Skip to main content

Recent Posts

General - (fb2k) / ADDED_TIMESTAMP - how to use?
Last post by minx_man -

I recently lost my Windows HDD which although a pain isn't the end of the world since I had my music library backed up. The issue is now that i've rebuilt windows and installed foobar2000 from scratch I forgotten how to do various things. Specifically what I'd like to do is have foobar display the "date added" in the playlist, and whilst I can see that most/all of my music files have an ADDED_TIMESTAMP tag I'm not sure how to add that info as a column. Presumably the data would need to be parsed since a typical value "129725982540083078" isn't recognisable as a date. I had this setup before I lost the HDD....

Thanks for reading

If you use guard against clipping and convert tracks separately, it can change loudness difference between tracks in album.
I tested a few tracks and I got these numbers;
Spoiler (click to show/hide)
But if (as DVDdoug said) it's good practice to use the guard option, then I would want to leave it in, better to be safe than sorry. On the other hand, when does the loudness difference caused by this clipping protection become noticable? Is it a cause for worry?
If transition between source tracks is gapless, you can get audible clicks between tracks after resampling.
I didn't hear any clicks between the linked files using f2k, but thanks for the heads up, I will surely check my files after each resample.
WAV files?
I'm using FLAC but still, even 24-bit vs 16-bit takes up twice as much space, so I'm trying to downsample everything that's bigger than CD quality, since that's my most used format anyway.
Theoretically, you wouldn't want to add any noise to vinyl.  But, the dither is very low level compared to the existing vinyl noise so it won't do any harm.
I've used ABX to check if I can notice any difference between the source file and the dithered file and I couldn't hear any difference on dead silent parts (tried both vinyl rips and bandcamp files), even with the volume set to max.
So I think I should leave dither on, to keep up the good practice!
If you have "image + cue"
I store my music as separate files for each song (gaps appended), It made more sense to me when I was not using f2k and had to choose the songs from their folder and it became a habit.
The OP says these are 44.1/24, right? Then ... no resampling!
Yes, It's weird but some of them really came in 44.1/24 instead of 44.1/16.  Also, sorry for my incorrect terminology, what I meant was "dither down" those files. I'm planning to use the same command line for them too since SoX seems to skip the unnecessary resampling process anyway.
BTW, I wonder what would be the reasons to use a less than stellar resampling. You do this once and for all
That's my thinking too, I will delete the source files after I'm done so I better do it right!
The OP says these are 44.1/24, right? Then ... no resampling!
To dither these down to 16bit does not introduce any clicks. Just saying...
General - (fb2k) / Re: New HDD and Mass lossless conversion
Last post by Porcus -
In addition to my reply #9: I would not only check the file count, but also the total time, or total number of samples.
And then I would sort by path - given that you have preserved file structure, that will give you the 37k files on (say) E:\ in path order followed by the 37k files on (say) F:\ in path order.
Different sources here:

* Vinyl rips.
I'd say they "need gaplessness", as a track transition click in the surface noise is likely more annoying than the surface noise itself.
OTOH, they are usually digitized a side at the time.  If you have "image + cue" then resampling the image should avoid all gaplessness issues, eh?
Also, needledrops probably should be volume-normalized (fb2k, Album Gain avoiding peak?).  At least it would be bad practice to record them without headroom.

* Bandcamp hi-rez's:
The OP says these are 44.1/24, right? Then ... no resampling! (I have a few 48k and 96k files from Bandcamp, but if the OP does not ...)
And it has happened to me that a single track on a Bandcamp album is in a different format, but then none of these have required gaplessness anyway. YMMV.

BTW, I wonder what would be the reasons to use a less than stellar resampling. You do this once and for all, it can't cost that much on your electricity bill?
The DLLs are from MSVS 2017 which is DLL-compatible with 2015.
and third party components can easily use dynamic linking too without worrying about user having to install runtimes or bundling huge additional runtime dlls.

What versions of MSVS are supported?
Edit: Wasn't one solution to use "Don't reset DSP between tracks" in foobar? Maybe you have set this and never had to wurry.
It also moves boundaries between tracks, and resulting files will have slightly different durations.

@lvqcl, care to test how well the LPC predictor from Vorbis would mitigate this in your SoX resampler?
Unforunately, I don't have much free time now, so... maybe, but in some unknown future.
General - (fb2k) / Re: Command-Line Decoder Wrapper
Last post by EpicForever -
I have small suggestion. Would it be possible, to have in configuration of the plugin checkboxes next to each format name, similarly to what is present in foo_input_ffmpeg? That would allow for very convenient switching between wrapper / ffmpeg wrapper / native decoding of some formats  just by unchecking / checking boxes.
Yes Case, I tested it recently and it works well, especially in latest version (thing with reporting format names is nice change). However - as I said previously - for most of my collection of some old mpg/av/otheri files I decided to go remux to .mkv, which is natively supported in foobar. Anyway, due to some specific bugs in several files (which caused mkvtoolnix to fail remux operation on them), foo_input_ffmpeg is still a must for me :) . So foo_input_ds and relying on "sometimes broken DirectShow" became obsolete for me.