HydrogenAudio

CD-R and Audio Hardware => CUETools => Topic started by: Rollin on 2018-10-16 15:34:00

Title: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Rollin on 2018-10-16 15:34:00
In CUETools 2.1.7 when i try to encode something to FLAC using libFLAC encoder this results in error "Arithmetic operation resulted in an overflow.."
Core i3 3245, Windows 7 SP1 32bit, NET Framework 4.6.1, MSVC 2015
Title: Re: CUETools 2.1.7 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Gregory S. Chudov on 2018-10-16 16:07:35
Thanks, i'll look into it.
Title: Re: CUETools 2.1.7 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2021-03-10 22:38:43
Getting this with v.2.1.8 on Win 10 x64 and Win 7 x86.

Went back to v.2.1.6 and all ok again.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: c72578 on 2021-03-11 20:57:27
I have tried to reproduce this issue by encoding dozens of albums using CUETools 2.1.8 and libFLAC using default settings.
So far not hit. The error "Arithmetic operation resulted in an overflow" has not appeared yet.
It would be good to have an album, where this error occurs, for debugging.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2021-03-12 00:39:09
In my case, it was Willie Nelson's Spirit.  I was using libFLAC at compression 6.

Is there a way to erase all settings back to default?
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: c72578 on 2021-03-12 05:49:32
Thanks for the info about an album, where the error occurs.
The default compression setting for libFLAC is "5".

The settings file is typically stored here:
"%APPDATA%\CUE Tools\settings.txt"
Which is "C:\Users\%USERNAME%\AppData\Roaming\CUE Tools\settings.txt"
If you backup and remove this settings.txt file, the default settings are restored after the next start of CUETools.exe.

Another possibility is to rename the file "user_profiles_enabled" to e.g. "user_profiles_disabled".
After the next start of CUETools.exe, a fresh default settings file is generated in a subdirectory, where the CUETools zip file has been unpacked:
"CUE Tools\settings.txt"
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2021-03-12 14:11:24
Thanks.  After deleting the %APPDATA% folders, v.2.1.8 rips without error.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2021-03-25 13:10:48
Sorry to say that I still see this error creating image & cue in CUERipper v.2.1.8 when using libFLAC compression 5 burst.  Switching to cuetools solves it.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: c72578 on 2021-03-25 17:54:38
Thanks for the additional information.
Does this mean, that the error only occurs on your system, when Burst mode is used and that it is OK in Secure mode?
Two more questions:
- With which drive model does this error occur?
- Does the error occur with every album or just single ones?
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2021-03-25 20:06:55
It's an "HL-DT-ST DVDRAM GH24NS50 ATA Device".  It's happened on 3 out of 3 CDs I've tried, after gap detection at about 2% completion.  Error on Burst, Secure, and Paranoid.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: c72578 on 2021-04-02 20:58:56
@Apesbrain In the meantime I can reproduce the issue "Exception: Arithmetic operation resulted in an overflow.", when encoding using libFLAC. It happens, when CUERipper is running in 32-bit mode. Could you please check on your Win10 64-bit system, if Task Manager shows "CUERipper" or "CUERipper (32-bit)", after CUERipper has been started?

X

Normally, CUERipper would run in 64-bit mode on a Win10 64-bit system. For testing, I could force CUERipper to run in 32-bit mode using
Code: [Select]
CorFlags.exe CUERipper.exe /32BITREQ+
from the Visual Studio Developer Command Prompt and reproduce the issue.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2021-04-03 00:18:58
Could you please check on your Win10 64-bit system, if Task Manager shows "CUERipper" or "CUERipper (32-bit)", after CUERipper has been started?
In Task Manager, it says "CUERipper" on my Win 10 x64 system.  These latest errors, however, were on my Win 7 x86 server.  I'll test both again.

EDIT: Confirmed, the issue seems only to be on the x86 system.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: c72578 on 2023-02-02 18:50:20
A fix has been prepared and a build of CUETools, which allows encoding using libFLAC also under 32-bit OS can be found here:
https://github.com/gchudov/cuetools.net/pull/251#issuecomment-1414203173

Feel free to test.
Title: Re: CUETools 2.1.7 & 2.1.8 libFLAC error "Arithmetic operation resulted in an overflow"
Post by: Apesbrain on 2023-02-03 00:36:01
Feel free to test.
Sorry, I no longer have any Windows x86 PCs.