a broken DLDI, and deleting it disabled auto-patching, so the correct DLDI that you patches NitroTracker with was used.

People: You could try disabling auto-DLDI-patching and using the Acekard+ DLDI patch (http://dldi.drunkencoders.com/index.php?title=AceKard%2B). Or Maybe writing only works if run from SD Card instead of the internal NAND.
Title: Re: 0.4 corrupting samples?
Post by: remouk on August 04, 2008, 16:12:15
Just to let you know, I have the same problem here.

- DS Linker with internal flash memory
- NitroTracker patched with the "NEO Flash MK5 (NAND Flash)" DLDI file, as I usually do
Title: Re: 0.4 corrupting samples?
Post by: 0xtob on August 16, 2008, 02:56:17
OK, the thing is: In the old days, chishm's FAT driver couldn't do non-aligned writes. They would result in noise samples (the same that some of you are experiencing). So, I added a buffer to my XM writing routine such that all writes would be aligned and samples would save fine. With the advent of DLDI, I thought the alignment problems were solved and I removed the buffer because it cluttered up the code. I tested XM writing on my hardware which works fine.

Judging from this thread it seems that some DLDI drivers are buggy in that they do not support unaligned writes. If you have this problem, please complain to your card manufacturer or whoever wrote the DLDI driver and have them fix it, or even better, fix it yourself if the DLDI source is available and make the DLDI available.

But since I don't think any flash card manufacturer will actually care, I'll add the wite buffer to the XM saving routine again so you can save uncorrupted XMs regardless of whether or not your driver can write unaligned. I have no time for NT coding at the moment, and I cannot promise a date for this fix, but you won't have to wait until 0.5 :)
Title: Re: 0.4 corrupting samples?
Post by: remouk on August 27, 2008, 18:40:11
Thanks for your detailed answer! Can't wait for the new version(s).
Title: Re: 0.4 corrupting samples?
Post by: eventide on October 13, 2008, 07:36:20
Hi there,

Can anybody tell me, which combination of nitrotracker and hardware module will work
without this problem? I want to buy a new module.

Thanks in advance!

Eventide.
Title: Re: 0.4 corrupting samples?
Post by: werto on October 17, 2008, 09:17:33
Hi there,

Can anybody tell me, which combination of nitrotracker and hardware module will work
without this problem? I want to buy a new module.

Thanks in advance!

Eventide.


Yo Eventide :)
here's a cool link all for you:

http://nitrotracker.tobw.net/index.php?cat_id=4#compatibility

wish you'll solve your problem :)

see ya'
peace and love
werto

ps:there's no DSTT in list, but i can say you that it works very fine with it (with the last two kernel :) ).
Title: Re: 0.4 corrupting samples?
Post by: john amino on March 25, 2009, 14:21:13
I have this problem....and i can't fix it. its driving me insane

does anyone have a version of 0.3 they could send me?
Title: Re: 0.4 corrupting samples?
Post by: john amino on March 25, 2009, 14:25:50
scrap that.....i need 0.4 working.

ARGH!

can anyone help? i was intending to use Nitrotracker in a live set in a few weeks

i'm using R4DS and DLDI Drop to do the patching.
Title: Re: 0.4 corrupting samples?
Post by: john amino on March 25, 2009, 14:51:14
...or post me up the details of your set up that work without this problem

cheers

xx
Title: Re: 0.4 corrupting samples?
Post by: john amino on March 26, 2009, 12:27:10
come on people!

someone must have at least a working set up they can share with me....you can't all be suffering from this bug??

just the card and the patch
Title: Re: 0.4 corrupting samples?
Post by: MatrixMaster3 on March 27, 2009, 01:29:14
Try contacting Azop Corp. I don't know what card he uses, but, he does use Nitro live. He might be able to help.
Title: Re: 0.4 corrupting samples?
Post by: john amino on March 27, 2009, 12:53:05
thanks, i'll do that
Title: Re: 0.4 corrupting samples?
Post by: zoikkis on April 07, 2010, 16:21:13
I have added this into the issue tracker: http://code.google.com/p/nitrotracker/issues/detail?id=15