Skip to main content
Recent Posts
1
FLAC / Re: Archive tester - Tool to facilitate testing collections of archives, incl. FLAC
Last post by Thomas C. -
Version 0.2.0.0 is available (big update).

Change log
Parallel testing (~multithreading)
  • Testing processes can now be ran in parallel (multithreaded). Default thread count is based on .NET's "Environment.ProcessorCount" which counts your computer's logical cores. There is a setting to change this manually too.
    Previous version was singlethreaded, relying on the fact the algorithms such as RAR5 are multithreaded. But some, for instance RAR4, are not multithreaded. This should bring a huge performance gain when testing a large number of files encoded with singlethreaded algorithms (or at least when testing algorithms are singlethreaded).
Multivolume: I still have a more flexible file extension/tester matching system in mind, but here is already something that should solve many cases.
  • *.001 are now testable through 7z.exe
  • *.r00 and *.r01 are now testable through UnRAR.exe, *.r01 is considered non testable if matching *.z00 exists in the same folder.
  • *.z00 and *.z01 are now testable through 7z.exe, *.z01 is considered non testable if matching *.z00 exists in the same folder.
  • Prevented case when all files ending with *.rar but belonging to the same multivolume archive would get tested along with each other. Which resulted in n2-1 redundant tests (if I got the formula right). Files ending with ".part1.rar" are still considered testable while all other volume, following pattern ".part[N].rar" are now considered non-testable.
  • ?: Should multivolume non-first volumes be counted as non-testable or simply ignored? For now they are counted as non-testable.
Self-extracting I still have a more flexible file extension/tester matching system in mind, but here is already something that should solve many cases.
  • *.exe are now optionally testable. *.exe first get tested with UnRAR.exe, if not a RAR archive, a *.exe file will then be tested by 7z.exe, if failed, it will be counted as failed test. There is a checkbox to test or ignore *.exe files. Default value: false.
Benchmarking:
  • Added "kB/s" indicator.
Targeting tuning:
  • In the "Inputs: Paths to be tested" field, you may now include/exclude specific files, the same way you would include/exclude a folder.
  • There is now a setting to automatically skip some folders by name (separated with "|"). Default value: System Volume Information | $RECYCLE.BIN. (In previous versions, "$RECYCLE.BIN" was hard-coded.)
  • Prevented cases when some file could be tester multiple times or counted multiple times as non testable files (typically if a folder and its parent folder were both set as inputs).
Stability and efficiency
  • Closing main form/application now more efficiently closes background threads.
  • Stopping during discovery process will now be much faster (kind of instant).
UI:
  • At first start, the input text field is now filled with a demo/tutorial, that should be able to scan you all computer if started.
  • Tests that have been stopped by clicking on "Stop" now count as skipped.
  • Possible future improvement: Progress indicators work fine is the full process is ran. Using "Skip" or "Stop" buttons result in skipping files, which may make progress indicators interpretation arguable. This does not look like a priority but may possibly be re-thought in the future. Probably in a smaller update.
UI consistency:
  • "Skip" button is now disabled during the discovery process and only activated during testing process.
  • Progress indicators that can only be evaluated after at least one file has been tested (such as "Est. total time") now read "?" before they can be evaluated, instead of "00:00:00.0000000".
  • "Launch/Testing" button now reads "Scanning" during the scanning process (instead of "Testing").
  • "Time spent on current file" is replaced with the count of active parallel tests when parallel testing.
  • .exe path text boxes and most setting buttons are now disabled while processing.
Dependencies and compatibility:
  • Tested with UnRAR.exe 5.60 and 7-Zip 18.05.
  • Update UnRAR.exe to version 5.60
Possible issue(s):
  • On my Win7 computer, stopping the testing process ("Stop" button") gets the application stuck with one running parallel test running (according to UI). This did not happen on the Win10 computer where I write the code. Cause unidentified so far.
3
Support - (fb2k) / Replaygain changes have instant effect - EXCEPT when using shortcut keys. Bug?
Last post by MaxDread -
Hi all

As far as I know, in the past - a month or two ago and previous to that - if you changed the replaygain setting (none, track, album) "on the fly" it had no effect  It would only take effect if the track was stopped and re-started, or when the next track came on 

So it was a pleasant surprise today when by accident I found that changing the RG setting does now make the change "on the fly" / instantly.  However, it only does this when using the drop down menu.  When I use shortcut keys to change the RG status it does not happen.  I tend to always use my shortcut keys, and I guess that's why I had not noticed the change. 

Whether it is a bug or not I don't know.  But is it something that can be fixed? 

Ta
5
Support - (fb2k) / Re: foobar2000 v1.4b19 - visualizations on toolbar don't have native borders anymore
Last post by j7n -
The screenshots posted by Rollin appear to be resampled, reducing the visibility of the effect, and adding new blended borders. But it is still there. The minor graphical flaw resulting from this flatness is that the bars are poorly defined where the gradient happens to randomly match the color/brightness of the window chrome - the first 25 percent of the level meter. A border is additional protection against this, why the transport icons look as they do, being white on white.

8
Scientific Discussion / Re: Help me understand why sound is one dimensional
Last post by Phanton_13 -
All this problem lies in the definition of dimension, as it varies with the context from the classical 3 spatial dimensions + time up to the multiple dimension in quantum physics. Sometimes you can argue the existence of hidden or omitted dimensions.

In the case you can argue that audio is 2 dimensional because it needs 2 elements to exist, magnitude (pressure) and time, after all audio is the information conveyed by the difference in pressure over time, even if its commonly represented as 1d vector because of its simplicity in this representation it exist a hidden dimension that is time because each position in the vector represents concrete time. It's possible to represent each point of audio as list of [pressure, time], this representation can led to variable sample rate audio, something exotic but plausible.

Also I can tell that normally a 1D vector is actually a 2D object because we normally omit the magnitude (weight, dimension) of the element in the vector to define its dimension. Basically we have a 1D vector of 1D element making it a 2D object. Or we can represent a vector by a list of [magnitude, position] demonstrating that a vector is a 1D representation of a list of 2D objects.

The devil of this discussion is hidden in the language and concepts.






9
Support - (fb2k) / Re: %CONTENT GROUP% mapping error
Last post by incifinci -
@infinci
I try to give you a logical explanation...
Thank you for your really interesting and valuable explanation. It seems to be right story. Unfortunately, 1. not fully right; 2. it only affects the surface, not the deep waves. As a result, what appears to be true, in the end became false. Half-truths are dangerous things, they deceive people.

Explanation little later, I am busy now.
10
Scientific Discussion / Re: Help me understand why sound is one dimensional
Last post by greynol -
Interesting thread! I'm not a mathematician, just an IT guy with barely enough math knowledge to be dangerous, but hear me out.

It always seemed to me that a waveform is something that can only be approximated. Unless it's generated by a function, that is (like a sine wave.) In music, the waveforms can only be approximated since the underlying sounds are very complex and chaotic.
Have you considered how waveforms not generated by a function are created?  It isn't like they are created by some algorithm used to approximate sound.

The actual complexity of a waveform makes it a very fractal-like structure. As such, it might make more sense to assign a fractal dimensionality to waveforms, which probably lies somewhere between 1 and 2?
Fractal-like?  That's quite a leap.
SimplePortal 1.0.0 RC1 © 2008-2018