John,
I tried to reproduce it with a very simple SoundFont, but I could not. It is obviously something that may require one more step which I did not perform. I will think about some kind of debug version for you that prints out more information. It will take some time.
Edited Instruments Splits in Viena, now all samples are broken
Re: Edited Instruments Splits in Viena, now all samples are broken
Here is version "C" of 0.996:
https://dl.dropboxusercontent.com/u/173 ... a0996C.rar
Do your best to make the fault appear. Run DBWIN32 and hand me the output.
I suggest you use "Save As..." to save the altered file.
https://dl.dropboxusercontent.com/u/173 ... a0996C.rar
Do your best to make the fault appear. Run DBWIN32 and hand me the output.
I suggest you use "Save As..." to save the altered file.
-
- Posts: 65
- Joined: Sat Mar 21, 2015 11:29 pm
Re: Edited Instruments Splits in Viena, now all samples are broken
Kenneth,
Here is the debug printout:
https://dl.dropboxusercontent.com/u/673 ... 0.996C.txt
However, the fault does not seem to have occurred this time.
I intend to keep using 0.996C and shall report if the fault reappears.
John Reimer
Here is the debug printout:
https://dl.dropboxusercontent.com/u/673 ... 0.996C.txt
However, the fault does not seem to have occurred this time.
I intend to keep using 0.996C and shall report if the fault reappears.
John Reimer
Re: Edited Instruments Splits in Viena, now all samples are broken
Hi Kenneth,
since quite a few versions (including the last one) it's impossible to add a sample from .WAV to replace an older one. Even though it warns about one existing with the same name, it won't replace it regardless of what I choose (nor does it respect the "keep root and tuning"). I unfortunately didn't record the debug output.
I just wrecked another soundfont and it might have been in relation to the issue I just mentioned.
since quite a few versions (including the last one) it's impossible to add a sample from .WAV to replace an older one. Even though it warns about one existing with the same name, it won't replace it regardless of what I choose (nor does it respect the "keep root and tuning"). I unfortunately didn't record the debug output.
I just wrecked another soundfont and it might have been in relation to the issue I just mentioned.
used to be "likeaguest"
-
- Posts: 65
- Joined: Sat Mar 21, 2015 11:29 pm
Re: Edited Instruments Splits in Viena, now all samples are broken
Hi Kenneth,
I believe that this time Viena 0.996C has reproduced the "Save after Purge" problem.
Here is the debug link:
https://dl.dropboxusercontent.com/u/673 ... Ctest2.txt
Regarding Mike's problem with replacing a sample with one of the same name, I have just tried to reproduce that with 0.996C, and I didn't get it to reproduce it. My experience with this is that Viena will accept the new sample, but will retain the old one as well, and retain the SAME name. This is a nuisance, as when I do a reassign, I'm not sure which version of the sample has been used. So I end up deleting both samples from the .sf2 file and adding the sample afresh.
I might add that I have been aware for quite some time, of a problem when trying to replace a sample with a new one of the same name. The problem goes away if I select (for viewing) some other sample from the sample list before trying to add the new version of the wanted sample. This workaround has become second-nature to me, which probably explains why I haven't mentioned it to you until now. Sorry.
John Reimer
I believe that this time Viena 0.996C has reproduced the "Save after Purge" problem.
Here is the debug link:
https://dl.dropboxusercontent.com/u/673 ... Ctest2.txt
Regarding Mike's problem with replacing a sample with one of the same name, I have just tried to reproduce that with 0.996C, and I didn't get it to reproduce it. My experience with this is that Viena will accept the new sample, but will retain the old one as well, and retain the SAME name. This is a nuisance, as when I do a reassign, I'm not sure which version of the sample has been used. So I end up deleting both samples from the .sf2 file and adding the sample afresh.
I might add that I have been aware for quite some time, of a problem when trying to replace a sample with a new one of the same name. The problem goes away if I select (for viewing) some other sample from the sample list before trying to add the new version of the wanted sample. This workaround has become second-nature to me, which probably explains why I haven't mentioned it to you until now. Sorry.
John Reimer
Re: Edited Instruments Splits in Viena, now all samples are broken
Thanks John, I'll analyze the output. In the mean time, if you can think what additional step(s) you had to take to arrive at this result, then please tell me.
I guess it might be a good idea to force the user to give the new same a unique name. The only problem is that the name cannot have more than 20 characters (silly limitation). Also, for stereo samples it is advised (but not required) that the name should end in the character "L" and "R" (or low case). This last latter is the only letter separating the stereo sample names from each other and if the name gets longer than 20 chars they may become dropped.
I guess it might be a good idea to force the user to give the new same a unique name. The only problem is that the name cannot have more than 20 characters (silly limitation). Also, for stereo samples it is advised (but not required) that the name should end in the character "L" and "R" (or low case). This last latter is the only letter separating the stereo sample names from each other and if the name gets longer than 20 chars they may become dropped.
-
- Posts: 65
- Joined: Sat Mar 21, 2015 11:29 pm
Re: Edited Instruments Splits in Viena, now all samples are broken
Kenneth,
Bad news. 0.996C has now corrupted the samples twice, on normal non-complicated addition of new samples. I haven't yet managed to get it to do this with the DBWIN32 active.
Fortunately the backup file came to my rescue.
John Reimer
Bad news. 0.996C has now corrupted the samples twice, on normal non-complicated addition of new samples. I haven't yet managed to get it to do this with the DBWIN32 active.
Fortunately the backup file came to my rescue.
John Reimer
Re: Edited Instruments Splits in Viena, now all samples are broken
John,
Looks like a very difficult issue to reproduce. How much I try it never happens to me. Ergo - I'm doing my tests the "wrong way." I will think of a way to document it when it occurs for you. There are ways.
Looks like a very difficult issue to reproduce. How much I try it never happens to me. Ergo - I'm doing my tests the "wrong way." I will think of a way to document it when it occurs for you. There are ways.
Re: Edited Instruments Splits in Viena, now all samples are broken
John,
Next time when it happens, please do your best to identify what has gone wrong. I mean, does it look like only one, a few, or all samples have been scrambled? Are the actual audio data destroyed or are the splits just linked to the wrong samples? What else can you think of?
Next time when it happens, please do your best to identify what has gone wrong. I mean, does it look like only one, a few, or all samples have been scrambled? Are the actual audio data destroyed or are the splits just linked to the wrong samples? What else can you think of?
-
- Posts: 65
- Joined: Sat Mar 21, 2015 11:29 pm
Re: Edited Instruments Splits in Viena, now all samples are broken
Kenneth,
What is frustrating is that it happens only very occasionally, with most additions causing no problem. What I do recall is that it is certainly not a case of matching up samples with the wrong splits. The actual samples are degraded, but I have not checked on how widespread the damage is..
This has not been a major problem to me, for I have learned to do some actions using Viena .078, which itself has a couple of eccentricities but causes no damage. But I acknowledge that this is no solution for you as the creator of Viena.
I have recently completed a major sample-making project, but there is another one just around the corner.
John Reimer
What is frustrating is that it happens only very occasionally, with most additions causing no problem. What I do recall is that it is certainly not a case of matching up samples with the wrong splits. The actual samples are degraded, but I have not checked on how widespread the damage is..
This has not been a major problem to me, for I have learned to do some actions using Viena .078, which itself has a couple of eccentricities but causes no damage. But I acknowledge that this is no solution for you as the creator of Viena.
I have recently completed a major sample-making project, but there is another one just around the corner.
John Reimer